Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
ARIEL Consortium Phase A Payload Study
ARIEL Payload Design Description
ARIEL-RAL-PL-DD-001 Issue 2.0
Prepared by: Date: Paul Eccleston (RAL Space) Consortium Project Manager
Reviewed by: Date: Kevin Middleton (RAL Space) Payload Systems Engineer
Approved & Date: Released by:
Giovanna Tinetti (UCL) Consortium PI
Page i Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
DOCUMENT CHANGE DETAILS Issue Date Page Description Of Change Comment 0.1 09/05/16 All New document draft created. Document structure and headings defined to request input from consortium. 0.2 24/05/16 All Added input information from consortium as received. 0.3 27/05/16 All Added further input received up to this date from consortium, addition of general architecture and background section in part 4. 0.4 30/05/16 All Further iteration of inputs from consortium and addition of section 3 on science case and driving requirements. 0.5 31/05/16 All Completed all additional sections except 1 (Exec Summary) and 8 (Active Cooler), further updates and iterations from consortium including updated science section. Added new mass budget and data rate tables. 0.6 01/06/16 All Updates from consortium review of final document and addition of section 8 on active cooler (except input on turbo-brayton alternative). Updated mass and power budget table entries for cooler based on latest modelling. 0.7 02/06/16 All Updated figure and table numbering following check. Added comments from KM & GT. Added section 8.4 on other cooler options. 1.0 02/06/16 All Added Executive summary. Prepared for release for MCR datapack. 1.1 23/12/16 All Draft issue to consortium with section authors identified 1.2 10/02/17 All Updated throughout with consortium input for updated designs ready for the MSR submission. 1.3 11/02/17 All Further updates for MSR based on consortium input 1.4 13/02/17 All Added sections on Mechanical, Thermal, detector readout modes, science overview & telescope assy from consortium authors. 1.5 14/02/17 All Added inputs on ACS and AIRS from consortium. Other minor editorial changes. 1.6 15/02/17 All Added comments from EP & KM reviews. Added section on TA thermo-elastic analysis. 1.7 15/02/17 All Final review comments incorporated 2.0 15/02/17 All Document released for MSR datapack
Page ii Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
DISTRIBUTION LIST ARIEL Payload Consortium External Co-PIs Study Engineering Team European Space Working Group Leads Agency Giovanna Tinetti Paul Eccleston Goren Pilbratt Giusi Micela Kevin Middleton Ludovic Puig Jean-Philippe Beaulieu Emanule Pace Astrid Heske Paul Hartogh Gianluca Morgante Enzo Pascale Tom Hunt Ignasi Ribas Vania Da Deppo Hans Ulrik Nørgaard-Nielsen Pino Malaguti Michiel Min Jerome Amiaux Mirek Rataj Pep Colome Bart Vandenbussche Jean-Louis Auguères Manuel Gudel Etienne Renotte David Luz Martin Frericks
Other Engineering Team As necessary for doc
Page iii Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
TABLE OF CONTENTS Document Change Details ...... ii Distribution List ...... iii Table of Contents ...... iv 1 Executive Summary ...... 11 2 Introduction ...... 12 2.1 Purpose ...... 12 2.2 Scope ...... 12 2.3 Applicable Documents ...... 12 2.4 Reference Documents ...... 12 3 ARIEL Science Background ...... 15 3.1 The ARIEL Science ...... 15 3.1.1 Background ...... 15 3.1.2 ARIEL Science Goals ...... 15 3.1.3 Observational Strategy ...... 16 3.2 ARIEL science requirements ...... 17 3.2.1 Wavelength coverage & spectral resolving power ...... 17 3.2.2 ARIEL performances requirements ...... 19 3.2.3 ARIEL core sample, observational strategy & sky visibility ...... 20 4 Payload System Design and Architecture ...... 23 4.1 Payload Architecture & Responsibilities ...... 23 4.2 The Need for Mechanisms in PLM ...... 24 4.3 Design Philosophy ...... 24 4.3.1 Modularity ...... 24 4.3.2 Material Selection ...... 24 4.4 Performance Analysis and Noise Budget ...... 25 4.4.1 Photometric Stability ...... 26 4.4.1.1 Frequency Bands of Interest ...... 26 4.4.1.2 Pointing stability ...... 27 4.4.1.3 Slit Losses ...... 29 4.4.1.4 Detector stability ...... 29 4.4.1.5 Thermal stability ...... 30 4.4.2 Noise Budget ...... 31 4.4.2.1 Contribution from star variability ...... 32 4.4.3 Compliance with requirements ...... 33 4.5 Detector Readout Modes ...... 35 4.5.1 FGS Detectors and Readout Rates ...... 35
Page iv Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
4.5.2 AIRS Detectors and readout rates ...... 36 4.5.3 SIDECAR ASIC as controller electronics ...... 36 5 System Optical Design ...... 37 5.1 Optical System Design Overview ...... 37 5.2 Module Optical Division ...... 41 5.3 Common Optics Design ...... 41 5.3.1 Dichroics ...... 41 5.3.2 Spectrometer Input Optics ...... 43 5.4 On-board Calibration Unit ...... 45 5.4.1 OBCU design ...... 46 5.4.2 Thermal source ...... 47 5.4.3 Electrical Interface ...... 47 5.4.4 Mass and Power Estimation ...... 47 5.5 System Level Straylight Control and Baffling ...... 48 5.6 Optical Budgets and Predicted Performance ...... 48 5.6.1 Overall Throughput Budget ...... 48 5.6.2 Wavefront Error Budget ...... 50 5.6.3 Pupil Alignment Budget ...... 51 6 System Mechanical Design ...... 53 6.1 Payload Module Structure ...... 53 6.1.1 V-Grooves ...... 53 6.1.2 Bipods and supporting struts ...... 54 6.2 Mechanical Interfaces Descriptions ...... 54 6.2.1 Main PLM Interface ...... 54 6.2.2 M1 interface to Telescope Structure ...... 54 6.2.3 Instrument Boxes to Telescope Structure ...... 54 6.3 Telescope Optical Bench Design ...... 54 6.4 Payload Radiator Mechanical Design ...... 55 6.5 Payload Module Mass Budget ...... 55 6.6 Preliminary Payload Module Mechanical Modelling Results ...... 56 7 System Thermal Design ...... 60 7.1 Baseline Thermal Architecture and Design ...... 60 7.2 Thermal Interfaces Definitions ...... 64 7.3 Payload Module Thermal Control Hardware ...... 66 7.3.1 V-Grooves ...... 66 7.3.2 Telescope Baffle and Instrument Radiator ...... 67 7.3.3 Thermal straps ...... 69 7.3.4 Thermistors ...... 70
Page v Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
7.3.5 PLM active thermal control system ...... 71 7.3.5.1 Detectors temperature control stages ...... 73 7.3.5.2 Telescope Primary Mirror temperature control stage ...... 74 7.3.6 Heaters ...... 74 7.4 Thermal Model and Analysis Results ...... 76 7.4.1 G/TMM Description ...... 76 7.4.2 Boundary Conditions ...... 77 7.4.3 Steady-State Analysis Results ...... 78 7.4.4 Harness preliminary thermal analysis ...... 81 7.5 Thermal Budgets ...... 81 7.5.1 Uncertainty Analysis...... 82 8 Active Cooler System ...... 83 8.1 Baseline Cooler Architecture ...... 83 8.1.1 Heat Exchanger Design...... 84 8.1.2 Compressor Set Design ...... 84 8.1.3 Gas Cleanliness and Ancillary Equipment ...... 85 8.1.4 Cooler Control Electronics ...... 86 8.2 Cooler Systems Architecture and Integration to Spacecraft ...... 87 8.3 Baseline Cooler Thermal Modelling and Performance Predictions ...... 88 8.3.1 Modelling Conclusions ...... 89 8.3.2 JT Cooler Development Activities Relevant to ARIEL ...... 90 8.4 Alternative Cooling Configurations ...... 90 8.4.1 40K Turbo-Brayton Cooler ...... 91 8.4.2 Pulse-Tube Coolers ...... 92 8.4.3 Sorption Coolers ...... 92 9 Electrical System Design ...... 93 9.1 Overall Electrical Architecture ...... 93 9.2 Power Budget ...... 93 9.3 Data Rate Budget ...... 93 10 Instrument Control Unit (ICU) Design ...... 96 10.1 AIRS & ICU Electrical Architecture ...... 96 10.2 Instrument Control Unit Electronics Baseline Design ...... 97 10.3 Telescope Control Unit Design Baseline ...... 99 10.4 Cold Front End Electronics System Design ...... 102 10.5 ICU Electrical Interfaces and Budgets ...... 104 10.5.1 ICU Interface Definitions ...... 104 10.6 ICU Mechanical Design ...... 104 10.7 Electrical Ground Support Equipment...... 106
Page vi Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
10.7.1 Overview ...... 107 10.7.2 Functional description ...... 107 10.7.3 The S/C DMS (OBC and SSMM) and PDCU simulator (SIS) ...... 108 10.7.4 Payload Simulator ...... 108 10.7.5 Detectors and ROIC Simulators ...... 108 10.8 ARIEL ICU On-Board Software ...... 109 10.8.1 ICU on-board software description ...... 109 10.8.2 Data processing ...... 110 11 ARIEL Telescope System Design ...... 111 11.1 Telescope Architecture Trade-Off ...... 111 11.2 Telescope Material Trade-Off ...... 111 11.3 Telescope Baseline Design ...... 111 11.4 Telescope Scientific Requirements ...... 112 11.4.1 Telescope FoV Requirements ...... 112 11.4.1.1 Assumptions ...... 112 11.4.1.2 Required FOV for the FGS ...... 113 11.4.1.3 Required FOV for the spectrometer ...... 114 11.4.1.4 Required FOV for the Telescope ...... 115 11.5 Telescope Characteristics ...... 118 11.5.1 Telescope Optical Performance ...... 119 11.5.2 Preliminary Mechanical Design ...... 121 11.5.3 Preliminary Thermal Design and Modelling ...... 122 11.6 Telescope Assembly Design & Accommodation ...... 123 11.7 Telescope Assembly Preliminary Mechanical Analysis Summary ...... 127 11.7.1 Analysis at M1 Mirror Level...... 127 11.7.2 Analysis at Telescope Level – Alignment Position ...... 129 11.7.3 Design limit loads (DLL) at Telescope Level ...... 131 11.7.4 Modal Analysis ...... 133 11.7.5 Preliminary Thermo-Elastic Analysis ...... 134 11.7.5.1 Stress on the bipods ...... 137 11.8 M2 Mirror Mechanism (M2M) Design Baseline ...... 138 11.9 Pathfinder Telescope Mirror Program ...... 139 11.9.1 PTM Plan ...... 139 11.9.2 PTM Progress ...... 140 11.9.2.1 Design and Analysis ...... 140 11.9.2.2 Fabrication ...... 141 11.9.2.3 Polishing Trials ...... 142
Page vii Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
12 ARIEL IR Spectrometer (AIRS) Design ...... 143 12.1 AIRS Design Architecture and functional analysis ...... 143 12.1.1 Spectrometer Architecture Trade-Offs ...... 144 12.2 AIRS Optical Module Design Baseline ...... 145 12.2.1 Optical Design ...... 145 12.2.2 Optical Performance Predictions ...... 147 12.3 Thermal and Mechanical Design ...... 150 12.3.1 AIRS Global Thermal and Mechanical Design ...... 150 12.3.2 AIRS-Focal Plane Assembly Specific Thermal and Mechanical Design ...... 153 12.3.3 AIRS-Focal Plane Assembly Detector ...... 155 12.3.3.1 Baseline US AIRS-Focal Plane Assembly Detector ...... 157 12.3.3.2 Option European AIRS-Focal Plane Assembly Detector ...... 158 12.4 AIRS Observation strategy ...... 159 12.5 AIRS Detection Chain Electronics ...... 165 12.5.1 Electrical System Overview ...... 165 12.5.2 Electrical system description ...... 165 12.5.3 AIRS-FPA Cold Front End Electronics ...... 166 12.5.3.1 Baseline Detector CFEE ...... 166 12.5.3.2 Option European Detector CFEE ...... 167 12.5.3.3 CFEE High level Requirements ...... 167 12.5.4 On-Board Data Processing ...... 168 12.6 AIRS AIV/AIT Philosophy ...... 169 13 Fine Guidance System ...... 171 13.1 FGS Key Requirements ...... 171 13.2 FGS Design Architecture ...... 171 13.3 FGS Optical Module Design Baseline...... 172 13.3.1 Optical Design ...... 172 13.3.2 Dichroic System Definition ...... 173 13.3.3 FGS Stand-alone Optical Performance Predictions ...... 174 13.3.4 NIR-SPEC resolution comparison for FGS with and without 200 nm WFE...... 176 13.3.5 Complete Optical Performance Predictions ...... 177 13.4 Mechanical Design ...... 179 13.5 FGS Detector System ...... 181 13.5.1 Baseline Detector Performance Parameters ...... 181 13.5.2 Other requirements ...... 183 13.5.2.1 Modes of operation of the detectors ...... 183 13.5.3 Detectors available and performance parameters ...... 183 13.6 FGS Control Unit (FCU) Hardware ...... 183
Page viii Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
13.6.1 FCU Design Architecture ...... 183 13.6.2 DPU Block Scheme ...... 184 13.6.2.1 DPU voltage requirements ...... 185 13.6.3 FGS DPU Power and Mass Budget ...... 186 13.6.3.1 FGS DPU Power Budget ...... 186 13.6.3.2 DPU Mass Budget ...... 186 13.6.4 PSU design ...... 186 13.6.5 DCU design ...... 187 13.6.6 FGS FCU Mechanical Description and power consumption ...... 187 13.6.6.1 FGS FCU general requirement ...... 187 13.6.6.2 FGS FCU mechanical information ...... 187 13.6.7 Overall FCU Power Budget...... 188 13.7 FGS Algorithms and Software Design ...... 188 13.7.1 Centroiding Algorithm ...... 188 13.7.2 Photometer Channel Data Processing ...... 189 13.7.3 Data Products and Telemetry ...... 189 13.7.3.1 Data rate in measurement mode...... 189 13.7.4 FGS Predicted Performance ...... 190 14 In-Flight Calibration and Data Processing ...... 191 14.1 In-Flight Calibration ...... 191 14.1.1 The Smooth Transition Philosophy ...... 191 14.1.2 Ground Test Calibration Plan ...... 191 14.1.3 In-Flight Calibration Plan ...... 192 14.1.4 Routine Calibration Phase ...... 193 14.2 On-Board Data Processing Requirements ...... 194 14.3 Ground Data Processing ...... 194 14.3.1 The ARIEL Science Ground Segment ...... 194 14.3.2 ARIEL Science Data Levels and Products ...... 195 14.3.3 ARIEL Science Data Processing Pipeline ...... 196 15 Payload Performance Modelling ...... 197 15.1 ExoSim Overview ...... 197 15.2 ExoSim Structure ...... 197 15.3 Reference Cases and Benchmarking ...... 198 15.4 Results Demonstrating Compliance to Key Science Performance Requirements ...... 199
Page ix Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
1 EXECUTIVE SUMMARY This document presents the design overview of the payload being developed for ARIEL (the Atmospheric Remote-sensing Infrared Exoplanet Large-survey), a candidate mission for the ESA M4 launch opportunity. The proposed ARIEL mission will conduct a large unbiased spectroscopic survey and begin to explore the nature of exoplanet atmospheres and interiors and, through this, the key factors affecting the formation and evolution of planetary systems. ARIEL will observe a large number (>500) of warm and hot transiting gas giants, Neptunes and super-Earths around a range of host star types using transit spectroscopy in the ~1.2 – 8 μm spectral range and broad-band photometry in the optical and Near-IR. We target planets hotter than 600K to take advantage of their well-mixed atmospheres. This document provides an overall summary of the baseline design derived during the phase A study. It gives an overview of the evolution of the design, although full details of the trade studies and alternative options that have been considered and studied are provided in many of the reference documents (also provided in the MSR datapack). Similarly, much more detailed design information and analysis results are available within the extensive collection of reference documents. The baseline integrated payload consists of an all-Aluminium off-axis Cassegrain telescope, feeding a collimated beam into two separate instrument modules. A combined Fine Guidance System / VIS-Photometer / NIR-Spectrometer contains 3 channels of photometry between 0.50 µm and 1.2 µm, of which two will also be used as a redundant system for provided guidance and closed-loop control to the AOCS. One further low resolution (R = ~10) spectrometer in the 1.2 µm – 1.95 µm waveband is also accommodated here. The other instrument module, the ARIEL IR Spectrometer (AIRS), provides spectral resolutions of between 30 – 100 for a waveband between 1.95 µm and 7.8 µm. The payload module is passively cooled to ~55 K by isolation from the spacecraft bus via a series of V-Groove radiators; the detectors for the AIRS are the only items that require active cooling to <42 K via an active Ne JT cooler. The payload mechanical design and the hot case of the thermal model results are shown below.
Page 11 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
2 INTRODUCTION This document presents the design overview of the instrument payload being developed in the frame of the assessment phase study for ARIEL, a candidate mission for the ESA M4 launch opportunity. For a background to the mission and the requirements on the spacecraft and payload see [AD1], [AD2], [AD3] and [AD5].
2.1 PURPOSE This document captures the design information relevant to the payload instrument being studied by a multi- national consortium of European institutes in frame of the assessment phase of the mission. This document contains a complete summary of the payload design information and references to other detailed design documentation where available.
2.2 SCOPE This document provides an overall summary of the baseline design for the ARIEL Payload Module derived during the study. It gives an overview of the evolution of the design to this point, although full details of the trade studies and alternative options that have been considered and studied during the phase A are provided in many of the reference documents. This document concentrates on the technical design aspects of the study output. Although the identification of the driving science requirements and how these impact the design is considered here, the scientific justification for the missions and the derivation of the science requirements is beyond the scope of this document, this is covered elsewhere, in [AD2] and [AD1]. Detailed technical information and trade-off justifications are contained in a number of self-standing technical notes which are referenced from this report. Information on the planning of instrument development program, technology status and planning for the payload module AIV are contained in separate documents, the initial versions of the payload DDVEP [AD4] and AIV [RD23] plans.
2.3 APPLICABLE DOCUMENTS The follow documents are applicable to the content of this document. They are referred to as [ADx] throughout the text. AD # APPLICABLE DOCUMENT TITLE DOCUMENT ID ISSUE / DATE 1 ARIEL Mission Requirements Document (MRD) ESA-ARIEL-EST-MIS- 1.3 RS-001 2 ARIEL Science Requirements Document (SciRD) ESA-ARIEL-EST-SCI- 1.3 RS-001 3 ARIEL Payload Interface Definition – Part A (PID-A) ESA-ARIEL-EST-PL-IF- 0.14 001 4 ARIEL Payload Design, Development, Verification ARIEL-RAL-PL-PL-002 1.0 and Engineering Plan 5 ARIEL Payload Requirements Document ARIEL-RAL-PL-RS-001 1.0 6 Science Operations Assumptions Document (SOAD) ESA-ARIEL-ESAC-SOC- 1.0 AD-001 7 ARIEL Payload Interface Definition – Part B (PID-B) ARIEL-RAL-PL-IF-001 1.0
2.4 REFERENCE DOCUMENTS The table below contains a list of other project documents which are referenced throughout this report. Note that external papers / documents are referenced as footnotes throughout the document and are not allocated an [RDx] number in the text.
Page 12 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
RD # REFERENCE DOCUMENT TITLE DOCUMENT ID ISSUE / DATE 1 Deleted at Issue 1.5 – now not called up from ARIEL-CEA-INST-RS-002 0.3 AIRS chapter 2 AIRS optical design trade-off analysis ARIEL-CEA-INST-DD-001 1.0 3 ARIEL Payload Module Mechanical Analysis ARIEL-MSSL-PL-AN-001 2.0 Report 4 ARIEL Payload Module Finite Element Model ARIEL-MSSL-PL-ML-001 2.0 5 ARIEL Baseline Telescope Optical Prescription ARIEL-RAL-PL-TN-001 3.2 6 ARIEL Telescope Material Trade-Off Analysis ARIEL-INAF-PL-TN-004 2.0 7 ARIEL Payload Consortium Management Plan ARIEL-RAL-PL-PL-001 1.0 8 Deleted at Issue 1.2 – superseded by AD5 ARIEL-RAL-RS-001 9 Alternative Telescope Design ARIEL-RAL-PL-TN-004 1.0 10 Minutes of Pathfinder Telescope Planning ARIEL-RAL-PL-MIN-006 1.0, 7th April Meeting 2016 11 ARIEL Performance Analysis Report ARIEL-CRDF-PL-AN-001 2.1 12 ARIEL Performance Model ARIEL-CRDF-PL-ML-001 2.0 13 Deleted at Issue 1.2 – now included in RD11 ARIEL-CRDF-PL-TN-001 14 ARIEL PLM Thermal Analysis Report ARIEL-INAF-PL-TN-003 2.0 15 ARIEL Dichroic Background Information ARIEL-RAL-PL-TN-003 1.0 16 EChO Baseline Telescope for Phase 0 SRE-F/2012.069 1.0, 22/05/2012 17 ExoSim Comparison to ESA Radiometric Model ARIEL-CRDF-PL-TN-002 1.2 18 ARIEL Targets: Mission Reference Sample ARIEL-UCL-SCI-TN-001 1.0 19 ARIEL Retrievals Technical Note ARIEL-UCL-SCI-TN-002 1.0 20 ARIEL Operations and Calibration Plan ARIEL-RAL-PL-PL-004 1.0 21 ARIEL IOSDC Organisation Plan ARIEL-INAF-GS-PL-001 1.0 22 ARIEL Payload Module MICD ARIEL-MSSL-PL-DRW-001 2.0 23 ARIEL Payload Level Assembly Integration and ARIEL-RAL-PL-PL-007 1.0 Verification (AIV) Plan 24 ARIEL Instrument Control Unit (ICU) Detailed ARIEL-INAF-PL-TN-001 2.0 Design Description 25 ARIEL Mirror Mechanism Requirements ARIEL-SEN-RP-3-001 Rev 1, 2017- Assessment and Design Information 02-08 26 ARIEL Harness Thermal Analysis, Assumptions ARIEL-INAF-TN-005 1.0 and Preliminary Results 27 Telescope Control Unit (TCU) Detailed Design ARIEL-ICE-PL-TN-001 1.0 Technical Note 28 ARIEL ICU Mechanical Interface Control ARIEL-INAF-PL-DRW-001 1.0 Document (MICD) 29 ARIEL FGS Design Description Document ARIEL-CBK-PL-DD-001 2.0 30 ARIEL Wavefront Distortion Analysis ARIEL-CRDF-PL-TN-003 1.2 31 Overall Optical Analysis Report ARIEL-RAL-PL-AN-001 1.0 32 ARIEL Throughput Budget ARIEL-RAL-PL-TN-005 1.0 33 Pathfinder Telescope Mirror Manufacturing and ARIEL-INAF-PL-TN-007 1.0 Preliminary test report 34 Telescope Tolerance Analysis Tech Note ARIEL-INAF-PL-TN-008 1.0 35 Telescope assembly structural and thermo- ARIEL-INAF-PL-TN-006 1.0 elastic analysis report 36 ARIEL Dichroic Study Results and Development ARIEL-DIAS-PL-PL-001 0.1 Plan 37 ARIEL Detector Readout Modes ARIEL-ATC-PL-TN-001 1.0 38 ARIEL Ground Data Pipeline Description ARIEL-INAF-GS-PL-002 1.0 39 ARIEL Ground Calibration Plan ARIEL-RAL-PL-PL-005 1.0
Page 13 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
RD # REFERENCE DOCUMENT TITLE DOCUMENT ID ISSUE / DATE 40 ARIEL PLM Level Optical Ground Support ARIEL-OXF-PL-TN-001 1.0 Equipment (OGSE) Conceptual Design 41 AIRS Detailed Design Description ARIEL-CEA-INST-DD-003 2.0
Page 14 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
3 ARIEL SCIENCE BACKGROUND
3.1 THE ARIEL SCIENCE
3.1.1 Background Thousands of exoplanets have now been discovered with a huge range of masses, sizes and orbits: from rocky Earth-size planets to large gas giants grazing the surface of their host star (Figure 1). However, the essential nature of these exoplanets remains largely mysterious: there is no known, discernible pattern linking the presence, size, or orbital parameters of a planet to the nature of its parent star. We have little idea whether the chemistry of a planet is linked to its formation environment, or whether the type of host star drives the physics and chemistry of the planet’s birth, and evolution. Progress with these science questions demands a large, unbiased spectroscopic survey of exoplanets. The ARIEL candidate mission has been conceived to conduct such a survey and to explore the nature of exoplanet atmospheres and interiors and, through this, the key factors affecting the formation and evolution of planetary systems.
Figure 1: Left: currently known exoplanets, plotted as a function of distance to the star and planetary radii (courtesy of exoplanets.org). The graph suggests a continuous distribution of planetary sizes – from sub-Earths to super-Jupiters – and planetary temperatures (see graph of the right) than span two orders of magnitude.
3.1.2 ARIEL Science Goals ARIEL will address the fundamental questions: – What are exoplanets made of? – How do planets form and evolve? through the direct measurement of the atmospheric and bulk chemical composition. ARIEL will focus on warm and hot planets, for which the atmospheric composition is more representative of the bulk one. ARIEL will observe a large number, i.e. ~ 1000, of warm and hot transiting gas giants, Neptunes and super- Earths around a range of host star types using transit spectroscopy in the ~1.2-8 µm spectral range and multiple-band photometry in the optical. We target in particular warm and hot planets to take advantage of their well-mixed atmospheres which should show minimal condensation and sequestration of high-Z materials and thus reveal their bulk and elemental composition (especially C, O, N, S, Si). Observations of these hot exoplanets will allow the understanding of the early stages of planetary and atmospheric formation during the nebular phase and the following few millions years. ARIEL will thus provide a truly representative picture of the chemical nature of the exoplanets and relate this directly to the type and chemical environment of the host star (Figure 2).
Page 15 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 2: Schematic summary of the various classes of atmospheres as predicted by Forget & Leconte (2014)1. Only the expected dominant species are indicated, other (trace) gases will be present. Each line represents a transition from one regime to another, but these “transitions” need tight calibrations from observations. The axes do not have numerical values as they are unknown. Solar System planets are indicated, together with a lava planet, an Ocean planet and a hot Jupiter. ARIEL will observe planets ranging from the Earth to the super-Jupiter masses, especially warm and hot ones: many atmospheric regime transitions are expected to occur in this domain.
3.1.3 Observational Strategy For this ambitious scientific programme, ARIEL is designed as a dedicated survey mission for transit, eclipse and phase-curves spectroscopy, capable of observing a large and well-defined planet sample within its 3.5- year mission lifetime. Transit and eclipse spectroscopy methods, whereby the signal from the star and planet are differentiated using knowledge of the planetary ephemerides, allow us to measure atmospheric signals from the planet at levels of ~10-100 ppm relative to the star (post-processing) and, given the bright nature of targets, also allows more sophisticated techniques, such as phase curve analysis and eclipse mapping, to give a deeper insight into the nature of the atmosphere (Figure 3). Transit spectroscopy means that no angular resolution is required and detailed performance studies show that a 1-metre class telescope is sufficient to achieve the necessary observations on all the ARIEL targets within the mission lifetime. The satellite is best placed into an L2 orbit to maximise the thermal stability and the field of regard. To maximize the science return of ARIEL and take full advantage of its unique characteristics, a three-tiered approach has been considered, where three different samples are observed at optimised spectral resolutions, wavelength intervals and signal-to-noise ratios. A summary of the survey tiers is given in Table 2 of the SciRD document [AD2] and a possible implementation is described in the Technical Note by Zingales et al., [RD18].
1 Forget, F., Leconte, J., Phil. Trans. Royal Society 372, #20130084 (2014)
Page 16 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 3: Methods adopted by ARIEL to probe the exoplanet composition and structure. Left: orbital lightcurve of the transiting exoplanet HAT-P-7b as observed by Kepler (Borucki et al., 20092). The transit and eclipse are visible. Right: slice mapping with ingress and egress maps as well as a combined map of HD189733b at 8 µm. These were achieved with Spitzer (Majeau et al., 20123).
3.2 ARIEL SCIENCE REQUIREMENTS
3.2.1 Wavelength coverage & spectral resolving power
Figure 4: Left: Molecular signatures in the 1-10 μm range at the required spectral resolving power proposed for ARIEL (R=100). Right: cloud signature in the 0.5-2.5 μm range: ARIEL will measure simultaneously the relative contributions of the “blue” and “red” filters in the visible, 1 NIR filter and the spectral contribution in the 1.25-7.8 μm range. Through these measurements ARIEL will detect the presence of clouds/hazes, and constrain cloud parameters such as altitude, thickness, particle-size. To fulfil the science requirements, ARIEL will be a specifically designed, stable payload and satellite platform with broad, instantaneous wavelength coverage to detect many molecular species, probe the thermal structure, identify/characterize clouds and monitor the stellar activity. The wavelength range considered
2 Borucki, W.J., et al., Science, 325, 709, 2009 3 Majeau, C., Agol, E. and Cowan, N. B., ApJ 747, id L20, 2012.
Page 17 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
covers all the expected major atmospheric gases from e.g. H2O, CO2, CH4 NH3, HCN, H2S through to the more exotic metallic compounds, such as TiO, VO, and condensed species (Figure 4). Wavelength range Resolving power Scientific motivation • Correction stellar activity (optimised early stars) Blue filter Integrated band • Measurement of planetary albedo 0.5 – 0.55 µm • Detection of Rayleigh scattering/clouds • Correction stellar activity (optimised late stars) Red filter Integrated band • Measurement of planetary albedo 0.8 – 1.0 µm • Detection of clouds NIR1 filter • Correction stellar activity (optimised late stars) Integrated band 1.05 – 1.2 µm • Detection of clouds • Correction stellar activity (optimised late stars) • Detection of clouds • Detection of molecules (especially TiO, VO, metal hydrides) NIR2 filter (spectrograph) R=10 • Measurement of planet temperature (optimised hot) 1.25 – 1.95 µm • Retrieval of molecular abundances • Retrieval of vertical and horizontal thermal structure • Detection temporal variability (weather/cloud distribution) • Detection of atmospheric chemical components • Measurement of planet temperature (optimised warm- hot) IR spectrograph – 100-200 • Retrieval of molecular abundances 1.95 – 7.8 µm • Retrieval of vertical and horizontal thermal structure • Detection temporal variability (weather/cloud distribution) Table 1: Summary of the ARIEL spectral coverage (left column) and resolving power (central column). The key scientific motivations are listed in the right column.
Page 18 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 5: Posterior distributions of various atmospheric trace gases, temperature and top cloud pressure obtained with the TauREx model (Waldmann et al., 20154). by retrieving the simulated spectra shown on the right, as observed by HST/WFC (green), JWST (red) and ARIEL (orange). Dashed lines in the histogram plots show the 1 sigma confidence intervals. The gases and atmospheric parameters retrieved include H2O, CO2, CO, CH4 and NH3, temperature, radius and derived mean molecular weight. There is still considerable degeneracy for retrievals from WFC3 spectra, but ARIEL and JWST data are expected to be very constraining. The information content for JWST and ARIEL in the case of bright sources is comparable, indicating that ARIEL will be able to characterize atmospheres to a similar degree of accuracy.
3.2.2 ARIEL performances requirements ARIEL’s top-level requirement is that the photometric stability over the frequency band of interest shall not add significantly to the photometric noise from the astrophysical scene (star, planet and zodiacal light). The frequency band over which the requirement applies is between 2.8×10-5 to ~3.7x10-3 Hz, i.e. between ~ 2
4 Waldmann, I.P. et al., 2015. Tau-REx I: A next generation retrieval code for exoplanetary atmospheres, ApJ, 802, 107
Page 19 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 minutes and 10 hours (see [RD11]). This implies having the capability to remove any residual systematics and to co-add the elementary observations from many repeat visits to a given target.
3.2.3 ARIEL core sample, observational strategy & sky visibility ARIEL will study a large population of hot and warm planets, already discovered by other facilities. In particular, it will focus on hundreds of gaseous objects (Jupiters, Saturns, Neptunes) and of super-Earths and sub-Neptunes around bright stars. There are over 200 currently known planets complying with these requirements. The current 200 known targets have been discovered mainly close to the ecliptic plane because provided by ground-based surveys, as shown in Figure 6, illustrating also the sky visibility for ARIEL. K2, Cheops and NGTS are expected to complete the search for planets around bright sources closer to the ecliptic plane. TESS and PLATO will extend the planet search closer to the ecliptic poles, which are where ARIEL has continuous coverage.
Figure 6: A plot illustrating the fraction of the year for which a given location in the sky (in equatorial coordinates) is visible to ARIEL, as seen from a representative operational orbit of ARIEL at L2. Red and green targets are the currently known best targets in term of stellar brightness and planetary parameters (green are the very best, including e.g. 55 Cnc e, HD 189733b, HD 209 458 b, GJ 436 b etc.), yellow targets are currently known transiting planets around stars observable by ARIEL, brighter sources would allow more efficient use of the ARIEL life-time.
To generate a core mission sample to be observed by ARIEL in 2026 during 3.5 years, a list of targets with different stellar types (F, G, K, M) and planetary parameters (size: Jupiters, Neptunes, sub-Neptunes, super- Earths, Earth-size; different temperatures) has been created. Said list was compiled using the statistics provided by the NASA Kepler mission combined with the number/types of stars in the Solar neighbourhood (see [RD18]). The required number of transits/eclipses to achieve the SNR/R needed to perform an accurate retrieval of the gas abundances and thermal properties has been calculated using the ESA Rad Model (Puig et al., 20145) and ExoSim (see [RD12]). These preliminary calculations have indicated that ~ 1000 planets can be observed during the mission lifetime (3.5 years) with the required SNR/R (see [RD18]) given in Table 2 of the SciRD document, see Figures 7, 8 & 9.
5 Puig, L. et al., The phase 0/A study of the ESA M3 mission candidate EChO, Experimental Astronomy 42, 2014
Page 20 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 7: Overview of the ARIEL Target sample. The graph shows how the ARIEL 3 Tiers, i.e. Reconnaissance Survey, Deep Survey and Benchmark planets are nested (see [RD18]).
Figure 8: Number of planets in the Reconnaissance Survey (Tier 1, top) and Deep Survey modality (Tier 2, bottom) divided in size bins. The plots indicate as well the % ARIEL lifetime needed to observe those planets and the number of transits/eclipses needed to observe a single planet. See [RD18].
Page 21 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 9: Tier 1 planets plotted as a function of size versus planetary density (top) and planetary versus stellar temperature (bottom). See [RD18].
Page 22 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
4 PAYLOAD SYSTEM DESIGN AND ARCHITECTURE
4.1 PAYLOAD ARCHITECTURE & RESPONSIBILITIES The baseline architecture for the ARIEL payload is illustrated in Figure 10 below. This diagram also shows the nationalities of the members of the payload consortium who are taking responsibility for each element. Further details of the consortium organisation can be found in the Consortium Management Plan, [RD7].
Figure 10: ARIEL Payload Schematic and Reponsibilities
A functional block diagram of the payload with a focus on the optical path is shown in Figure 19. The baseline architecture splits the payload into two major sections, the cold payload module (PLM) and the items of the payload that mount within the spacecraft service module (SVM). The major items are: • Cold PLM:
o Telescope system, incorporating M1, M2 & M3 mirrors, a re-focusing mechanism on the M2 mirror (M2M), the telescope structure and baffles.
o An optical bench / metering structure onto which both the telescope items and the other instruments are mounted.
o A set of common optics including fold mirrors for packaging, the dichroics to split the FGS and spectrometer light, formatting optics to inject the light into the spectrometer correctly, and a common calibration source for the payload.
o The ARIEL IR Spectrometer (AIRS), including all optics and structure plus detector and cold front end electronics (cFEE).
Page 23 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
o Fine Guidance Sensor / Visible Photometer / Near-IR Spectrometer (FGS/VISPhot/NIRSpec), including all optics and dichroics to split into the 4 separate channels, prime and redundant detectors and cold front end electronics.
o Thermal hardware: active cooler coldhead for Neon JT cooler, passive radiator for cooling of FGS detectors and all cFEE, V-grooves and support structure to isolate the cold PLM from the warmer SVM and solar thermal loads. • Warm SVM mounted units:
o Instrument Control Unit (ICU) housing the AIRS warm front end electronics (DCU), the central data processing unit (DPU) for the spectrometer data, a power supply unit (PSU) and the telescope control unit (TCU) for M2M mechanism control and thermal monitoring and control of the telescope.
o FGS Control Unit (FCU) electronics incorporating the FGS/VISPhot/NIRSpec wFEE, the control and processing electronics and software for determining the pointing from the FSG data and transmitting this information to the spacecraft.
o Active cooler system: . Cooler control electronics . Cooler compressors . Cooler gas handling panel incorporating fill connections, filtering etc.
4.2 THE NEED FOR MECHANISMS IN PLM The baseline design of the PLM includes one active mechanism in the payload. This is the refocusing mechanism on the telescope M2 mirror, used for ensuring alignment and image quality of the telescope system is good after launch and cooldown. The previously baselined inclusion of a tip-tilt mechanism on the M4 mirror to be used to close the pointing loop has been removed in the second part of phase A. The jitter analysis of the payload has demonstrated that the pointing requirements that are achievable by the S/C are more than enough to allow the photometric stability and noise requirements to be met (see section 4.4 below and [RD11] for details) The baseline mechanisms are using heritage from previous studies and programs that have been undertaken by commercial companies under ESA contracts (TRP etc). The baselined consortium partner for the M2M (Spain) have heritage from the Gaia and Euclid telescope re-focusing mechanisms. See section 11.1 for more details.
4.3 DESIGN PHILOSOPHY
4.3.1 Modularity The baseline design architecture has been selected to maintain a high degree of modularity in the design. This helps both technically and programmatically in allowing independent development of the instrument / module / subsystem designs and in giving the maximum flexibility for future changes. To this end the optical design of the modules are largely decoupled from one another by the selection of a either a collimated (nominally) aberration free beam or a focus as the interface between modules. A common optical bench (separate from the structures of the individual modules and the telescope primary mirror) has been selected as the design baseline. The assumption of responsibility for the OB by the consortium team allows the instruments and telescope to be built, assembled, aligned and tested as a unit and pre-calibrated prior to delivery to ESA. The co-alignment of the modules (critical to the success of the mission due to the shared field of view) can then be assured and checked at the earliest possible stage.
4.3.2 Material Selection The design baseline is that all of the cryogenic components of the payload architecture are manufactured from a common material, Aluminium 6061 alloy. This ensures that the design has a matched CTE, allowing warm alignment of the payload to proceed, with a high degree of confidence that this will be maintained Page 24 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 when cooled to operating temperatures. All reflective optical elements in the design are also assumed to be manufactured from Aluminium for the same reason. This builds on the significant design heritage within Europe of building all Aluminium space instruments for cryogenic operation such as Herschel SPIRE & PACS and JWST MIRI. Building in Aluminium also provides a robust design approach with minimum technical risk (and straight-forward options for rework if and when this becomes necessary during AIV activities, unlike with other options such as SiC). Details of the trade-off conducted are contained in [RD6].
4.4 PERFORMANCE ANALYSIS AND NOISE BUDGET We have investigated the overall science performance of the ARIEL Payload. The outcome of this study is summarised here, and discussed in detail in the Performance Analysis Report [RD11]. Transit spectroscopy and multi-band photometry has been so far conducted using general-purpose, space-based instruments. Some success has been obtained using ground-based instrumentation as well. These measurements however suffer from a high level of systematic error due to a number of issues such as pointing jitter, thermal and optomechanical stability, wavelength and photometric calibration, and detector stability. We have designed ARIEL to be an instrument to perform time series spectroscopy and photometry from the visible to the mid-IR with a stability of better than 100ppm during one transit observation. Key aspects which allow ARIEL to obtain its stable performance are: 1) Simultaneous observations of the same transit event by all photometric and spectroscopic channels; 2) Continuous observation of the transit event such that the measurement is conducted in a thermally and photometrically stable condition. 3) A payload design which makes ARIEL resilient to major sources of systematics or make it possible their removal in post processing.
This is summarised in the table below, which lists the most important sources of noise and systematics identified, along with the approach used to mitigate their impact on the detection, and on the overall photometric stability. Type of uncertainty Source Mitigation Strategy Detector noise Dark current noise Choice of low-noise detectors Readout noise Gain stability Calibration, post- processing data analysis, choice of stable detectors. Persistence Post-processing decorrelation. Continuously staring at a target for the whole duration of the observation. Thermal noise Emission from telescope, Negligible due to surface emissivity properties and common optics and all optical in-flight temperatures of the payload. elements Temperature fluctuations in time Negligible impact by design Astrophysical noise Photon noise arising from the Fundamental noise limit, choice of aperture size (M1 target diameter). Photon noise arising from local Negligible over ARIEL band zodiacal light Stellar variability with time Multi-wavelength stellar monitoring, post- processing decorrelation Pointing jitter RPE and PDE effects on the Small RPE and PDE, Nyquist sampling, post- position, Spectral Energy processing decorrelation Sistribution, and detector intra/inter pixel response Slit losses Spectrometer input slit sufficiently large Table 2: Summary of noise sources and systematic errors
Page 25 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Below we demonstrate how the ARIEL design is capable of achieving its photometric stability requirements using time domain simulations and an advanced data reduction pipeline. We use the end-to-end ExoSim simulator to comprehensively model the ARIEL performance. The simulator implements a detailed instrument model and generates photometric and spectroscopic lightcurves similar to those ARIEL will provide during science operations. An overview of ExoSim in provided in Section 15 with more details in [RD11, RD12] and references therein. Its timelines are processed with a data reduction pipeline to estimate the magnitude of noise processes and systematics, to reconstruct the spectral light curves and study the overall performance. Details about the pipeline implementation are also in [RD11].
4.4.1 Photometric Stability The photometric stability of the ARIEL system is critical to achieving our science goals. Using ExoSim it is possible to assess of all aspects of the payload and satellite stability in order, among other things, to establish the optimum observing and calibration programme for the mission. This complex task requires the knowledge and resources of the scientific and technical members of the consortium and ESA. The photometric stability of the instrument throughout consecutive observations lasting several hours is mainly governed by the source of uncertainties listed in Table 2 and discussed below using three model stars: HD219134 (brightest target, AD1, R-PERF-020), HD209458 (bright target, R-PERF-025), and GJ1214 (faintest target, AD1, R-PERF-010).
4.4.1.1 Frequency Bands of Interest The observing strategy of ARIEL is that of time resolved spectroscopy and photometry. By obtaining series of time consecutive spectra ARIEL will trace the transit/eclipse event of an exoplanet through time, yielding a lightcurve of the transit/eclipse for every single spectral resolution element. Figure 11 (a) shows an example of a secondary eclipse of a typical hot-Jupiter planet. Figure 11 (b) shows the signal observed by ARIEL over the duration of 6 planetary orbits of a hot-Jupiter. From these figures it can easily be seen that time- correlated noise has the greatest impact on the retrieved science at temporal variation frequencies comparable to those of the transit/eclipse event, or a multiple thereof. Figure 11 (c) shows the frequency domain representation of Figure 11 (b) given a variety of orbital periods. It is clear that the desired signal is contained in discrete frequencies and their respective overtones. It is also apparent that frequency ranges beyond these shown in Figure 11 (c) are of no concern to the science objective and can safely be filtered out using pass-band filters without impairing the shape or amplitude of the lightcurve feature. This leads to the concept of ‘crucial frequency bands’ within the photometric stability must be kept at a level defined in the SCIRD R-SCI-210 (and MRD R-PERF-160) to ensure the success of the mission. The example of Figure 11 finds the ‘crucial frequency band’ to be from 1.9×10-4 to ~1.7x10-3 Hz (compatible with R-SCI-210), outside of which slow moving trends and high-frequency noise can effectively be filtered. The overall critical frequency band for ARIEL is determined by the longest observation expected and the need to Nyquist- sample the -5 -3 -2 highest expected frequencies. It is specified in R-PERF-160 as between 2.8x10 Hz and 3.3x10 Hz (1.1x10 Hz) for faint (bright) targets. Photometric stability over longer periods is not required as multiple transits of the same target can be stacked together after these are calibrated to the level required in R-PRD-0520, R- PRD-0530, R-PRD-0550 (absolute calibration).
Figure 11: (a) Left: Secondary eclipse lightcurve of a hot-Jupiter type exoplanet with eclipse duration of 720 min. Noise at 10-4 level was added. (b) Centre: Time series of 6 orbits of a hot-Jupiter (akin to HD189733b). The deep troughs are limb-darkened transits, smaller troughs are secondary eclipses and sinusoidal variations are due to the planetary phase curve as the planetary day-side rotates in and out of view. White noise of the level of 10-4 was added. (c) Right: Power spectra of time series shown in (a) for different orbital periods. Blue: Period = 120 days, Green = 2.21 days (akin to HD189733b), Red = 0.4 days. The sensitive frequency range extends from 1.9×10-4 - 1.7x10-3 Hz.
Page 26 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
4.4.1.2 Pointing stability Pointing stability of the telescope is quantified in terms of Absolute Performance Error (APE), Performance Drift Error (PDE) and Relative Performance Error (RPE) for AOCS solution considered in the study. These pointing drifts manifest themselves in the observed data product via two mechanisms: 1) the drifting of the spectrum along the spectral axis of the detector, from here on referred to as ‘spectral jitter’; 2) the drift of the spectrum along the spatial direction (or ‘spatial jitter’). The effect of jitter on the observed time series is the introduction of noise, characterized by the power-spectrum of the telescope pointing. It is correlated in time, if the power-spectrum is non-white. The amplitude of the resultant photometric scatter depends on the amount of spectral/spatial displacement of the spectrum, the monochromatic PSF of the instruments, the detector intra-pixel response and the amplitude of the inter-pixel variations. Slit losses are made negligible by adopting wide enough slits at the input of each spectroscopic Pointing jitter as a source of photometric error operates on two different time-scales. When a detector is read by correlated-double-sampling (CDS), pointing jitter only affects significantly the second non- destructive read (NDR). Therefore, the relevant frequency is set by the exposure rate. Moreover, because jitter is due to a random motion of the overall spacecraft, it affects all illuminated pixels in the same way, and it is therefore 100% correlated at all wavelengths in every channel. A detailed study of photometric uncertainties arising from pointing jitter is provided in [RD11]. Main findings and design solutions are summarised here: • Jitter frequencies faster than the exposure rate have the net effect of enlarging the effective monochromatic PSF at the focal plane. The size of the PSF changes slightly from exposure to exposure, but the total energy collected in the unit time does not incur in significant losses. Therefore, this source of photometric error is negligible when compared to other experimental uncertainties. • Jitter frequencies slower than the exposure rate are a non-negligible source of photometric errors. In the case of spatial-jitter, photometric uncertainties arise from the combined effect of a wobbling spectrum sampled in presence of intra- and inter-pixel variations. In the case of spectral-jitter, uncertainties arise mostly by the modulation jitter imposes on the spectrum in the focal plane dispersion direction. Compared to spatial-jitter, spectral-jitter can result in larger photometric instabilities, and if left uncorrected can severely impair the quality of the final science result. The effect is particularly significant in the case of spectral jitter in areas of steep flux gradients whether due to stellar lines or black-body variations. • Jitter photometric uncertainties are proportional to the source intensity. Assuming that the exposure rate is set such as to fill the detector well depth (WD), then the photometric uncertainty is proportional to the detector WD times the usable bandwidth of the jitter, which decreases with increasing integration time. This translates into the fact that pointing jitter is more relevant for bright sources (short integrations) than it is for faint targets (longer exposures. • As discussed in [RD11] the effect can be mitigated by a combination of instrument design, AOCS stability and post-processing analysis which takes advantage of the correlated nature of this systematic component. • Spatial-jitter can be made negligible by requiring that signals (photometer + spectrometer) are fully Nyquist sampled. This results in a monochromatic PSF with widths spanning two or more detector pixels such that intra-pixel responses have no significant impact on the photometric stability budget. Inter-pixel variations do have an impact which is made negligible by flat-fielding quantum-efficiency variations in post-processing to the required level (~0.5%). • Spectral-jitter can be decorrelated in post-processing to such a level which results in a negligible contribution to the total noise budget (usually less than 10% of the photon noise variance of the star). This can be done by either using the spectral information alone (effectively correlating one exposure with all other exposures to estimate the shift) or by using the information provided by the FGS.
Page 27 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 12: Photometric uncertainties from pointing jitter (green and pink solid lines) compared to photon noise variance from the source (blue line) for bright and faint sources in AIRS CH0 and CH1. The total noise variance includes here all noise sources such as photon, readout, dark current, jitter noise(dashed line). Left panels show the raw effect. Right panels show the reduction obtained after decorrelation of jitter noise The analysis is summarised in Figure 12. The pointing jitter used is that specified by R-AOCS-030 in the MRD for bright and faint targets. It is interesting to note how effective decorrelation is in reducing the effects of Page 28 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 spectral jitter, by close to two orders of magnitude in noise variance. The same is not observed for spatial jitter. For a perfect focal plane with no QE variations, spatial jitter noise would be absent. For a real focal plane, intra-pixel variations have negligible effect when the signal is spatially Nyquist sampled. Therefore, the major contribution comes from QE variations across the focal plane (inter-pixel variations) which cannot be mitigated by the decorrelation technique used. It is possible to implement more sophisticated data analysis techniques to reduce even further the residuals (e.g. Independent Component Analysis, use of information from all focal planes at the same time, etc.). However, this is left to subsequent phases in the project. It has not been investigated here as the simple decorrelation technique employed results sufficient to demonstrate compliance with requirements.
4.4.1.3 Slit Losses Both AIRS channels make use of an input slits, while photometric channels and NIRSpec make no use of a slit. Slit losses in AIRS channels are negligible because the slits are by design sufficiently wide to result in negligible loss of throughput in presence of APE, RPE and PDE. This is shown in Figure 13. The important here is not the absolute value of the slit loss, which depends from the APE, but the modulation. Therefore, noise depends from the gradient of the slit loss and it is always less than 10ppm (68% C.L), well within requirements.
Figure 13: Slit losses as a function of the absolute pointing error are show for different wavelengths (color coded) of AIRS CH1. Bottom panels show the estimated photometric stability achievable vs observing time for bright (left panel) and faint (right panel) AOCS modes. Line styles code absolute pointing error chosen while colors code wavelength as in top panel. The required photometric stability is always achieved, even at short integration times and in AIRS CH0, not shown here.
4.4.1.4 Detector stability The effects of temperature drifts on the detector dark current stability is discussed next. Once temperature controlled, MCT detectors are known to manifest high QE stability. Studying the Teledyne H1RG, Bezawada & Ives6, find an upper limit of 500ppm over several hours, limited by the stability of the
6 Nagaraja Bezawada and Derek Ives, "High-speed multiple window readout of Hawaii-1RG detector for a radial velocity experiment", Proc. SPIE 6276, High Energy, Optical, and Infrared Detectors for Astronomy II, 62760O (June 15, 2006); doi:10.1117/12.670244
Page 29 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 source used in their measurements. More recently Clanton et al.7 have demonstrated an upper limit of 50 ppm (also limited by the experimental setup used in their measurements) with indications that the stability can be as low as 10-20ppm. This exceeds ARIEL requirements of 100ppm.
4.4.1.5 Thermal stability Emission from optical surfaces are negligible in ARIEL when compared to the signal of even the faintest target (R-PERF-010). While the detector characteristics vary as a function of its operational temperature, the FPA enclosure, and therefore the detector, is operated at a sufficiently low and stable temperature, eliminating this variability. To verify this, the relevant derived requirements which are captured in [AD5] are: R-PRD-1200 Telescope 70K ±1K (±5K) R-PRD-1210 Optical Bench 55K ±1K (±5K) R-PRD-1240 AIRS & FGS Optics 55K ±1K (±5K) R-PRD-1250 FGS detectors 70K ± 0.05K R-PRD-1260 AIRS detectors 40K ± 0.05K
All variations intended peak-to-peak. A variation in temperature of any of the optical elements result in its thermal emission to be modulated. The major effect in the temperature variation of the detector is that of a changing dark current. This is shown in Figure 14. Temperature variations assumed as those in the table above, using the more pessimistic values for optical elements in brackets above to demonstrate that ARIEL is insensitive to these effects). Figure caption explains. The detector model used for these simulations is a Teledyne with a cutoff wavelength at 10.6 micron, operated at a temperature of 42K (see McMurtry, C., et al., 2013 for details and their Figure 5). This is an overly pessimistic case for ARIEL, yet the photometric variations induced are negligibly small when compared to the flux of the even the faintest target.
Figure 14: This diagram shows the peak-to-peak signal variation relative to the faintest target star signal induced by temperature fluctuations. Instrument emission variations correspond to ΔT = 10K peak-to-peak of all optical elements (telescope, filters, prism, etc.) in the instrument light path. Inner sanctum variations are estimated for ΔT = 2K peak-to- peak. Dark current variations assume an identical detector for both AIRS channels and ΔT = 100mK peak-to-peak. For
7 Clanton et al, PAASP, 124, 713, 2012
Page 30 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
AIRS CH0 this is an overly pessimistic estimate. The lower cut-off wavelength of AIRS CH0 and the low dark current of the H1RG would result in a negligibly small dark current fluctuation noise in this band. All noise source are below 2% peak-to-peak indicated by the grey horizontal line (less than 1% RMS). NIRSpec and photometric channels are not shown because effects are too small.
4.4.2 Noise Budget The overall noise budget is shown in Figure 15 which gives the current best estimate of the complete instrument noise performance from ExoSim for every sizing stars in the MRD (R-PERF-010, R-PERF-2020 and R-PERF-025). All cases are photon noise limited by the target star. The following general considerations can be made. Pointing jitter is the second most dominant noise component in NIRSpec and at the blue end of the AIRS channels. Detector dark currents used are 1, 30 and 50 e/s in VNIR, and AIRS CH0 and CH1, respectively. Their associated Poisson noise is important when observing faint targets (VISPhot and AIRS). It should be noted that Teledyne detectors have significantly less dark current than assumed here, therefore this result is pessimistic. Detector readout noise variances presented are based on a simple CDS and are the second most important noise component at the red-end of AIRS for bright targets and for some of the photometers. It should be noted that this is an overly pessimistic assumption when observing targets such R-PERF-010 (faintest) and R-PERD-025 (bright) as the detector will be read up-the-ramp with slope fitting. This would result in negligible readout noise in these cases, but here CDS are shown as worst case. All other noise sources are negligible.
Page 31 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 15: ARIEL post-processing noise budgets. Panels from top to bottom show the budget when observing the brightest, bright and faintest target (R-PERF-020, R-PERF 025, R-PERF-010), respectively.
4.4.2.1 Contribution from star variability
Page 32 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
ExoSim can simulate the time domain signature from stellar effects such as convection-driven variability and spots. Details of models implemented can be found in [RD11] where results from convection-driven results (pulsation and granulation) are fully detailed. For this type of variability, a solar type star and an M type star have been simulated to bracket the extremes expected in the ARIEL target list. It is found that across all ARIEL wavelengths the effect is always smaller than 50ppm as shown in the table below. This translates to less than 1% of target star photon noise variance in the raw timeline, leading to the conclusion that this type of stellar noise is negligible for ARIEL photometry and spectrometry and decorrelation of the effects (although possible because the high correlation across all bands) is not necessary to meet requirements (< 10% RSS, R-PERF-190). The equivalent analysis for stellar spots is included in the yellow book. Table 3: Stellar convection noise/signal as ppm (for a channel or spectral bin). Median values shown for spectroscopic channels. Signal is the mean count for a single exposure, and noise is the standard deviation of the signal resulting from stellar variation. Channel GJ1214 HD209458 AIRS Ch1 15.4 9.6 AIRS Ch0 24.5 9.8 NIRSpec 32.2 11.2 FGS2 36.6 19.6 FGS1 43.6 21.7 VISPhot 13.2 40.5
4.4.3 Compliance with requirements Compliance against the MRD requirements in R-PERF-160 where the total noise variance model is defined as ( 0 + ) × (1 + ) +
The parameters in the model are given𝑉𝑉𝑉𝑉𝑉𝑉 as≤ follows:𝑁𝑁 𝑧𝑧𝑧𝑧𝑧𝑧𝑧𝑧 𝑋𝑋 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁
Channel Wavelength X Nmin (micron) [e-/s/spectral bin] VPhot 0.5 – 0.55 0.3 400 FGS1 0.8 – 1.0 0.3 400 FGS2 1.05 – 1.2 0.3 400 NIRSpec 1.2 – 1.95 0.3 = 17x λ3 AIRS-0 1.95 – 3.9 0.2 = 20 x λ3 AIRS-1 >3.9 0.2 = 5 x λ3
Figure 16 compares the values of the parameters X and Nmin found with ExoSim with those in R-PERF-160. We find that the instrument always performs within requirements and often outperform requirements significantly. Recalling that read noise components are estimated here assuming CDS, while faint and bright targets will be read up-the-ramp with slope fitting, implies that AIRS read noise are pessimistic for these cases. Similar information is presented in Figure 16is shown in Figure 17 in terms of total noise variance and relative noise power spectral density for direct comparison with equivalent figures in MRD. Any deviation from compliance in the noise variance panel of Figure 16 is misleading and not correct. This because, for instance, larger noise than requirements is good, if it is photon noise from the target. Therefore the relevant quantity for compliance is the relative noise power density panel in Figure 17, showing that the ARIEL design is fully compliant at all wavelengths.
Page 33 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 16: top panel – X MRD requirements (median and maximum values) are shown by the thick horizontal grey lines. The total X-values estimated with ExoSim are shown by the solid coloured lines for each sizing star. The jitter and detector read noise (assuming CDS) components of X are shown respectively by the star and circle symbols. Cyan rectangles mark the medians of X in each band for all sizing stars showing that both median and maximum X values fall below requirements. Bottom panel – Nmin MRD requirements are shown by the thick grey lines while the black circles show values estimated with ExoSim. The trapezoids in each panel show the location of ARIEL’s photometric and spectroscopic bands.
Figure 17: Top panel – total noise variance in one second of integration. MRD requirements are shown by solid lines, which for the spectrometers correspond to MRD median X-values. Dashed lines show the requirements corresponding to MRD maximum X-values. ExoSim estimates are shown by data points. Bottom panel – relative noise (the ratio of instrument noise to the star signal) estimates of ExoSim are shown by the solid lines. MRD requirements are marked by dashed lines which for the spectrometers correspond to MRD median X-values. Dot-dashed lines show the requirements corresponding to MRD maximum X-values. Trapezoids in both panels mark the location of ARIEL’s photometric and spectroscopic bands, these are not the requirements.
Page 34 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
4.5 DETECTOR READOUT MODES
4.5.1 FGS Detectors and Readout Rates The Baseline detectors for the FGS system are the standard substrate removed H1RG, 2.5µm cut-off detectors from Teledyne, see section 13.5 for details. The FGS system will use two such detectors; one detector is used for FGS1 and VIS-Phot channels and another detector for FGS2 and NIR-Spectrometer channels. The Table below shows the required window sizes to be readout for different modes of FGS. The maximum window sizes and the readout rates are expected in auto-calibration mode. Maximum window sizes and readout Typical window sizes and readout Channel rates in auto-calibration mode rates in measurement mode FGS1 150x150 pixels at 10Hz CDS 30x30 pixels at 10Hz CDS VIS Phot 150x150 pixels at 10Hz NDR 30x30 pixels at 10Hz NDR 30x30 pixels at 10Hz CDS (for bright FGS2 180x180 pixels at < 10Hz CDS objects) NIR Spec 160x60 pixels at <<10Hz NDR 160x60 pixels < 10Hz NDR Table 4: Window sizes for different channels of FGS
The FGS1 and FGS2 the window regions are required to be read out at 10Hz CDS frame rate whilst the visible photometric and the NIR spectrometer channels are required to be read out non-destructively using sample up the ramp scheme (SUR/NDR) at 10Hz window frame rate. The H1RG detectors are very versatile detectors which offer complete flexibility to implement different readout modes applicable to both window and full frame readout modes. The two window regions on an array can be readout using a window output or using multiple parallel outputs in standard readout mode, by skipping the unwanted rows and read out the rows corresponding to the window regions normally. The H1RG array can be reset using a global reset or a reset method based on a line-by-line or pixel-by-pixel basis. A standard CDS frame is obtained by a reset followed by the two reads keeping the same clocking for reset and read phases. However, at a minimum detector integration time (i.e. no delay between the reads), the signal integration efficiency (i.e. the signal integration time compared to the CDS frame time) is only 33.3%. For efficient CDS readout, the H1RG allows read, reset, read mode on a line-by-line or pixel-by-pixel basis. That is a currently selected row can be first read out (the second read of the current CDS exposure) then reset and read out again (the first read of the next CDS exposure) before moving onto the next line. The overhead in this method is only a time required to read a single line plus time to do a fast line reset. This mode offers a high signal integration efficiency, close to 100%.
Figure 18: CDS readout using line reset in Read-Reset-Read mode The FGS1 plus visible photometric windows (or FGS2 plus NIR-Spectrometer windows) can be readout at the required window rates using read-reset-read mode at a reasonably low pixel rates with low read noise. For more details on different readout modes and readout rates, please see the technical note on H1RG readout modes [RD37].
Page 35 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Minimum readout rate in Auto- Minimum readout rate in Channel Calibration Mode (multiple Measurement Mode (using parallel outputs) window output) FGS1 (CDS @10Hz) + VIS- 288 kHz 27 kHz Phot (NDR @10Hz) FGS2 (CDS @ 10Hz) + NIR- 268 kHz 114 kHz Spec (NDR @10Hz) Table 5: Minimum readout rates for FGS in different operating modes
4.5.2 AIRS Detectors and readout rates The baseline detectors for AIRS are two independently optimised H1RG arrays for the two channels (CH0 and CH1) and the requirement is to read out the window regions each up to a maximum of 10Hz rate. The window readout approach for AIRS is similar to that of the FGS. The following table shows the window sizes and the minimum pixel rate for a 10Hz window rate. The baseline operational readout mode is to read all AIRS pixels (within the window) out at 60 kpix/s. Further details can be found in chapter 12. Pixel rate for 10Hz (NDR) frame Pixel rate for 10Hz (NDR) Window size Channel rate (multiple parallel outputs) frame rate (using window (Cols x Rows) output) Channel-0 270x64 pixels 173 k pixels/s 41 k pixels/s Channel-1 100x64 pixels 64 k pixels/s 41 k pixels/s Table 6: Window sizes for different channels of AIRS
4.5.3 SIDECAR ASIC as controller electronics The baseline array controller electronics for both FGS and AIRS are the SIDECAR ASICs from Teledyne. The ASIC qualified for space use is available in EUCLID package and hence this can use all the available outputs from the detector. The required pixel rates for the FGS and AIRS are within the ASIC slow speed operation (up to 500 kHz pixel rate in each channel for a 16 bit resolution) limits. The ASIC can be configured to take in multiple parallel outputs of the FGS detectors during auto-calibration mode and then switch to take in window output from the array during measurement mode to save power. Similarly AIRS can also configure the ASIC either to use parallel outputs from the array or to use the window output mode.
Page 36 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
5 SYSTEM OPTICAL DESIGN
5.1 OPTICAL SYSTEM DESIGN OVERVIEW Figure 19 shows a functional block diagram of the payload optics. The optical system can be broken down into a number of separate modules, namely the telescope, the common optics, AIRS and the FGS. The interfaces between these modules are defined in [RD31].
INSTRUMENT OPTICAL BENCH
FGS / NIR-Phot / LRS FGS-DET-1 FGS DETECTOR 1 TELESCOPE ENTRANCE TEL-M1 BAFFLE FGS-1 VIS-PHOT TEL-M2 TEL-M3 COM-M5 TELESCOPE FGS-L2 MIRROR 0.5 – 0.55 um PRIMARY M1 TELESCOPE TEL-M4 M5 INPUT APERTURE SECONDARY M2 TERTIARY RECOLLIMATED BEAM FGS-1 LENS FGS-L1 1100 X 730 mm M3 FOLD MIRROR 20 x 13.3 mm 0.8 – 1 um NIR-PHOT LENS M4 FGS-LP2 T RECOLLIMATING M2 REFOCUS FGS-LP1 MECH. MIRROR T LONG PASS FILTER > 0.55 – 1 um TELESCOPE ENTRANCE LONG PASS FILTER BAFFLE FGS-D4 FGS-M3 < 0.5 – NIR-PHOT / FGS-2 R 0.55 um CASSEGRAIN FOCUS DICHROIC FOLD MIRROR R TELESCOPE OPTICAL BENCH < 0.5 – 1 um FGS TELESCOPE FGS-D5 R COM-D1 T FGS-PR1 FGS-M1 FGS-M2 FGS-D3 NIR- VIS/NIR 1.05 – TELESCOPE MODULE COMMON FGS SLIT SPEC / NIR- NIR-SPEC < 0.5 – 1.9 um TELESCOPE TELESCOPE DET 1 / 2 1.9um DICHROIC FGS-2 SPECT PRIMARY SECONDARY DICHROIC D1 DICHROIC PRISM R T 1.25 – 1.9 um 1.05 – 1.2 um FGS-DET-2 T FGS DETECTOR 2 IR 1.95 – 7.8 um FGS-M4 FGS-M5 FGS-2 FOLD MIRROR FOLD MIRROR TELESCOPE FGS
FGS BEAM DIVISION AND FOCAL PLANES
FGS / NIR-Phot / NIR-Spec MODULE
AIRS COMMON OPTICS CH0 R R=100 SPECTROMETER COM-D2 COM-M6 CH0-L1 CH0-M1 CH0-L2 IR CH0 COM-M7 (Ch0) CH0-PR1 CH0-DET COMMON 1.95 – INPUT FOLD MIRROR DICHROIC 3.9 um CH0 FORE-OPTICS SLIT COLLIMATING FOLD CAMERA CH0 PRISM DETECTOR D2 LENS MIRROR LENS
AIRS CH0 AIRS CH1 T
IR CH1 3.9 – 7.8 um AIRS MODULE
COM-M7 (Ch1)
CH1 FORE-OPTICS SUB-ASSEMBLY COMMON DETECTOR COMMON OPTICS MODULE
AIRS CH1
R=30 SPECTROMETER CH1-SP1 CH1-L1 CH1-M1 CH1-L2 CH1-PR1 CH1-DET INPUT SHORT SLIT COLLIMATING FOLD CAMERA PRISM PASS DETECTOR LENS MIRROR LENS FILTER
AIRS MODULE
Figure 19: Optics functional block diagram
The telescope is a Cassegrain design (parabolic primary M1 and hyperbolic secondary M2) with a third mirror M3 used to recollimate the beam. A fourth mirror M4 directs the collimated beam onto the optical bench. The telescope optical design is described in more detail in [RD5]. Figure 20 shows the telescope optical layout also including the first common optics mirror, M5. M4 is positioned to put the output collimated beam 275 mm from the optical surface of the primary mirror.
Page 37 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
275 mm
TANGENT PLANE TO PLANE OF CHIEF RAY AT MIRROR SURFACE CENTRE OF FIELD
M1
YOPT
M5
0.1°
TELESCOPE LOS M3
ZOPT M2 CASSEGRAIN FOCUS ORIGIN (VERTEX OF PARENT M4 M1)
Figure 20: Scale drawing of the telescope – view in YOPT-ZOPT plane. The 0.1° offset is exaggerated for clarity. The coordinate system is explained in RD5. The aperture stop, located at the primary mirror, M1, defines the elliptical entrance pupil, of size 1100 mm x 730 mm. The entrance baffle, a cylinder extending the length of the optical bench, limits M1’s view of the sky. In combination with placing the stop at the first optical surface (M1), this provides the first line of defence to block out-of-field light. An additional baffle is positioned over the ‘top’ of M2 (as viewed in Figure 20) to block any direct view of the sky from M2 past the end of the entrance baffle. M2 has a refocus mechanism with three degrees of freedom as a baseline (focus and tip/tilt). The purpose is to correct for one-off movements due to launch loads and cool-down and potentially to make occasional adjustments (for example to compensate for any long term drifts in structural stability). To determine the optical focus position a two-step process is used to close the loop. First, a suitably bright star is observed and the peak amplitude of the PSF is monitored on FGS-1 and FGS-2 as M2 is moved. Although affected by a large WFE relative to their central wavelengths, the PSF in these bands is expected to have a well-defined lobe. Once the optimal focus is found in this way, the telescope attitude can be commanded to place the same star onto the slit of AIRS. We can then move M2 and monitor the peak amplitude of the spectra in all spectroscopic channels. The optimal focus position is obtained when the peak signal is maximum as this condition maximises the SNR of the science detection. As M2 is moved, also the position of the star (and its spectrum) moves on AIRS and FGS. Because FGS closes the AOCS loop, this ensures that the star is kept onto the slit at all times during focussing operation. Scanning the spacecraft such that the source is moved orthogonally to the slit will allow to identify when the source is crossing the slit centre. When that occurs, the position of the same source on the FGS-1 and -2 focal planes is recorded as being coincident with the bore-sight of AIRS. The Cassegrain focus after M2 provides the possibility of inserting a field stop (discussed later in section 5.5) to aid stray-light rejection. After the Cassegrain focus, the beam is recollimated by M3. The baseline design results in a recollimated beam of size 20 mm x 13.3 mm, which is directed onto mirror M4, used to direct the beam onto the optical
Page 38 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 bench. Figure 21 shows the telescope and common optics. Figure 22 shows the common optics viewed in the plane of the optical bench. The telescope is required to be diffraction limited at 3 μm, which equates to an rms WFE of about 200 nm. This is expected to be dominated by M1 surface error. More details of the analysis of telescope performance can be found in RD34 and RD35.
M5 BEAM TO FGS AIRS SLIT Ch1
AIRS SLIT Ch0 M6
D1
D2
M1
EXIT PUPIL
M7 Ch0 M2 M7 Ch1 M3
M4
Figure 21: Scale drawing of the telescope and common optics – view in XOPT-YOPT plane
Page 39 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
M5
FGS INPUT CH1 INPUT 90 mm SLIT
CH0 INPUT SLIT
M6
D1
D2
M7 CH1 M7 CH0
M4
Figure 22: Common optics as viewed in the plane of the optical bench. The beam from M4 travels to another fold mirror M5 which deflects the beam back at a modest angle to allow accommodation of the dichroic beamsplitters and AIRS input optics. The beam from M4 runs parallel to the sides of the optical bench (vertically, parallel to the local y-axis shown in Figure 22) and the elliptical pupil on the optical bench is orientated such that the major axis is parallel to the plane of the bench. The calibration source is situated behind M5, injected into the beam through a small (1 mm diameter) hole in the centre of the mirror. The first dichroic, D1, is used to split light between the FGS and AIRS, with a transition at about 2 um. The dichroics are laid out to minimise angle of incidence, which is always kept below 30°. A dichroic specification has been produced [RD15] and was used to elicit design proposals from vendors. The short wave beam from D1 passes to the FGS, which provides two guiding channels along with a visible photometric band and a low resolution (R>10) NIR spectrometer. Section 13 gives further details of the FGS design. The long-wave beam from D1 is further split by D2 into two wavebands, of one octave each (1.95 μm – 3.9 μm and 3.9 μm – 7.8 μm). Each of these two paths is focussed onto the spectrometer entrance slit by an identical off-axis parabola M7, the function of which is to deliver the correct input f-numbers. The mirror focal length is chosen to deliver a beam with f/12 (major pupil axis) and f/18 (minor pupil axis). An additional fold mirror M6 is used to space the centres of the AIRS input slits apart by 90 mm – the agreed interface with AIRS. The beam geometry is arranged so that both slits have the same x-coordinate (following the local coordinate frame shown in Figure 22). Page 40 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The two spectrometers have independent optical channels. Channel 0 gives R=100 over the shorter waveband (1.95 – 3.9 um) while channel 1 gives R=30 over the longer waveband (3.9 – 7.8 um). The spectrometer uses a prism design; the reasons for selection of the design are explained in [RD2]. Section 12 gives further details of the AIRS design.
5.2 MODULE OPTICAL DIVISION The optical design is divided into a number of modules, as indicated in Figure 19. These are the telescope, the common optics, the FGS and AIRS. This division enables clear interfaces to be defined between each module (explained in more detail in RD31. The telescope comprises all of the optics from the entrance aperture of the telescope up to, and including, the M4 mirror. The common optics comprise the M4 and M5 mirrors, the dichroics (D1, D2) that split the beams for the FGS and AIRS Ch0/Ch1, and the fore-optics (M6, M7) that produce a focussed spot as an input for AIRS. The FGS comprises all of the optics following on from the collimated, reflected beam from D1. AIRS comprises all of the optics following on from the focussed input beam delivered by the fore-optics M7.
5.3 COMMON OPTICS DESIGN The function of the common optics is to: • split light into three channels (FGS; AIRS Ch0 and AIRS Ch1); • deliver the collimated beam to the FGS; • deliver focussed beams, of the correct f/number, at the AIRS input slits.
5.3.1 Dichroics The function of splitting is carried out by the two dichroics, D1 and D2. A specification for these components is given in RD15 and a summary is presented in Table 7 below. Dichroic Purpose R ≥ 0.9 (0.95 goal) R = 0.5 T ≥ 0.9 (0.95 goal) Name λ (μm) T = 0.5 λ (μm) λ (μm) Com-D1 Division of FGS from Spectrometer <0.50 – 1.90 1.95 2.0 – >7.8 Channels
Definition of short wavelength edge of AIRS Ch0
Definition of long wavelength edge of NIR-Spec
Com-D2 Sub-Division of Spectrometer <1.95 – 3.8 3.9 4.0 – >7.8
Definition of short wavelength edge of AIRS Ch1
Definition of long wavelength edge of AIRS Ch0
Table 7: Summary of dichroic properties
The dichroics are specified to have a minimum transmittance or reflectance of 90% (with a goal of 95%) over the wavelengths shown in Table 7. Both dichroics are specified to have a broad enough transition region to ensure at least one spectral element of overlap between AIRS Ch0 and Ch1 and between the FGS NIR-Spec and AIRS Ch0.
Page 41 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The dichroic specification has been refined during the study based on feedback from vendors. A throughput budget has been constructed [RD32] showing the expected in-band throughput from the end-to-end dichroic combination and also the degree of out-of-band blocking. Further details are given in section 5.6 and RD32. The baseline design requires a dichroic used aperture of 30 mm for D1 and D2 and is based on a 10 mm thick substrate. Possible materials for the substrates are ZnS Cleartran (or Cleartran for short), ZnSe and germanium. Table 8 lists the properties of these three potential candidates for the substrate material. Of particular interest is the sensitivity to alignment change with temperature, which will induce an offset in the collimated beam during cool-down and, potentially, a change in pointing if the substrate is not perfectly parallel, or a change in focus if the beam is not perfectly collimated. Table 8 presents an assessment of the alignment change induced with temperature for each substrate material, based on calculating a factor that represents the change in optical path difference with temperature. The data on CTE and dn/dt should be treated with caution as both can vary considerably from room temperature to cryogenic temperatures, but this at least gives a first assessment of potential substrates. It can be seen that, even allowing for these uncertainties, germanium is an order of magnitude more sensitive to alignment change with temperature. It also has an absorption band close to the 1.95 um cut-on wavelength of D1. For these reasons it is judged unsuitable. Both ZnSe and Cleartran are potential substrates. ZnSe is softer, making it easier to polish but more susceptible to marking (at least in its uncoated form – in reality durability will be limited by the coating). Given that their refractive indices are quite similar, both materials can act as a substrate for similar coating designs, meaning that the choice comes down to polishing and handling considerations. The main driver is achieving a good quality surface polish to limit scattered stray light, along with low bulk scattering within the material. On this basis ZnSe was chosen as the substrate material as it is likely to yield a better polished finish.
Material Refractive index Transmission CTE (K-1) dn/dt (K-1) Sensitivity of (approximate range (um) alignment to average over temperature waveband) change n.CTE + dn/dt (K-1) ZnSe 2.45 0.6 - >10 7.1E-6 @ 5.5E-5 @ 7.2E-5 273 K 153 K ZnS Cleartran 2.3 0.5 - 10 6.3E-6 @273 4E-5 @ 5.4E-5 K 153 K Germanium 4 1.9 - >10 6.1E-6 @ 396E-6 4.2E-4 298 K Table 8: Summary of material properties. ZnS dn/dt data taken from J. Phys Chem Ref Data, Vol 13, No 1, 184. All other data from manufacturers’ / vendors’ datasheets.
Details of the dichroic designs obtained from vendors are given in [RD36] and used in calculation of the overall performance as presented in [RD31]. Figure 23 shows the predicted performance of D1 and D2. Further dichroics (not part of the common optics) are used inside the FGS and AIRS in order to define band edges. Details are given in [RD15], [RD36] and [RD32].
Page 42 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
D1 and D2 dichroics 1
0.9
0.8
0.7
0.6
COM-D1 R 0.5 COM-D1 T
Throughput COM-D2 R 0.4 COM-D2 T
0.3
0.2
0.1
0 0 1 2 3 4 5 6 7 8 9 10 Wavelength (μm)
Figure 23: Predicted performance for D1 and D2 (credit: University of Reading)
5.3.2 Spectrometer Input Optics The spectrometer input optics are based on the use of an off-axis paraboloid to bring the collimated beam to a focus at the spectrometer slit. Two identical mirrors M7Ch0 and M7Ch1 are used to deliver a beam with a focal ratio of f/12 and f/18. With the elliptical collimated beam of 20 mm x 13.3 mm this requires a focal length of 240 mm. The nominal image quality of the telescope and common optics combination is limited by the aberrations of the off-axis parabola, which are determined by the combination of mirror focal length and off-axis distance. After some iteration of the layout an off-axis distance of 55 mm was chosen. Figure 24 and Figure 25 show the rms WFE at the focus of the input optics over a 30 arcsec FOV. The worst case is about 0.026 μm, or λ/115 at 3 μm. Thus the nominal design aberration of the telescope and common optics is insignificant in the context of a requirement to have a diffraction limited telescope at 3 μm.
Page 43 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 24: rms wavefront error over a 30 arcsec square field of view at the AIRS Ch0 input at 3 um
Page 44 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 25: rms wavefront error over a 30 arcsec square field of view at the AIRS Ch1 input at 3 um
5.4 ON-BOARD CALIBRATION UNIT An on-board calibration unit (OBCU) is placed in the common optics. When the calibrator is operated, it provides a uniform illumination sources on all focal planes. This will be used during commissioning to monitor variations in QE, as an alternative to diffuse astronomical sources, and to transfer over time the calibration obtained on the ground. During science operation, the focal plane flat field calibration can be monitored in a similar way if necessary. The OBCU makes use of a thermal source and 4 light emitting diodes (LED) feeding a 40mm diameter integrating sphere with its output port behind M5. Light is injected through a 1 mm diameter hole in the M5 mirror, which in the 20mm x 13.3 mm collimated beam results in negligible loss of throughput, i.e. less than 0.4%. A thermal source emitting a graybody spectrum is used for the AIRS channels. JWST-MIRI and NIRSpec use tungsten filament sources for the same purposes in a temperature range from 500 K (on JWST MIRI8) to 1600 K (JWST NIRSpec9). Even when operated at temperatures as high as 1600 K, the Planck spectrum of the thermal source does not provide a signal sufficiently bright at VNIR wavelengths, however LEDs can be used at additional input ports of the same integrating sphere. The absence of an input slit in the ARIEL NIRSpec allows using LEDs, which are narrow band, to illuminate the whole focal plane. An alternative solution involves replacing the thermal source with a series of LEDs extending to the red-end of AIRS channels. Visible to mid-IR LEDs are routinely used in a similar way at cryogenic temperatures and constitute
8 Wright, G., et al., PASP 127, 595, 2015 9 Bagnasco, G., et al., SPIE 6692, 2007
Page 45 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 an interesting option, although one with a low TLR in terms of space qualification of the sources. The main disadvantages are increased power dissipation and complexity in the mechanical interface with the integrating sphere and electrical interface. Therefore, the first solution is adopted in the baseline design because of its high TRL, along with the relatively simple interfaces.
5.4.1 OBCU design Light emerges from a 1 mm diameter hole in M5, at the centre of the collimated beam in the common optics. The surface brightness is defined by the input optical power, the diameter of the integrating sphere, and by the surface area of the input and output ports. The total power received by each of photometer and spectrometer is = ( ) where the throughput is the product of the output port surface area, As, and the solid angle subtended𝑖𝑖𝑖𝑖 𝑠𝑠 𝑠𝑠 by𝑠𝑠 the input optics of each instrument (M7 CH0 and M7 CH1 in AIRS, and the primary of the Gregorian𝑄𝑄 𝐴𝐴 Ω telescope𝐼𝐼 𝜆𝜆 in VNIR) and their pupils. AIRS have input slits, and the focal plane measures the spectrum of the calibrator. A monochromatic light source illuminates a region of the focal plane of area, AFP. VNIR does not have a slit, and the calibration signal is uniformly diffused on the focal plane of both NIRSpec and photometers. The power density at the focal plane, IFP, is estimated using Zemax modelling. The relevant parameters are listed in the table below, which also shows the efficiency, εs, representing the fraction of the total input power reaching the focal planes.
VisPhot FGS1 FGS2 NIRSpec AIRS AIRS CH0 CH1
AFP - - - - 0.25x1.94 0.2 x 1 (mm)
-4 -4 -4 -4 IFP./QIN 1.7x10 2.5x10 2.5x10 2.5x10 0.08 0.04 (mm-2)
εs 0.5% 0.5% 0.5% 0.5% 4% 0.8%
-3 -4 Ωs (sr) 2.6 x 10 7.4x10
Figure 26: Signal per detector pixel from OBCU. These parameters have been used to estimate the signal expected from the OBCU in each detector pixel. Assumed are: 1) Gold coated integrating sphere 2) MIRI-type tungsten source (emissivity 0.2, efficiency from self-shielding 0.6, temperature 1100 K) with a 10% neutral density filter. 3) Four off the shelf (to be space qualified in phase B) cryogenic LEDs operating at central wavelengths 0.525 µm, 0.910 µm, 1.07 µm, and 1.55µm.
The radiated optical power by each source at the input port of the sphere are 6 mW in AIRS (combined CH0 and CH1 bands), 2.6 mW in VISPhot, and 2 mW in each FGS-1, FGS-2, and NIRSpec bands. With these assumptions made, the expected signal at the focal planes is shown in Figure 26. This configuration allows to obtain a measurement with a SNR larger than 30 in one second of integration. Useful considerations can be made: this solution allows to monitor the flat field calibration to 1% in less than 10 s in all spectroscopic channels. Repeated calibration flashes can be done to increase the SNR. As the source is flashed while not observing science targets, this does not translate in a reduction of observing efficiency. The max dynamic range experienced is on the CH1 detector and it is about 1, similar to a star. Operating the thermal source at a higher, but reasonable temperature (1600 K) does not allow to avoid the use of LEDs in the VNIR channels, but would result in a higher power dissipated and increased risk related to the aging of the filament.
Page 46 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
5.4.2 Thermal source The proposed thermal source is a tungsten filament which borrows from the infrared calibration source heritage of the MIR Instrument on board JWST. The source itself is a wound tungsten coil, spot welded with copper- clad nickel-iron core alloy. The geometry and shape of the filament is shown in Figure 27 below and shows the presence of two filaments in the same cavity (one for redundancy).
Figure 27: The two filaments are shown at the centre of the assembly in both pictures (top forward and Front view). The two glass beads which achieve mechanical bonding of the filaments are also visible. A forward semi-spherical cavity can be used as well as a PC (Parabolic Concentrator) which terminates in a waveguide feeding section (output). A microscope picture of the filaments is shown at the bottom.
5.4.3 Electrical Interface The current drive through the filament is used in a 4-wire configuration to allow current and voltage monitoring. The filament can be driven with currents in the range 0-15mA in order to produce a maximum temperature of approx 1100K. We estimate that in order to illuminate the AIRS spectrometer with a calibration flash providing more than 30 s-1/2 SNR, the source needs to reach a temperature of ~ 1100K. Assuming a stable current driver with 16 effective bit DAC a current resolution of 1.192 nA is achievable which corresponds to a temperature resolution of T ~ 20 mK which in turn suggests control of the output power to better than one part in 104. With a higher resolution DAC it would be also possible to monitor stability of the detector acquisition system if needed.
5.4.4 Mass and Power Estimation The calibration unit only has an estimated mass ~10g. The additional mass of a small calibration sphere and potentially a solid cylindrical thermal link to the optical bench will still be kept under 100g. The total electrical power dissipated by the thermal source is estimated to be 30 mW of which about 6 mW are radiated. The 4 LEDs dissipate 300 mW of which 8.5 mW are radiated. The short operation of the sources while not observing science targets will not incur in a reduction of performance or observing time.
Page 47 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
5.5 SYSTEM LEVEL STRAYLIGHT CONTROL AND BAFFLING A full end-to-end model of the system stray light is likely to be very challenging and outside the scope of this assessment phase, given the need to faithfully model stray light paths through both the telescope and AIRS and the FGS. Indeed, there is a risk that construction of such a model could provide either a false sense of security or an overly pessimistic prediction unless extreme care is taken over issues such as correct representation of scattering properties and sufficient sampling of rays. Our approach is to take a modular approach to the design for stray light, concentrating on making the individual systems (the telescope, FGS and AIRS) as insensitive to stray light as possible by good design practice (such as placement of suitable stops and baffles, choice of appropriate designs and use of appropriate materials). A quantitative stray light analysis has not been made but thought has been given to including features that will reduce the sensitivity of the optical design to stray light, and these are described below. The first step in stray light reduction is to make the telescope as insensitive to stray light as possible. The first line of defence is the cylindrical baffle at M1, which limits M1’s view of the sky. The system aperture stop is located at M1, and this provides a well-defined pupil. Given that M1 is both a large surface and the first optical surface, it is likely that the stray light response of the instrument will be dominated by scatter from its surface. Thus, achieving sufficiently low surface roughness of the polished surface will be important. The PTM is specified as have an rms surface roughness of 10 nm or less [RD33]. There is a Cassegrain focus after M2 and this allows a field stop to be included in the system. This stop will be slightly oversized compared to the 50” telescope FoV (section 11.4.1.4) and will serve to attenuate both scattered light from M1 and M2 and to block any out-of-field sources close to the edge of the FoV. The field stop also serves to block any unwanted views that downstream mirrors may have of either M1 or M2. There is an accessible image of the aperture stop in the collimated space after M4, shown as the exit pupil in Figure 21, [RD5] and this allows the placement of Lyot stop at this location. In combination with the field stop at Cassegrain focus, this is good practice in a reflective telescope as it further suppresses any stray light paths. Inclusion of a Lyot stop also has the advantage of providing a ‘clean’ entrance pupil (unaffected by any pupil aberrations) for AIRS and the FGS. At the dichroics, D1 and D2, a substrate material has been chosen that enables a high quality polish to be obtained along with low bulk scattering (section 5.5) to limit generation of stray light by these optics. The spectrometer input optics will deliver a focussed spot at the AIRS input split. In theory a slit is not required at the spectrometer input (it is a ‘slitless’ spectrometer with resolving power defined by the PSF size rather than the slit width). In practice a slit will be included, forming a further field stop to limit stray light particularly from field stars. At the start of this study, the baseline spectrometer design used a grating. One of the major efficiency losses in such a design is scattered stray light from the grating, the magnitude and direction of which can be vary depending on manufacturing tolerances, making it difficult to predict. This has now changed to a prism, which will have much lower, more predictable levels of scattered stray light. The FGS telescope has also changed since the beginning of the study and now comprises and off-axis Gregorian telescope (the previous design was on-axis). This allows placement of an additional field stop at the intermediate focus, further limiting stray light.
5.6 OPTICAL BUDGETS AND PREDICTED PERFORMANCE
5.6.1 Overall Throughput Budget At system level, a throughput budget is maintained to calculate the overall instrument efficiency. This is detailed in RD32. The throughput budget accounts for efficiency losses due to: • surface reflections; • internal transmission;
Page 48 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• ageing; • contamination; • filter efficiencies.
The following contributions are not accounted for: • detector QE; • slit losses • spreading of the PSF due to a non-diffraction limited WFE.
These effects are accounted for separately in the overall performance analysis [RD11].
The estimated end-to-end throughput is shown in Figure 28 and Figure 29
End-to-end throughput for AIRS Ch0 and Ch1 1
0.9
0.8
0.7
0.6
0.5 AIRS total - Ch0 AIRS total - Ch1 Throughput
0.4 Compliance
0.3
0.2
0.1
0 1 2 3 4 5 6 7 8 9 10 Wavelength (um)
Figure 28: End-to-end throughput for AIRS Ch0 and Ch1. The requirement from the MRD is also shown.
Page 49 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
End-to-end throughput for the FGS 1
0.9
0.8
0.7
0.6
VIS-PHOT total 0.5 FGS-1 total FGS-2 total Throughput
0.4 NIR-SPEC total Compliance
0.3
0.2
0.1
0 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 Wavelength (um)
Figure 29: End-to-end throughput for the various FGS channels. The requirement from the MRD is also shown.
5.6.2 Wavefront Error Budget The WFE error budget for the instrument is given in RD31. A summary is given in Table 9 below. This results in a PSF at the AIRS detectors which is diffraction limited at about 4 μm. Table 9: rms WFE budget rms WFE Contributor (nm) Notes WFE assumptions Telescope Manufacturing errors Set to give ~ 200 nm total M1 reflected WFE 160 Primary mirror (OAP) for telescope M2 reflected WFE 63 Secondary mirror (hyp) Lambda by 10 at 633 nm M3 reflected WFE 63 Tertiary mirror (OAP) Lambda by 10 at 633 nm M4 reflected WFE 32 Plane mirror Lambda by 20 at 633 nm Nominal design aberrations WFE in output collimated beam Telescope 20 from M4 From Zemax design Assembly and alignment Target value for flight mirror - Gravity deformation of M1 50 PTM factor ~2 larger Misalignment in flight 50 Residual after correction by M2M
Telescope total WFE 200 Page 50 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
rms WFE Contributor (nm) Notes WFE assumptions
Common optics Manufacturing errors M5 reflected WFE 32 Plane mirror Lambda by 20 at 633 nm M6 reflected WFE 32 Plane mirror Lambda by 20 at 633 nm M7 reflected WFE 63 AIRS fore-optics (OAP) Lambda by 10 at 633 nm D1 reflected / transmitted WFE 63 Plane dichroic Lambda by 10 at 633 nm D2 reflected / transmitted WFE 63 Plane dichroic Lambda by 10 at 633 nm Nominal design aberrations To give total telescope + fore- AIRS fore-optics 77 optics = 80 nm From Zemax design Assembly and alignment AIRS fore-optics 63 Residual after alignment Lambda by 10 at 633 nm
Common optics total WFE 155
AIRS Manufacturing errors AIRS collimator 63 Doublet lens Lambda by 10 at 633 nm AIRS prism 63 Prism Lambda by 10 at 633 nm AIRS camera 63 Doublet lens Lambda by 10 at 633 nm Nominal design aberrations AIRS 70 From Zemax design From Zemax design Assembly and alignment AIRS 63 Residual after alignment Lambda by 10 at 633 nm
AIRS total WFE 145
nm (telescope + common optics + Total rms WFE 292 AIRS)
Equivalent diffraction limited wavelength 4.1 um
5.6.3 Pupil Alignment Budget Pupil misalignment will translate to a loss or variation in throughput, depending on whether it is a static or time-varying (dynamic) error. If we express the pupil shear as Δ/R, where Δ is the movement of the pupil from nominal and R is the pupil radius, experience from other cryogenic instruments of a similar scale (for example JWST MIRI) suggests that an overall Δ/R of around 3% to 4% should be feasible for the static part of the budget. Over the
Page 51 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 collimated beam (diameter on the order of 20 mm) this translates into a total misalignment of about 0.3 to 0.4 mm. Table 10 presents a budget for static pupil alignment. Dynamic alignment values are expected to be substantially smaller. The values in the table assume stop size is precisely matched to beam size. Sensitivity to pupil misalignment can be reduced further by slight oversizing of field and Lyot stops (at the expense of a small increase in stray light).
Throughput (100% = perfect Degree of freedom % pupil shear allocation Pupil misalignment (mm) alignment) Alignment of M1 aperture stop with telescope Lyot stop (after M4) 2.0% 0.21 98.7% Alignment of common dichroics, D1, D2 1.5% 0.16 99.0% Alignment of AIRS entrance pupil with Lyot stop 1.5% 0.16 99.0% Alignment of FGS entrance pupil with Lyot stop 1.5% 0.16 99.0% Total static pupil misalignment 3.3% 0.35 97.9% Table 10: Budget for static pupil misalignment. Errors are assumed uncorrelated and are added in quadrature.
Page 52 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
6 SYSTEM MECHANICAL DESIGN
6.1 PAYLOAD MODULE STRUCTURE The main mechanical units composing the PLM are: − the V-Grooves − the bipods and supporting struts − the Telescope Assembly (Telescope optical bench, mirrors, struts and baffle) − the Instrument Enclosure (modules, common optics, radiator)
6.1.1 V-Grooves The V-Grooves (VGs) are high efficiency, passive radiant coolers, providing the first stage of the PLM cooling system. The Planck mission has definitely demonstrated their efficiency as passive cooling systems. Parasitic heat from warmer sections of the S/C is intercepted by the VGs and radiated to space after multiple reflections between the adjacent shields. To achieve this, VGs surfaces must have a very low emittance coating, a high reflection/mirroring material needed to reflect heat radiation. Only the upper surface of the last VG (VG3), exposed to the sky, is black coated with a high emissivity material to maximize the radiative coupling, and so heat rejection to deep space. VGs consist of a set of three specular shields, composed of six half circles arranged in a “V-shaped” configuration, angled along the diameter parallel to the S/C X axis (Figure 30). A constant angle of 7° has been assumed as the inclination between V-Grooves, resulting in a set of 7°-14°-21° for the three shields, separated by a gap of 100 mm at the vertices. V-Grooves are mechanically designed as a simple sandwich of Aluminum alloy (series 1000 or 6000) layers. A honeycomb cell structure 10 mm thick, with 10 mm (or less) cell size and ribbon thickness of 1 mm, is packed between two 1 mm thick layers. This thermo-mechanical configuration has several advantages: − achieve great thermal shielding in a relatively compact volume and reduced mass; − simple construction, totally passive, high reliability; − no need for deployment, vibration free. The average mass estimation for the V-Grooves arrangement is in the 3-5 kg/m2 range, depending on the cell size and thickness selected. Details of the surface finishes for thermal control are in section 7.
Figure 30: ARIEL PLM mechanical design
Page 53 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
6.1.2 Bipods and supporting struts The PLM is supported by three bipods mounted onto the PLM/SVM interface plate (shown in figure above). The interface is shown in the PLM MICD, [RD22]. One bipod is at the front of the telescope baffle in central position. The other two are on the rear side of the Telescope Assembly, supporting the OB and the Baffle on two points. The VG’s are thermally attached to the three bipods and are mechanically supported by eight GFRP struts. Bipods preliminary thermo-mechanical configuration is based on the Planck design. They are assumed as hollow cylinders, made of GFRP (R-Glass + Epoxy), a low conductive material with good structural properties. To increase their mechanical stiffness the inner volume of the cylinders is filled with low thermally conductive rigid foam. The eight supporting struts for the V-Grooves are positioned in order to optimally support the radiators. They are designed as hollow GFRP cylinders extending from the SVM/PLM interface to the lower surface of the last VG.
6.2 MECHANICAL INTERFACES DESCRIPTIONS
6.2.1 Main PLM Interface The main PLM interface is designed using thermal isolating bipod struts. These struts modelled from tubular glass fibre reinforced plastic. The flexures at the end of each strut are modelled as titanium components. The detailed geometry of the flexures has been investigated as part of the telescope structural analysis in order to ensure that they are both flexible to maintain the kinematic mounting and strong enough to support the PLM for launch and thermal contraction of PLM/SVM.
6.2.2 M1 interface to Telescope Structure M1 is to be supported from the telescope structure via 3 whiffle tree mountings, these are manufactured from aluminium and may need to provide a small amount of thermal isolation between the TOB and M1. These mounts need to be optimised during the next phase of the project as part of the telescope structure light-weighting program.
6.2.3 Instrument Boxes to Telescope Structure M3 will be mounted directly from the telescope optical bench. M4 is positioned under the instrument enclosure to fold the light beam 40mm above the instrument mounting face of the TOB. The first item in the instrument suite is the calibration source. This is passively cooled by the telescope optical bench. The light injection of the calibration source is planned to be at the centre of M5. The source consists of a light source and small integrating sphere. Next in the chain is the FGS system. This is also cooled directly from the bench. The input for the FGS is split from the main beam using Dichroic 1. Next in the detector chain is the two channel AIRS system. The detectors for AIRS will require specific cooling, an active cooler is to be mounted inside the instrument enclosure. The AIRS detector will be thermally isolated from the telescope structure and the support optics and electronics will be mounted directly to the telescope structure. On the top of the instrument enclosure the instrument radiator is mounted to complete a light tight housing. This is used to provide a passive cold environment for the instruments and telescope. This radiator is currently 0.6m x 0.7m which is estimated by the current thermal model to be very conservative. The large instrument enclosure allows for flexibility in the optical design and space to carry out on-ground calibration and alignment.
6.3 TELESCOPE OPTICAL BENCH DESIGN The Telescope Optical Bench (Figure 306) is a key mechanical component as it forms the main structure of the PLM: the telescope M1 is mounted to its front surface and telescope beam projects from its bottom edge forward to support the M2 near the front bi-pod. On the other side the Instrument enclosure has been incorporated into the bench design, increasing the bench thickness and thus stiffening the structure. The design of this platform, and the units mounted on it, will be as isothermal as possible except for the Page 54 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 components that need thermal decoupling. The bench, the modules and the common optics supports are to be manufactured predominantly with aluminium in order to minimize possible defocus effects that could occur during thermo-elastic contractions. At present the bench is assumed to be built in Al6061 to ensure a uniform thermo-mechanical behaviour of the whole Telescope Assembly. This is a key issue for the design of the whole assembly and has be addressed as part of the thermo-mechanical analysis of the telescope sub-system. The Optical Bench is directly supported on the two rear bipods by means of titanium flexures. The final shape and dimensions of the flexures have been defined via the telescope sub-system study and have been confirmed to be adequate by the system level structural model. The detailed design/stressing and light- weighting of the TOB will continue into phase B. The current design assumes that the component is manufactured using conventional techniques, however further investigation can be made into structural optimisation of this component.
6.4 PAYLOAD RADIATOR MECHANICAL DESIGN The upper half of the Telescope Baffle and the Instrument Box top side face directly the cold sky during nominal operations and work as efficient radiator surfaces for heat rejection, improving the PLM passive cooling performance. For this reason, their surface shall be optimized in terms of IR emissivity. At present the Baffle is designed as a simple sandwich of Al6063 (TBC) alloy layers. This thermo-mechanical design reduces mass while ensuring at the same time a very good thermal conductance in all three directions and mechanical stiffness. The top half section outer layer, directly exposed to space, is covered by a black (or white) painted honeycomb cell layer (see Chapter 7.3.2) with the same characteristics of the internal structure. The Baffle is mechanically supported primarily from the telescope stiffening tubes which run from the TOB and interface with the telescope beam. The top face of the Instrument enclosure is used as a dedicated radiator for detectors. In this preliminary design configuration, the radiator is mounted directly on the enclosure side walls which are now part of the TOB. The radiator closes out the volume of the enclosure with a top face area around 0.37 m2. This area has been shown to provide sufficient cooling by the PLM thermal analysis. The radiator is assumed to be a 12 mm thick black (or white) painted honeycomb cell structure of Al6063 alloy, 20 mm (or less) cell size and ribbon thickness of 1 mm mounted on an Aluminum layer at least 2 mm thick.
6.5 PAYLOAD MODULE MASS BUDGET The PLM mass budget is shown in Table 11 below. Note that further mass optimisation is to be carried out on the telescope structure, further optimisation and investigation into more aggressive light weighting will be carried out in phase B. Mass maturity margin of 20% is applied across the board. The instruments however carry further mass margin in case of further design developments. CBE Mass Nominal Mass (with Contingency Item (kg) contingency) (%) Telescope Subsystem 263.3 300.0 TOB* 106.6 111.9 5 Telescope Beam 29.2 35.0 20 Telescope Baffle 12.0 14.4 20 Telescope Heaters, Harness 6.0 7.2 20 M1 85.0 102.0 20 M1 Supports 16.9 20.2 20 M2 1.5 1.8 20 M2 Refocus Mecahanism 5.0 6.0 20 M3 0.2 0.2 20 M3 Mounting 1.0 1.2 20 Instrument Subsystems FGS 1.6 2.0 25 AIRS 7.7 9.3 25
Page 55 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
CBE Mass Nominal Mass (with Contingency Item (kg) contingency) (%) Common Optics + Calibration 2.0 2.4 20 source Instrument Radiator 6.0 7.2 20 Instrument Enclosure 0.0 0.0 20 JT Cooler Cold Head 1.5 1.8 20 Payload Cryo Harnesses 6.5 7.8 20 Thermal Shield Assembly
PLM Support Struts 16.0 19.2 20 V-grooves 53.3 64.0 20 PLM Total: 357.9 413.6 15.6
Service Module Components Instrument Control Unit 10.5 12.7 20 FGS Electronics 5.5 6.6 20 Cooler Compressors and Plumbing 8.0 9.6 20 Cooler Drive Electronics 8.0 9.6 20 SVM Total: 32.0 38.5 20
Grand Payload Total: 389.9 452.1 * TOB carries reduced margin as current design considered highly conservative in terms of mass The instrument enclosure and M3 mounting have all or the majority of their mass now included in the TOB as they have been incorporated into a monolithic design. Table 11: ARIEL Payload Mass Budget Estimates and Margins
6.6 PRELIMINARY PAYLOAD MODULE MECHANICAL MODELLING RESULTS The payload module was modelled within Patran/Nastran using bar and shell elements. This analysis was based upon worst-case masses carrying full margin in order to determine the normal modes of the full PLM. Full details of the structural analysis can be found in ARIEL-MSSL-PL-ANA-001 [RD3]. The model used is provided as ARIEL-MSSL-PL-ML-001 [RD4].
Figure 31: Finite Element Model
Page 56 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The sizing of the primary support struts are from the detailed telescope design. These have been increased slightly in stiffness and the flexures have been revised from aluminium to titanium. The thermal isolating properties have remained fundamentally unchanged. All of the sub-systems have been subject to detailed design studies so all masses have been revised and used in the revised analysis. 20% maturity margin is still carried to ensure a conservative design. The first lateral mode is predicted at 38.5Hz which higher than the requirement of 30Hz +15%. The first longitudinal mode is associated with the V-groove, however this is a very minor mode with very low effective mass. It is associated with the bottom V-groove where the largest overhang occurs and local stiffening of the plane could be easily introduced using very little mass to move this mode up in frequency. The first significant longitudinal mode is seen at 63.8Hz which is just higher than the required 50Hz + 15%. However as the model carries all mass margin as well this is seen a safe longitudinal mode at this stage of the project. Mode Mode Lateral/Longitudinal Associated sub-system Number Frequency 1 38.5 Lateral (Y axis) M1/TOB first mode 2 39.6 Lateral (X axis) M1/TOB second mode 3 41.9 Lateral (Y axis) V-Groove 4 43.4 Longitudinal V-Groove 5 50.6 Lateral V-Groove 6 52.1 Lateral V-Groove 7 56.6 Lateral V-Groove 8 60.5 Lateral V-Groove 9 62.1 Lateral V-Groove 10 63.83 Longitudinal Telescope third mode Z axis Table 12: Main PLM Modes
Figure 32: Mode 1 shape Mode 1 is a lateral mode of the telescope and is associated with the PLM moving in the plane of M1 and the TOB in the Y direction. This is where the majority of the mass of the PLM is concentrated and in this direction the rear bi-pods are relatively flexible. This mode is not seen as a problem. The sizing case for the PLM struts is the thermal cool down case where there is differential thermal contraction between the PLM and SVM.
Page 57 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Mode 10 is the first significant longitudinal mode and is associated with the bending of the telescope beam. This mode is higher than the required minimum for longitudinal modes and can be readily tuned using the bending stiffness of the telescope beam. All of these modes have been achieved with CBE based upon current design with 20% mass maturity margin added.
Figure 33: Mode 10 shape
Page 58 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 34: Modal effective mass fraction from fixed modes run
Page 59 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
7 SYSTEM THERMAL DESIGN
7.1 BASELINE THERMAL ARCHITECTURE AND DESIGN The spacecraft thermal design (Figure 35) is based on a cold Payload Module (PLM) sitting on the top of a warm Service Module (SVM). The mission thermal control is accomplished by a combination of passive and active cooling systems. The SVM is thermally controlled in the 270K-290K range for nominal operations of all the S/C subsystem units. The function of the cold PLM is to shield the scientific instrumentation (the ARIEL instrument and the telescope system) from the warm section of the S/C and to provide it with the required cooling and thermal stability. Passive cooling is achieved by a high efficiency thermal shielding system (Figure 35 and Figure 36) based on a multiple radiators configuration that, in the L2 environment, can provide stable temperature stages down to the 50 – 60 K range. At 1.5 million km from the Earth in the anti-Sun direction, the L2 orbit allows to maintain the same spacecraft attitude relative to the Sun-Earth system, while scanning the whole sky during the mission duration. Limiting the allowed Solar Aspect Angle (SAA) range, ARIEL will operate in a very stable thermal environment keeping always protect from the Sun/Earth/Moon illumination the coldest section of the PLM. For this reason the SAA allowed during nominal observations will be limited to ±5° around the S/C X-axis and to ±25° around the Y-axis. An extra margin due to possible contingencies, respectively of 1° and 5°, has been assumed on these values: the thermo-mechanical architecture of the PLM has been designed within a total envelope of ±6° around the spacecraft X-axis and ±30° around the Y-axis (Figure 36 and Figure 37).
Figure 35: PLM Thermal Architecture Scheme The SVM upper surface, the main thermo-mechanical interface of the PLM to the S/C, is covered with a low emissivity MLI shroud and acts as the first main radiative barrier between the PLM and the warm units in the service module. The geometrical configuration of the PLM passive stages and the maximum Solar Aspect
Page 60 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Angles allowed during the mission are strongly related. The SVM interface is assumed, at this stage of the study, as a perfect Sun shield for the PLM in the thermal analysis. This assumption will be verified in the next thermal analysis and design steps as the first Sun shielding stage and the PLM passive cooling system must be mutually designed. The passive cooling system is based on a three V-Grooves combination (Figure 35 and Figure 36). They represent the first cooling stage of the PLM. Mechanically supported on the SVM by bipods and other insulating struts, their shape, geometrical configuration and optical properties allow an efficient rejection of heat to cold space (Figure 37). Past missions (such as Planck, Spitzer Telescope and Herschel) have demonstrated that, in an environment such as L2, is possible to passively reach and maintain temperatures down to the 50K range with loads up to more than 1W. These three radiators, called V-Groove 1, 2 and 3 (VG1, VG2, VG3), work in sequence at temperatures around 190K, 120K and 60K respectively, providing stable temperature references for the Instrument units, for parasitic heat leaks (harness, struts, radiation) interception and for cryo-cooler pre-cooling. The last V-Groove, VG3, defines the coldest passive environment of the PLM and accommodates the Instrument modules and the Telescope Assembly (Figure 36). The V-Grooves structure, light but rigid, is composed of an Al sandwich with an internal honeycomb structure. The thermo-optical efficiency of this radiators system relies on angled highly-reflective surfaces open to space, rejecting radiation after a number of reflections between the angled shields (see Figure 37). Only the upper surface of the third radiator (V-Groove 3), always exposed to deep space during operations, is coated with a black painted honeycomb structure to maximize heat rejection. The telescope is required to operate at a temperature <70K. The Telescope Assembly (TA), enclosed in the cold environment established by the last V-Groove, acts as an extra passive stage using its large Baffle and Optical Bench (TOB) as radiating surfaces. These radiators, all arranged in the same black painted open honeycomb structure configuration, greatly improve the efficiency and the performances of the PLM passive cooling.
Figure 36: ARIEL PLM thermo-mechanical architecture The ARIEL Instrument modules are integrated on the Optical Bench and their cooling is achieved in two different ways, following the different temperature requirements of each frequency band. The detectors of two channels for the Fine Guidance System (FGS) / Visible Photometer (VISPhot) / Near IR Spectrometer (NIRSpec), located in a single module box (the FGS box) are passively cooled to T ≤ 70K by a dedicated radiator represented by the top surface that closes the modules cavity on the Optical Bench. This radiator, fully enclosed in the cold radiative environment set by the last V-Groove, always faces the cold space during operations. The ARIEL Infra-Red Spectrometer (AIRS) detectors must be operated colder, below 42K (see Table 13), with the goal of reaching a temperature around 36K, to minimise detector noise. Maintaining this temperature, with a load of tens of mW, require the implementation of an active cooling system. The cryocooler baseline relies on the Planck mission and EChO study heritage: a JT cold end fed by a Planck-like mechanical compressor using Neon gas isenthalpic expansion to achieve the required low temperature and
Page 61 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 heat lift (see Chapter 8 dedicated to the cooler in this document). The design of the ARIEL active cooling system is under the responsibility of the Rutherford Appleton Laboratory (RAL) in the UK.
Figure 37: ARIEL S/C attitude and SAA in orbit The general scheme of the PLM, shown in Figure 35, indicates the six main thermal interfaces of the PLM identified in the analysis. The design of the PLM thermal architecture is driven by the required operating temperature of each unit and interface, in combination with the expected loads at the thermal stages. These values are reported in the following Table. Thermal Interface (TIF) temperature and expected load1 [mW] Instr. SVM VG1 VG2 VG3 TOB/BAF JT CE ARIEL PLM thermal requirements Rad. (TIF 0) (TIF1) (TIF2) (TIF3) (TIF 4) (TIF 6) (TIF5) T_Op T_Op T_Op T_Op T_Op T_Op [K] T_Op [K] [K] [K] [K] [K] [K] T_Op ∆T4 2802 ± 10 Payload Unit ≤ 180 ≤ 120 ≤ 70 < 60 < 60 < 40 [K] [K] 1903 ± 10 Telescope < 70 ± 1 - - - - TBD5 - - FGS Optics ≤ 60 ± 0.5 ------FGS-1 detector ≤ 70 ± 0.1 - - - - - 206 - FGS-1 CFEE ≤ 70 ± 2 - - - - 1007 - - FGS-2 detectors ≤ 70 ± 0.1 - - - - 206 - FGS-2 CFEE ≤ 70 ± 2 - - - - 1007 - - AIRS Optics ≤ 60 ± 0.5 ------AIRS detector 0 ≤ 42 ± 0.05 ------156 AIRS detector 1 ≤ 42 ± 0.05 1007 156 AIRS CFEE ≤ 62 ± 2 - - - - 1007 - - Parasitic leaks1 (struts + harness + piping + NA NA -10000 TBD8 TBD8 TBD8 50 10 5 radiation) Total load1 (no margin) [mW] -10000 TBD TBD TBD 450 50 35 Total load1 (w/ 50% margin) [mW] -15000 TBD TBD TBD 675 759 52 Notes: 1 Based on thermal analysis of present PLM design 2 Conductive interface to SVM 3 Radiative interface to SVM
Page 62 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
4 Peak to peak value over a typical observation time (7 hours) 5 M1 thermal control system dissipation, if needed, is expected to be < 2W 6 FPA loads estimated with margin. The AIRS numbers already include the contribution of the FPA support and harness; ~5 mW are the allocated for temperature control 7 Worst Case dissipation 8 To be evaluated by thermal analysis 9 Radiator with A ≥ 0.3 m2 radiator Table 13: Main thermal requirements and expected loads for the ARIEL payload Besides the third V-Groove, the two main PLM radiators are the instrument cavity cover and the Telescope Baffle. The baseline design of their external surfaces is based on a black painted open honeycomb cell structure to maximize IR emissivity (ε ≥ 0.9). Instrument radiative thermal control is achieved by properly selecting the thermo-optical properties of the exposed surfaces. The radiative environment for the modules is set by the Optical Bench cavity and the Instrument Radiator that shield the channel modules and the common optics from the external environment. The remaining surface of the OB, directly exposed to cold space, is black painted (TBC). The internal surface of the bench cavity that accommodates the instrument modules and optics requires a black coating to minimize optical stray-light leaks (TBC). For the same reason, the module boxes are assumed externally treated with black paint or high emissivity coatings. In general, most of the ARIEL PLM units require high emissivity coatings to maximize passive cooling performances. As anticipated in the present issue of the Mission Analysis Guidelines document, there is the possibility of Sun intrusion in the PLM after launch, during the LEOP (TBC). For this reason transient simulations have been carried out with the TMM to evaluate the impact of direct solar illumination on the most sensitive units (detectors and optics). The preliminary results indicate that the temperature of these units is not rising to dangerous values if the exposure is limited to periods not longer than 60 minutes. The thermal inertia of the PLM helps in keeping the internal temperature within a safe range. If prolonged Sun exposures will be predicted a different approach for the thermo-optical design of radiating surfaces may be required. A possible alternative could be represented by flat surfaces (no open honeycomb structure) coated by lower solar absoprtance layers such as white paint. The FGS and the AIRS modules share a similar thermal design (see Figure 39). Both channels are integrated in a box that includes optical elements and a detector assembly, composed by the Focal Plane Assembly (FPA) and the cold front end electronics (CFEE). Due to electrical performance issues the cryo-harness connecting the Cold Front End Electronics (CFEE) to the detectors cannot be longer than 10 - 15 cm max. From this follows that the CFEE shall be mounted on each module box in proximity of the detectors working at the same temperature of the Optical Bench units. By tuning their thermal coupling to the OB it could be possible, if needed for performances optimization, to use their internal dissipation to keep them at a slightly higher temperature. CFEE loads are finally rejected to space by the main radiator stages. The FGS / VISPhot / NIRSpec channels are integrated in a single Module Box and work in the same temperature range, with the optical units and the detectors at a temperature, respectively, ≤ 60K and ≤ 70K. For the detectors Focal Plane Assembly (FPA) the general rule is the colder, the better. For this reason the detectors are cooled by a dedicated passive radiator stage, the Instrument Box top face (TIF5 in Figure 35) located inside the cold environment set by the third V-Groove and the Telescope Assembly. This radiator is mechanically supported on the Instrument Box by means of insulating supports and is under Instrument Consortium responsibility. The FGS detectors are thermally decoupled from the Module Box and high conductive links connect them, through their thermal control stages, to this radiator. The Module Box of the FGS channels is thermo-mechanically linked to the bench by means of a conductive interface. In this configuration, at steady state, the FGS optical units are expected to thermally equilibrate with the Optical Bench. The AIRS detectors technology baseline requires lower operating temperatures, on the order of 40K (T ≤ 42K with a goal of 36K), to achieve the required sensitivity. This temperature is reached by using the Ne JT cooler, capable of producing a heat lift up to of 50 mW or more. The AIRS module optics shall operate at a temperature similar to the other channels, 60K or below. For this reason, while the AIRS Module Box is thermally coupled to the OB, the FPA needs to be carefully insulated from the box, to limit the heat leak to the JT cooler cold end. In order to provide the required cooling to the AIRS detectors, the cold end heat
Page 63 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 exchanger is located in proximity the FPA, to minimize the thermal link distance, and supported on the Optical Bench by insulating struts. In general, each detector stage is thermally decoupled from the relative module box, to maximize performances of the FPA in terms of absolute temperature and stability. Coupling to the temperature reference stage is achieved through a Thermal Control Stage (TCS): this is an active closed loop thermal control system composed by a heater + thermometer couple driven by the ICU/TCU and located on the SCA’s frame. The thermal control stage of each detector is then directly coupled to the relative reference temperature stage (JT cold end or instrument radiator) by means of high conductivity thermal straps made by high purity Aluminum. The warm electronics and cooler compressor are located in the SVM. All harness and piping from SVM to Instrument channels should be thermally linked to all passive stages (VG1, VG2, VG3) to minimize parasitic leaks on the instrument cooling stages. In particular, cryo harness heat leaks to detectors shall be controlled by thermally optimizing the cables design with respect to the required electrical performances (see Chapter 7.4.4).
7.2 THERMAL INTERFACES DEFINITIONS In the ARIEL PLM thermal architecture seven main Thermal Interfaces (TIF’s) have been identified (see Figure 35): one to the S/C and the others internal to the PLM. The system thermal design has been based on the flow down of the basic instrument requirements (Table 13) to the main thermal interfaces (Table 14). The IF temperature values are fixed by the detectors/optics operating point and by the assumed total conductance from these units to their interfaces. The thermal stability requirements of the interfaces over a typical exposure time is set to ensure the required stability of the module units. The stability across longer periods, such as seasonal or orbital changes over mission, must be taken into account when dimensioning the interfaces and the relative couplings. The requirements in Table 14 are defined to ensure best thermal performances of the optical and detector systems over longer periods and full mission lifetime.
Figure 38. SVM thermo-mechanical interfaces dimensions The PLM interface to the SVM is the S/C platform: a 2700 mm circular surface cut in the Y axis direction at a distance of 2400 mm from the X axis (Figure 38). For thermal analysis purposes this platform is assumed as split in two interfaces: a conductive plate and a MLI radiative shroud. The platform is considered, for both interfaces, a fixed boundary, fully shielding the cold PLM from Sun and warmer section of the S/C, with the following characteristics: - Conductive coupling:
Page 64 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
o temperature range 270 – 290K o stability over an observation period +/- 3 K over 10 hours - Radiative coupling:
o IR emissivity = 0.05 o temperature range 180 – 200K with stability over an observation period +/- TBC K All conductors linking warm units in the SVM to the PLM are considered coupled to this interface (including harness and cooler piping) but the main conductive couplings of the PLM with this interface are the bipods 2 2 and V-Groove struts. The contact surface of each bipod foot is around 50 cm (for the struts Ac ≈ 5 cm ) and the assumed surface contact conductance for these couplings is 800 W/m2-K, a conservative value for bare interfaces (no thermal filler) with A < 100 cm2. For a cryogenic instrument, internal and external thermal interfaces are the key to a successful design. The detectors must be thermally decoupled from the OB, by resistive supports to improve their stability, while thermal contact with the cooling stage (JT cold end or cold radiator) must be maximized. This requires that the cooler cold end is located nearby the AIRS module to minimize the thermal path of straps (see Figure 39). The JT cold tip is mounted at a convenient position on the Optical Bench by means of insulating supports. The Instrument Radiator is integrated on the Optical Bench through insulating connection points. Presently, the allocated area for this radiator is defined by the dimension of the open cavity that encases the instrument modules, but there is still margin to increase its size, if needed. High conductivity thermal straps are used to connect the FGS and AIRS channel detectors to their operating temperature reference, respectively the Instrument Radiator and the JT cold heat exchanger.
Figure 39: Modules thermo-mechanical configuration on the OB The required conductance across each thermal interface is evaluated by analysis running thermal model simulations (see Chapter 7.4). A typical value of 500 W/m2-K is assumed as the average surface thermal conductance of machined metallic interfaces in the temperature ranges expected for the ARIEL PLM. This value was achieved with enough margin in several Planck couplings in the 20-55K range, based on M4 bolts, using spring washers and, in some cases, a filler (gold sheet). The required conductance values and limits at the ARIEL TIF’s have been bounded and checked by running preliminary parametric analysis. A justification of the assumptions made on the main conductive links is reported in the Technical Note describing the TMM/GMM. The high level specifications of the thermal couplings of PLM units to the main interfaces are summarized in the following table. The values refer to a single coupling.
Page 65 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
1 ∆T (K) G across IF Contact I/F Unit I/F to T (K) Vs time (W/K) Area (cm2) TIF0 SVM SVM top platform < 290 < ± 3 < 10 < 100 TIF1 VG1 Warm Harness, cooler piping ≤ 200 ≤ ± 3 ≥ 0.5 ≥ 10 TIF2 VG2 Warm Harness, cooler piping ≤ 120 ≤ ± 1 ≥ 0.5 ≥ 10 TIF3 VG3 Warm Harness, cooler piping ≤ 70 ≤ ± 1 ≥ 0.5 ≥ 10 Module Box, CFEE, Mirrors, Harness ≥ 0.5 ≥ 10 Telescope mirrors < 0.1 < 100 TIF4 TOB/Baffle ≤ 60 ≤ ± 1 Detectors FPA < 0.001 < 2 JT cold end < 0.01 < 5 TIF5 I Radiator FGS detectors ≤ 60 ≤ ± 0.1 ≥ 0.1 ≥4 TIF6 JT cold end AIRS detectors < 40 ≤ ± 0.1 ≥ 0.1 ≥2 Notes: 1 Peak to peak value over a typical observation time (10 hours) Table 14: TIFs main requirements Each detector FPA is supported on a flange that works as the active Thermal Control Stage (TCS) by thermally dimensioned supports to achieve a tuned thermal break. TIF5, for FGS detectors is located on the back side of the radiator. TIF6, for the AIRS FPA, is the JT cold end heat exchanger, supported on the bench by insulating struts and is located in proximity of the spectrometer box side where the detectors are positioned. The assumed contact area is 2 cm2. The conductance values defined in this study and assumed for the thermal analyses can be achieved using standard materials and solutions adopted already in previous experiments (e.g. MIRI and Planck, [RD14]): GFRP for insulating struts (Ti alloy can be used for specific fixtures) and 5N Al (or Cu) for the high conductivity links. Standard Al alloys (such as 6063 or 6061) are used for most of PLM structures and units as well as for the telescope structure. Stainless steel (TBC) bolts (A2-70) not smaller than M4 shall be used for the main mechanical couplings to Spacecraft and to the TOB. In general, to optimize the thermal contact, the maximum bolt dimension allowed by the mechanical allocations and design should be used. If needed, spring washers and a thermal filler (Gold or Indium foil for example) could be considered to improve conductance.
7.3 PAYLOAD MODULE THERMAL CONTROL HARDWARE The ARIEL Thermal Control Hardware (TCHW) includes all passive or active components that are used to reach and maintain the operating temperatures of the PLM units within their required ranges. The high level list of the PLM TCHW items is composed by: − V-Grooves (including bipods and struts) − Instrument Radiator (top face of the box) − Telescope Baffle (upper half) − Thermal straps − Thermistors (monitoring and control) − Detectors temperature control stage (TCS) − Telescope Mirrors temperature control stage − Heaters The first five items are fully passive components while the other three can be considered part of the active thermal control system.
7.3.1 V-Grooves V-Grooves (VG’s) design is a key issue of the ARIEL thermal performance as they represent the first cooling stage of the PLM. VGs are high efficiency, passive radiant coolers, whose performances in a cold radiative environment such as L2 has been definitely demonstrated by the Planck mission. ARIEL V-Grooves system
Page 66 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 consists in a set of three specular shields, composed by six half circles arranged in a “V-shaped” configuration, angled along the diameter parallel to the S/C X axis (Figure 40). Their objective is to intercept radiative and conductive heat leaks from warmer sections of the S/C and spread it away to deep space after multiple reflections between each VG pair. Each VG shield is tilted by a certain angle with respect to the adjacent one, creating a divergent radiative path for the reflected thermal rays. Figure 40 shows a very simplified schematic of this concept. In the present thermal configuration the three VG angles are 10°-15°- 20°. The thermo-optical properties of the VGs are of essential importance since they are the key parameters for thermal isolation and heat rejection to space. To achieve the required performances, VGs surfaces must have a very low emittance coating, a high reflection/mirroring material needed to reflect heat radiation. A 3N pure aluminium coating with a Vapour Deposited Aluminium (VDA) can reach an emissivity in the IR band of less than 0.05, as measured on the Planck PLM. The upper surface of the last VG (VG3), exposed to the cold sky, is covered by an open aluminium honeycomb structure coated with high emissivity black paint (MAP® PUK, or Aeroglaze Z306) to maximize the radiative coupling, and so heat rejection to deep space. The bipods (Figure 41) and V-Grooves struts are hollow tubes made of GFRP (R-glass + epoxy), with rigid foam inside to increase their mechanical performances while limiting thermal conductance across the stages. The bipod legs are 620 mm long, with a diameter of 50 mm and a thickness of 5 mm. The pod feet are made of Al alloy with Ti alloy (Ti6Al4V) fixtures. The smaller front bipod legs are 220 mm long, with a diameter of 30 mm and a thickness of 3 mm. During flight operations all bipod legs will always face the cold sky so their external surface is black painted to maximize self heat rejection to space. This configuration ensures a very limited heat leak through the length (G < 10-4 W/K).
Figure 40. ARIEL V-Grooves scheme Even if bipods and struts are good thermal insulators, to minimize heat leaks to the PLM colder stages, they must be thermally coupled to the V-Grooves. Each mechanical support is connected, by means of at least two thermal straps, to each VG as shown in Figure 41. For further parasitic loads reduction all bipods and struts are black painted to increase self-rejection to space. For the same reasons, the harness from the warm electronics in the SVM to the cold units is thermally coupled to each VG in order to reject its conductive load. The JT cooler piping dissipates the gas pre-cooling load on the V-Grooves by means of heat exchangers located on the shields. As a conservative case for the thermal analysis we have considered the harness interception load on VG3 and a pre-cooling power of 320mW entirely dissipated on one node of the last VG only. Even in these conditions, a thickness of less than 1 mm of the Aluminum skin is sufficient to limit the temperature gradient over the shields to less than 3K between the hot spots (cooler & harness heat exchangers) and the rest of the panel.
7.3.2 Telescope Baffle and Instrument Radiator The V-Groove-based design provides a cold and stable environment for the PLM units: telescope, instrument and cryocooler cold end. In this volume all main surfaces exposed to cold space can work as radiating units to increase margins and performances of the PLM passive design. For this purpose the main surfaces are the Instrument Radiator (the top face of the Instrument Box) and the top half of the Telescope Baffle. Potentially also the exposed areas of the Optical Bench or the side walls of the Instrument box can help in
Page 67 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 this direction and it is assumed that all these surfaces will be coated by a high IR emissivity layer (black paint, MAP® PUK, or Aeroglaze Z306).
Figure 41. Bipods configuration The top half of the Telescope Baffle will always directly face the cold sky during nominal operations and, given its large surface, offers a great chance of improving the passive performances of the ARIEL S/C. For this reason the surface shall be optimized in terms of IR emissivity: at present the Baffle is designed as a black painted sandwich of Al6061 alloy layers. A honeycomb cell structure 12 mm thick, with 10 mm cell size and ribbon thickness of less than 0.1 mm on the external surface, supported on a 1 mm thick skin (see Figure 42). The internal surface of this skin, that faces the telescope cavity inside the baffle, is black painted also for stray-light control purposes. This thermo-mechanical design reduces mass while ensuring at the same time a very good thermal conductance in all directions and mechanical stiffness. The baffle has a mass of ~10 kg and is mechanically supported on the two stiffening arms of the telescope structure close to the middle position along its main axis. In this way thermo-elastic contraction will not apply stress to any other unit. The baffle is also thermally connected with high conductance straps to the Optical Bench itself to provide a single large passive stage (TIF4) for all instrument units.
Figure 42: Example of Planck Baffle thermo-optical configuration
Page 68 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The top face of the Instrument Box is used as a dedicated radiator for detectors (TIF5). The FGS and NIR Photometer channel requirements (see Table 13) in terms of operating temperature, stability and dissipated loads need a cold radiator dedicated to detectors cooling only. This extra passive stage, called Instrument Radiator, like the Telescope Baffle, is designed and developed under the responsibility of the Consortium. The radiator is mounted directly on the optical bench, covering the cavity that accommodates the channel modules, by means of insulating supports. Even if the operating temperature of the radiator will be very similar, if not equal, to the TOB and baffle, this configuration would decouple the radiator from any possible instability generated by components functioning in the Box, helping the thermal control process of the FGS channels FPA. In this way the radiator is fully devoted to cool the channel detectors, with parasitic leaks due only to its own supporting struts and thermal monitoring harness. The expected active load due to detector stages is around 30 mW. The conductive load on the radiator due to struts and harness is estimated to be less than 10 mW. In total, 75 mW (including 50% margin) is the expected load to the cold radiator, in the worst case. At present, the Instrument Radiator approximated dimensions are 600 mm x 635 mm, with a top face area around 0.37 m2, with the exposed surfaces of the Optical Bench contributing for an extra 0.3 m2. In case a larger margin is needed, a wider surface (in the 0.4 – 0.5 m2 range) of the Instrument Box top could be easily fit in the allocated volume on the bench if needed. The radiator orientation is parallel to the OB with an angle around 12° with respect to the vertical direction. Its optical properties are the same of the Telescope Baffle as the outer surface is based on the same radiative configuration: a 12 mm thick black painted honeycomb cell structure of Al6061 alloy, 10 mm (or less) cell size and ribbon thickness of 0.1 mm mounted on 1 mm thick skin. As demonstrated by Planck, this thermo-optical configuration is capable of rejecting more than 300 mW per square meter in L2 orbit. The ARIEL G/TMM results seem to confirm this assumption. The internal surface of the radiator is, at this stage, assumed to be black painted due to stray- light control. A low emissivity coating, if acceptable, may help in gaining some more margin on the radiator performances, in case is needed, by limiting possible radiation loads from the warmer parts of the OB.
7.3.3 Thermal straps The main conductive links of the ARIEL PLM units are based on high purity 5N Al braids (wires or foils). Because of the high thermal conductivity of pure Al at low temperatures, and its low density, it is possible to maintain dimensions, and mass, of the braids within allocations. The straps are used to thermally connect: − the bipods to the VGs for conductive parasitic leaks interception: one (or two) straps per bipod leg per VG; − the Telescope Baffle to the Optical Bench: four straps; − the channels FPA’s to their temperature reference (radiator or cooler cold end): one per detector. The total conductance across a strap is the combination in series of conductance through the flanges and braids. In the 40-60 K range, the average conductivity of high purity Al (5N) is around 1000 W/m-K. The total conductance of a metallic braids thermal link is commonly assumed to be
=
𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑘𝑘𝑘𝑘 𝑝𝑝 𝑠𝑠 𝑒𝑒 where η indicate efficiency factors that take𝐺𝐺 into account∙ 𝜂𝜂 realistic𝜂𝜂 𝜂𝜂 inefficiencies in the density of wires (or foils) per unit area, in the effective length of the link (due𝐿𝐿 to turn and bends) and in the welds of the braids to the end flanges. Assuming typical conservative values for these factors and imposing a safe estimation of the distance between the units, it is possible to evaluate dimensions and mass of the braids required to achieve the required G across each link.
Page 69 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Total G Effective Section End flanges Units N of straps Length (m) Tot mass (g) (W/K) Length (m) (cm2) dim (cm) Bipods ≥ 0.2 18 0.05 0.1 34.8 20x20x3 291 Tel Baffle > 1.0 4 0.2 0.25 755.9 40x40x3 2219 FGS-1 Det ≥ 0.1 1 0.3 0.35 58.2 20x20x3 63 FGS-2 Det ≥ 0.1 1 0.3 0.35 58.2 20x20x3 63 AIRS Det ≥ 0.1 1 0.2 0.25 41.6 20x20x3 35 Total mass (g) 2671 Total mass with margin (50%) 4006 Table 15. ARIEL thermal braids preliminary dimensions
7.3.4 Thermistors The ARIEL number and specifications of the thermometers have been defined on the basis of the present knowledge of the PLM units and TIF’s design. Detailed temperature monitoring is achieved by the combination of direct measurements with units thermal prediction analysis, correlated with all ground test results at sub-system and system level. For the instruments units and the cooler cold end monitoring, Cernox sensors are the baseline (TBC). For the telescope and other PLM units (such as the V-Grooves) PRT’s or diodes can be used, using Cernox only for thermal control purposes (mirrors, critical interfaces etc.). On large surfaces (benches, M1), when thermal gradient need to be monitored with accuracy, thermocouples may become a possible alternative to reduce the number of thermistors. A preliminary selection indicates the following types: - Cernox 1050-1070 - DT-670 Silicon diodes - PT100 PRT The full number of sensors needed to monitor the PLM from the ICU/TCU in flight has not been finally defined yet. This number must be kept as low as possible in order to minimize the number of wires and harness complexity. At the same time the thermistors shall be enough to ensure, in combination to the units thermal mapping resulting from the correlated TMM, a complete monitoring of the PLM during flight operations. A thermometer is installed in correspondence of each TRP, main thermal interface or critical item. At this stage, all thermistors are assumed to be fully redundant, with the backup items connected to the redundant ICU/TCU. The readout/acquisition rate of the temperature of critical items (detectors, control stages, cooler cold end, etc.) shall be 1 Hz. The fully passive units (such as the V-Grooves, the Baffle, the Bench) can be monitored at a relaxed rate (with periods of tens of seconds), especially if they result dominated by low frequency variations. All thermistors reading shall be based on 4-wires measurement with connections to the readout electronics arranged in shielded twisted pairs to minimize EMI from external sources. A resolution of at least 20 mK and an accuracy of 50 mK are required for the PLM units when they are in their operational range. The thermistors used for detectors thermal control should have a resolution of at least 10 mK for the detectors and 100 mK for the M1 in a narrow (few K) range around the operating temperature. A first estimation based on the present design can be provided (TBC): - 26 nominal (+26 redundant) Cernox sensors - 20 nominal (+20 redundant) diodes or PRT In summary, a total of nearly 50 sensors shall be acquired by the ICU/TCU (TBC): approximately 60% of them will be read at 1Hz frequency while the remaining at a slower rate (TBD).
Page 70 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Extra thermal sensors may/will be integrated on the PLM by the spacecraft Prime Contractor but these will be monitored by the S/C electronics and will not be under Consortium responsibility. The preliminary evaluation of thermal sensors for PLM monitoring is summarized in Table 16.
7.3.5 PLM active thermal control system One of the tasks of the ICU/TCU is the thermal control needed to stabilise the units during normal scientific operations. The process is based on reading the PLM thermistors and controlling the heaters according to the general specifications summarized in this chapter. All other thermal control operations such as decontamination and survival are, at the moment, considered under S/C responsibility. Thermal monitoring for these procedures will be based on thermistors that are part of the S/C. In this way the two thermal monitoring/control chains (S/C and PLM) are considered independent, each responsible of the relative sensors/heaters. The PLM units that may require active control are the telescope mirrors and the detector focal planes. At this stage of the design, it is assumed that the primary mirror M1, the secondary M2 and M3 shall be actively controlled. The instrument channel detectors require active thermal control, as the stability requirement is demanding. Thermal control is based on a feed-back loop, Proportional-Integral-Derivative (PID) type controller, operated with a frequency rate of 1Hz (TBC) as the relevant fluctuations expected on the controlled units should have typical time periods much longer that 1 second. All possible faster fluctuations, for example cooler compressor induced (around 40 Hz, TBC) if relevant, are filtered out passively by the thermal inertia of the units. Controlling every second will allow to respond efficiently to all expected oscillations. Each control loop should be able to read the temperature and adjust heater power, if needed, within a 1s cycle. All thermal control stages are assumed to be fully redundant: each one allocates nominal and redundant monitoring thermistors and heating resistances. For each control stage (the four detectors FPA, M2 and M3), a thermistor reading shall be acquired and a heater adjusted every second. Mirror M1 dimensions may need a more accurate monitoring. In this case the control loop may be based on the average reading of more than one sensor, possibly up to all five allocated for M1. Only one heating line is expected for the mirror. Multiple heaters (in series or in parallel, TBD) will be energized by this single line to optimize heat distribution.
Page 71 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Unit Position Number Type Resolution Accuracy TCS 1 (+1 Red) 0.025K for T ≤ 60K 0.050K for T ≤ 60K Cernox FGS-1 DS 1 (+1 Red) 0.1K for T > 60K 0.25K for T > 60K Box (TBC) 1 (+1 Red) DT-670 0.25K 0.25K TCS 1 (+1 Red) 0.025K for T ≤ 60K 0.050K for T ≤ 60K Cernox FGS-2 DS 1 (+1 Red) 0.1K for T > 60K 0.25K for T > 60K Box (TBC) 1 (+1 Red) DT-670 0.25K 0.25K 5 mK for 32K< T <42K 0.050K for T ≤ 60K TCS 1 (+1 Red) Cernox 0.025K for T ≤ 60K 0.25K for T > 60K 0.1K for T > 60K AIRS 0.025K for T ≤ 60K 0.050K for T ≤ 60K DS 1 (+1 Red) Cernox 0.1K for T > 60K 0.25K for T > 60K Box (TBC) 1 (+1 Red) DT-670 0.25K 0.25K 0.025K for T ≤ 60K 0.050K for T ≤ 60K JT cooler Cold end 3 (+3 Red) Cernox 0.1K for T > 60K 0.25K for T > 60K IF to detectors straps (hot spot) 2 (+2 Red) 0.025K for T ≤ 60K 0.050K for T ≤ 60K I Rad Cernox (TBC) Opposite positions (max grad) 2 (+2 Red) 0.1K for T > 60K 0.25K for T > 60K IF to Radiator 1 (+1 Red) IF to Bipods 3 (+3 Red) 0.025K for T ≤ 60K 0.050K for T ≤ 60K OB Cernox (TBC) Opposite positions (max grad) 2 (+2 Red) 0.1K for T > 60K 0.25K for T > 60K Critical optical units 2 (+2 Red) 0.025K for T ≤ 60K 0.050K for T ≤ 60K M1 5 (+5 Red) Cernox (TBC) 0.1K for T > 60K 0.25K for T > 60K Telescope M2 + M2M 2 (+2 Red) DT-670 (TBC) 0.25K 0.25K Baffle 3 (+3 Red) DT-670 (TBC) 0.25K 0.25K Bipods IF 3 (+3 Red) DT-670 (TBC) 0.25K 0.25K VG3 Precooling/harness IF 3 (+3 Red) DT-670 (TBC) 0.25K 0.25K VG2 Bipod/harness/piping IF’s 3 (+3 Red) DT-670 (TBC) 0.25K 0.25K VG1 Bipod/harness/piping IF’s 3 (+3 Red) DT-670 (TBC) 0.25K 0.25K
Total ARIEL PLM thermistors 46 (+46 Red) = 26 Cernox + 20 Diodes
Table 16: Preliminary list of ARIEL PLM thermistors The detailed design and power requirements of the temperature control stages must rely on a realistic estimation of the possible level and time spectrum of the fluctuations on each sensitive unit or at its interfaces. The TMM can provide indications of the power (peak/max, average and power resolution) needed to be applied to the mirror(s) and detectors during operations, once the expected instabilities at the main interfaces are considered. A preliminary evaluation of the power needs for the ARIEL PLM units has been carried out by analysis. The power allocated for detectors thermal control is 5 mW (TBC), as the SCA’s are “light” units in terms of thermal capacitance. Moreover the control stage will be designed to minimize the control power by using Page 72 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 tuned thermal resistances towards the source of instability. For the telescope control, the required power depends mainly on the final mass of the mirror(s) and on the time behaviour of the fluctuations. The light- weighting level of the flight M1 will have an impact on its thermal behaviour and stability. With the present thermo-mechanical design the TMM prediction for M1 max decontamination power is around 6W, with a 50% margin that includes also the fraction for M2 and M3. In the other S/C Modes the thermal control may be limited to: − cooldown phase during transfer to L2 (first few weeks of flight). For decontamination purposes the mirrors temperature should be kept around or above 170K (TBC) with the telescope baffle at 160K, for approximately the first two weeks of flight operations. With the present thermo-mechanical design the TMM prediction for M1, M2 and M3 max decontamination power is around 40W (with a 50% margin). This is the power needed to take M1 temperature to 170K once the mirror is already operating at 50K, in case a decontamination run is needed during cold operations. The control logic for decontamination can be based on a simple proportional loop; − Survival/Eclipse Modes. In these operative Modes the thermal control shall be able to maintain the units temperature in their survival range (TBD). Survival heaters control can be based on a proportional control loop. Survival/decontamination heaters control should be under S/C responsibility and will not be part of the ICU/TCU functions. T_Op Op T range ∆T in time Control power Payload Unit [K] [K] [K] [W] Telescope M1 < 70 50 – 70 ± 1 0 – 6 FGS-1 detector ≤50 50 – 70 ± 0.1 0 – 0.005 FGS-2 detector ≤50 50 – 70 ± 0.1 0 – 0.005 AIRS detectors ≤42 35 – 42 ± 0.05 0 – 0.005 Table 17: PLM units active thermal control required performances To date the required stability for the telescope mirror (see Table 17) is a band 2K wide (± 1K) around its operating temperature in the 50 – 70K range, in a time reference of one observation run (typically 10 hrs TBC). For the detectors we assume that their stability level, in the time reference of one observation run (typically 10 hrs TBC), will be ±0.1K (a 200 mK wide range) around the operating T for the FGS and NIR photometer and ±0.05K (10 mK peak-to-peak) for the AIRS.
7.3.5.1 Detectors temperature control stages Detectors thermal control is achieved by the combination of passive and active damping processes. The passive component exploits the thermal inertia (R and C) of the detector system components (struts and frames) to damp temperature oscillations during their propagation from the instability source (the Instrument radiator or the JT cold end) to the detectors. The finer active control is accomplished by a PID type controller on the Thermal Control Stage (TCS). This is composed by the detector supporting flange and the active closed loop thermal control system: a heater plus thermistor couple driven by the warm electronics. Since detectors thermal stability is critical for instrument performance, a fully redundant system with two identical heater and sensor pairs is foreseen. Each detector stage is thermally decoupled from the relative module box or optics, to ensure optimal performance in terms of absolute temperature and stability. Coupling of the detectors to the temperature reference stage (cooler cold end or radiator) is achieved through high conductance links (high purity Al straps) that connect to the TCS. The TCS works as an intermediate stage between FPA and cooling point, so it must be partially decoupled from both sides to act as a passive filter by increasing the system time constant and to minimize the power needed for active control. The optimal conductance for the ARIEL detectors control stages has been evaluated by thermal analysis. Model results indicate that for the T control stage a conductance on the order of 0.1 W/K to the cooling stage (see Table 15) and a thermal coupling 10 times more insulating (G ~ 0.01 W/K) to the SCA’s, are an acceptable compromise, in terms of control
Page 73 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 power and absolute temperature requirements. A simple schematic view of the TCS configuration is shown in Figure 43.
7.3.5.2 Telescope Primary Mirror temperature control stage The telescope M1 is supported on the optical bench by means of three structure (Figure 44). The detailed design of these struts shall to take into account the opto-mechanical requirements to ensure the telescope optical stability. The struts are made of the same alloy of the mirror (Al6061) but the thermal conductance can be controlled at the three contact points by means of thermal washers. According to the results of the present analysis, a conductance GM1 < 0.1 W/K (TBC) is sufficient for mirror temperature control. Together with the high thermal mass of the mirror, this level of conductive decoupling from the bench allows an efficient filtering of the instabilities expected, at this stage, at higher frequencies (with periods on the order of seconds and minutes). The slower fluctuations (with periods longer than thousands of s) that could be transmitted to the optics will be smoothed by the active control system (Figure 44). This is composed by a heating line that could supply either strip heaters or film resistances (in series or parallel TBD) distributed inside the cavities of the light-weighting machining on the mirror large back surface to ensure a more even heat distribution. Control is based on a PID type feed-back loop responding to the temperature readings coming from a set of sensors located in key positions (inside the M1 back surface or on the struts) to monitor the thermal status of the mirror in all directions.
Figure 43: Detector modules thermal control stage scheme
7.3.6 Heaters Heaters are needed for thermal control of the channels detector and telescope mirrors, plus decontamination/survival. The control heaters for the detectors could be designed to operate also for decontamination and SCA’s annealing purposes (TBC), should this need arise. For PLM thermal control film and cartridge heaters will be used. Type and resistance is still TBD at the moment. For this reason, models or resistance range has not been selected yet: thermal analysis will provide the power range and the power resolution (minimum step of power to be adjusted by TCU) needed. On this basis the heaters (type, resistance) will be selected in collaboration with ICU/TCU team. In principle all heaters are assumed to be on the 28V power line and they should be sized so that the main performance specification are satisfied even at the minimum expected voltage value from the S/C on this line during operations.
Page 74 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 44: Telescope M1: heaters and thermistors can be located inside the light-weighting cavities and/or on mirror supports At present the total number of resistors to be controlled by the ICU/TCU is 5: − 3 nominal heaters (+3 redundant) for detectors thermal control − 1 nominal (+ redundant) heating line for the telescope primary mirror (two if also M2 needs control)
o each heating line can feed up to 3 to 5 resistors connected in series or parallel to distribute heat more evenly on the large mirror surface − 1 nominal (+ redundant) heating line for the telescope decontamination/PLM survival
o each heating line can feed several resistors connected in series or parallel For detector control, power is expected to be in the range 0 - 2W (if detector annealing is an option), or smaller 0 - 0.1W, for each heater with a 100 uW power resolution, as the maximum power allocated for control during normal operations is 5 mW. For telescope control the range is on the order of 0 - 6W (TBC), due to the higher thermal mass of the mirror, with a lower resolution requirement (0.1 W, TBC). The power resolution will be defined on the basis of the expected fluctuations on the mirror. The control heaters will be operated with a PID type logic and connected to the control electronics by shielded twisted pairs to minimize EMI. The number of heating lines (as some of them could energize more than one resistor) are summarized in the following table.
Page 75 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Resistance Range Unit Position Number Type Resolution (mW) (Ω) (W) 0 – 0.1 0.1 (TBC) in 0-100 mW range FGS-1 TCS 1 (+1 Red) Cartridge/Film TBD 0 – 2 100 in the 0 - 2W range 0 – 0.1 0.1 (TBC) in 0-100 mW range FGS-2 TCS 1 (+1 Red) Cartridge/Film TBD 0 – 2 100 in the 0 - 2W range 0 – 0.1 0.1 (TBC) in 0-100 mW range AIRS TCS 1 (+1 Red) Cartridge/Film TBD 0 – 2 100 in the 0 - 2W range Telescope M1 1 (+1 Red) Cartridge/Film TBD 0 - 10 100 (TBC) in the full range 0 – 40 Decon/survival PLM units 1 (+1 Red) Cartridge/Film TBD TBD (TBC) Total PLM heaters 5 (+5 Red) Table 18: Preliminary list of ARIEL control heaters
7.4 THERMAL MODEL AND ANALYSIS RESULTS
7.4.1 G/TMM Description The GMM is based on the CAD model of the PLM. Small geometries and details which are not relevant for thermal analysis purposes are not considered in the GMM. In the model the main radiative surfaces and representative supporting structures between the different stages are simulated. In order to simulate the interaction with the Service Module, the model includes the SVM/PLM interface assumed as a boundary. The SVM/PLM I/F is composed by two surfaces with the dimensions indicated in Figure 38). The conductive one is called “SVM top plate”, operates at room temperature and is the main interface of the bipods and V-Grooves struts. The other surface is the “SVM Radiative shield”, located in between the SVM top plate and the VG1, and represents a purely radiative coupling with the rest of the model. The SVM top plate is a disc of diameter 2700 mm, cut to 2400 mm in the ±YARIEL axis. Figure 45 shows front views of the entire model in ESATAN-TMS. The SVM top plate is coloured in orange, the SVM Radiative shield in yellow. These two shells have assigned optical coating MLI which has a low value of emissivity (0.05). A more detailed description of the PLM TMM/GMM and the thermal analysis results is reported in a dedicated Technical Note [RD14].
Figure 45: ARIEL PLM geometric model view from two sides. On the right panel the radiator is the top of the instrument cover. The thermo-mechanical configuration of the telescope structure and the optical bench with the instrument modules represent a simplified version of the CAD design but includes all relevant thermal characteristics of each unit (Figure 46).
Page 76 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 46: A exposed view of the Instrument optical bench geometric model with the main units In order to study and predict the system thermal performance, several model runs in different boundary conditions have been performed. The main results are reported in [RD14]. Here we report the results of the steady-state analysis representing the PLM in nominal science operations in the two extreme temperature cases at the SVM interface.
7.4.2 Boundary Conditions For each significant set of boundary conditions associated with a scenario (radiative case) a thermal-analysis case is created and solved. Table 19 reports the boundary conditions assumed for the nominal operation cases, the only two reported in this document. The power dissipation of the active units are summarized in Figure 47. The pre-cooling load of JT cooler (320 mW as per RAL estimation) is assumed to be fully dissipated on a single node of VG3, as a sort of worst case. The JT cooler heat exchanger is, at this stage, simulated as a diffusion node with a fixed heat lift. Even if the baseline cooler for ARIEL will be capable of producing a cooling power up to 50 mW or more, an extra margin of 25% has been assumed on this number. In the present thermal design of the PLM, the JT cold tip can operate in the 35-36K range with 37.5 mW of heat lift produced. Analysis case Radiative case Boundary conditions Type of solution Cold case at nominal SVM top plate @ 270K (fixed) Earth-Sun L2 point Steady-state operational SVM Radiative shield @ 180K (fixed) Hot case at nominal SVM top plate @ 290K (fixed) Earth-Sun L2 point Steady-state operational SVM Radiative shield @ 200K (fixed) Table 19: Main steady-state operational cases boundary conditions
Page 77 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 47. Power dissipations for operational Cold and Hot cases The AIRS SCA’s active load assumed for the thermal analysis (Table 13) is 15 mW (with margin) but that number already includes the parasitic leaks (struts, harness) to the FPA that are the dominant contribution over the actual electrical dissipation (see AIRS Design Description Document). The TMM computes independently the heat leaks due to the struts (the main parasitic load ~6 mW). For this reason in the TMM the dissipation associated to the SCA’s is 10 mW each, which corresponds to an extra 35% margin assumed over the actual active load of the detectors stage.
7.4.3 Steady-State Analysis Results Table 20 shows the solved steady-state temperatures of the cold and hot case for the nominal conditions, with the S/C orbiting at the Earth-Sun L2 point in nominal attitude and loads conditions. The cold and hot case are different just for the fixed boundary temperatures applied at the SVM/PLM I/F. Hot case T Req Margin PLM Unit Cold case T [K] [K] [K] [K] SVM top plate (fixed boundary temperature) 270.00 290.00 - - SVM Radiative shield (fixed boundary temperature) 180.00 200.00 - - VG1 148.6 158.1 180 10 VG2 97.4 103.1 120 10 VG3 55.1 56.6 70 10 Telescope M1 49.4 50.4 70 10 Telescope M2 48.8 49.8 70 10 Telescope Baffle 48.2 49.1 60 10 TOB 49.6 50.5 60 10 Instrument Box 49.6 50.5 60 10 FGS CFEEs 50.1 51.0 70 10 FGS detectors 49.5 50.3 70 10 Instrument radiator 48.9 49.8 60 10 AIRS CFEEs 50.1 51.0 62 10 AIRS detectors 35.1 36.0 42 5 JT cold end 34.9 35.9 40 5 Table 20: Steady-state average temperatures of PLM units for the cold and hot case in nominal conditions. In the table above, the temperature results are compared to the requirements with the margins to be assumed at this stage of the design process. The temperature of all passively and actively cooled units are fully compliant to the requirements including margins.
Page 78 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The unit temperatures are graphically shown in the next figures, comparing the Cold and Hot cases: general views of the PLM are compared in Figure 48, while Figure 49 and Figure 50 show the details of the coldest units in both cases.
Figure 48: Cold Case (upper panel) and Hot Case(lower panel) steady-state results in nominal conditions, general views of the PLM model.
Figure 49: Cold case steady-state results in nominal conditions, detailed views of the cold PLM. Left view shows the radiator, OB and baffle temperatures. Right side: the module boxes are hidden to show AIRS and FGS CFEEs, detectors and the JT cold end.
Page 79 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 50: Hot case steady-state results in nominal conditions, detailed views of the cold PLM. Left view shows the radiator, OB and baffle temperatures. Right side: the module boxes are hidden to show AIRS and FGS CFEEs, detectors and the JT cold end. Figure 51 reports a detailed view of the Telescope Assembly in the Cold Case, to show the thermal uniformity reached by the passive cooling. M1 maximum gradient is lower than few mK.
Figure 51: Telescope Assembly nodes temperature in the Cold Case
The loads to the main PLM internal interfaces are evaluated by balancing the input/output heat fluxes to/from each node through all conductors. The most relevant heat exchanges and rejection capabilities are reported in Table 21. PLM IF main heat flux Cold case [W] Hot case [W] VG1 heat rejection to space 7.6 9.8 VG2 heat rejection to space 1.1 1.4 VG3 heat rejection to space 1.0 1.1 Telescope Baffle heat rejection to space 0.7 0.8 Instrument radiator heat rejection to space 0.096 0.103 TOB heat rejection to space 0.037 0.039 Conductive heat flux from bipods’ heads to TOB 0.12 0.13 Conductive heat flux from TOB to Mirror 1 0.010 0.012 Conductive heat flux from TOB to Baffle 0.350 0.360 Conductive heat flux from CFEEs (both FGS and AIRS) to TOB 0.37 0.36 Conductive heat flux to FGS detectors 0.017 0.019 Conductive heat flux to AIRS detectors 0.017 0.017 Table 21: Heat exchange at the main internal interfaces and between units for Cold and Hot cases in nominal conditions.
Page 80 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The V-Groove 3 is capable of rejecting more than 1W at a temperature of 55K, while the telescope baffle and the instrument radiator dissipate to space, respectively, 0.7 W and 0.1W.
7.4.4 Harness preliminary thermal analysis Detectors wiring plus all the service harness (general HK, thermal control, heaters, thermistors etc.) consists of a high number of electrical connections. An appropriate selection of materials combined with an optimal combination of gauge wiring and harness length allows to keep the conducted load from the warm SVM to the coldest stages within few tens of mW. In order to minimize parasitic loads to TOB, radiator and cooler cold end, all the passive stages should be used for heat leak interception. In particular, the last V-Groove (VG3) is the main and final harness thermalization stage to minimize the load of the FEE on the detectors. At the same time, working as a conductive link, the warm harness can also be used to efficiently remove cold electronics heat load towards the VG3, if needed. A preliminary estimation of the possible heat loads due to instrument wiring on the stages, has been carried out on the basis of conservative assumptions (ref. [RD26]). The results, in terms of heat leaks to the thermal stages, are summarized in Table 22. Scaling the values resulting from the EChO harness conservative thermal analysis, a preliminary estimation of the ARIEL harness leaks has been calculated: VG1 VG2 VG3 TOB JT cold end Harness (TIF1@190K) (TIF2@120K) (TIF3@60K) (TIF5@50K) (TIF6@40K) [mW] [mW] [mW] [mW] [mW] Detectors control1 182 101 64 8.0 TBD Thermal control2 64 35 22 2.7 0.8 M2 control 10 5 3 0.4 - Calibration source 10 5 3 0.4 - Total load 266 147 93 11.5 0.8 Total load with 50% margin 399 220 140 17 1.3 Table 22: Heat loads due to harness The contribution of the AIRS cryo harness connecting the CFEE’s to the SCA’s (around 1 mW) is, at this stage, considered part of the detector system study budget and is already included in the active load estimation of the two FPA’s (see Table 13) assumed for the thermal analysis. In the next advanced phase of the PLM design, a detailed analysis will allow to trade thermal performance with the electrical properties, reliability, mass and complexity for harness design optimization.
7.5 THERMAL BUDGETS The ARIEL PLM loads, especially for what concerns detectors dissipation, mechanical supports, piping and harness leaks, have been evaluated on the basis of the present knowledge of the units performances and of the heritage from more advanced projects (MIRI, Planck, Euclid etc.). Conservative estimations have been assumed and verified by thermal analysis with the PLM G/TMM. The budget resulting at the external thermal interface and at the main six internal interfaces from the analysis activity is reported in Table 23 and Table 24. The values correspond to the loads calculated by the TMM in the two main reference cases: Cold and Hot. The fact that the heat fluxes reported in Table 24 correspond to interface temperatures lower than required, indicate that extra margin is available for optimizations in the next phase.
Page 81 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
External Req T Cold Case T Cold Case Load Hot Case T Hot Case Load Location IF to Interface [K] [K] [mW] [K] [mW] SVM Bipods and VG ≤ 290 270 -6040 290 -8720 Conducted struts TIF0 SVM VG1 (mainly) ≤ 290 180 -1430 200 -2460 Radiative Table 23: ARIEL external thermal interface budget
Internal Req T Cold Case Cold Case Hot Case Hot Case Location IF to Interface [K] T [K] Load [mW] T [K] Load [mW] Heat leaks from SVM (harness, TIF1 V-Groove 1 ≤ 200 149 7470 158 11180 bipods, struts, piping, radiation) Heat leaks from warmer stages TIF2 V-Groove 2 (harness, bipods, struts, piping, ≤ 120 97 1130 103 1400 radiation) Heat leaks from warmer stages TIF3 V-Groove 3 (harness, bipods, struts, piping, ≤ 70 55 980 57 1100 radiation) + JT precooling TIF4 TOB/Baffle Channels FEE, parasitics, rad ≤ 60 50 740 51 810 Instrument FGS detectors + T control stage TIF5 ≤ 60 49 100 50 100 Radiator loads + parasitic leaks AIRS detectors + T control stage TIF6 JT cold end loads + parasitics leaks < 40 35 37.5 36 37.5 (struts+harness+rad) Table 24: ARIEL internal thermal interfaces budget The thermal model results, heat fluxes across the interfaces and units temperature, indicate a reliable PLM thermal architecture, compliant to the requirement including the present level of margin. The next steps of the thermal analysis will be aimed to an optimized design and a more reliable assessment of uncertainties and margins at the interfaces.
7.5.1 Uncertainty Analysis The uncertainty analysis of the thermal study is summarized in the ARIEL PLM Thermal Analysis Report ([RD14]).
Page 82 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
8 ACTIVE COOLER SYSTEM
8.1 BASELINE COOLER ARCHITECTURE The active cooling system for ARIEL has been baselined as a closed cycle Joule-Thomson (JT) system with Neon as the working fluid. The cooler makes use of advanced compressors, heat exchangers and gas handling systems which have been developed by STFC’s RAL Technology Department. Cooler development at RAL has been ongoing for over 30years and has resulted in a range of mechanical cryocoolers covering temperatures from 2K to 80K and above. At the forefront of the current state of the art, these coolers have underpinned a large variety of high profile scientific and operational missions; ranging from space exploration and the origins of the universe, to earth observation, climate change and weather forecasting. Notably, the group provided the 4K-JT cooler for the Planck spacecraft. Key to success is the high reliability and lifetime of the cooler mechanisms, which stems from the philosophy of non-lubricated, non-contacting moving parts, enabled by a flexure bearing suspension system operating in the infinite fatigue strength regime; there have been no failures in space. The JT cooler system comprises several reciprocating linear motor compression stages to perform an expansion of the gas across a JT orifice to produce the cooling. Neon is selected as the working fluid because its boing point (27.05K at 1atm) is well matched to the temperature requirements for ARIEL. Note that the melting point is very close (24.55K) which places some restrictions on the temperature range over which cooling can be provided. The system makes use of several pre-cooling stages that are available from the spacecraft radiators. Counter-flow heat exchangers are used between these stages to reduce the heat rejected to the radiators. An ancillary panel carries gas handling and measuring equipment as well as filters and a reactive getter to ensure gas cleanliness. The cooler is controlled by a set of drive electronics which provide the electrical input power for the compressors, perform all controlling functions such as active vibration cancellation and return the cooler housekeeping data. The basic system layout with two stages of compression and a single JT cooling interface is shown below. Also shown is a set of disconnection plates and connecting pipework that allow the system to be broken into several pieces to aid integration.
JT T Temperature sensor F Res T P Pressure sensor 28K
F Filter F T 70K
F T Cooler Drive 180K Electronics Disconnection 300K Disconnection Flow P Getter F high pressure inlet T T P low pressure return
Compression stages Ancillary panel Connecting Pipework (two shown)
Figure 52: JT Cooler Schematic Page 83 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
8.1.1 Heat Exchanger Design The heat exchangers are pipe-in-pipe counter-flow types that make use of the cold low pressure return gas to remove enthalpy from the warm high pressure inlet gas in order to reduce the heat rejected to each of the pre-cooling stages. At the interfaces on each of these stages there are filters to ensure good thermal contact to the stage and also to trap impurities in the gas. The design is essentially identical to that used for the Planck 4K-JT cooler, and the RAL 2K-JT cooler (currently under development for ATHENA) but with the pipe lengths and diameters optimised to suit the operating conditions and system requirements for ARIEL. The pre-cooled gas is routed to the distributed elements on the focal plane that require cooling in a series of final JT interfaces which terminate in a filter prior to the JT orifice, through which the gas is expanded to produce liquid. The liquid is retained in a sintered reservoir at each of these interfaces to prevent sloshing and flash evaporation affecting the temperature stability. Parts of the heat exchanger system are shown below for the RAL 2K-JT cooler in its test configuration.
Stage filter and heat exchanger system Final stage filter, JT expansion orifice and liquid reservoir 15K/150K interface 2K interface 100g 125g 69mm x 35mm x 21mm 79mm x 61mm x 24mm
Figure 53: CAD Models of potential V-Groove (left) and Coldhead (right) heat exchangers
8.1.2 Compressor Set Design The compressors are linear motor reciprocating piston mechanisms, with the flexure bearings maintaining the dynamic clearance seal between piston and bore at 10um over the operating stroke. The motor efficiency is a trade-off with mass, which from a system perspective is usually targeted to be greater than 80% under nominal compressor operating conditions. A range of mechanism motor modules from several hundred grams to several kilograms are available that cover operation over a wide range of power requirements. The motor modules are identical for both JT and Stirling/Pulse tube type cryocoolers but a set
Page 84 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 of valves and buffer volumes are incorporated in the head of the JT compressors in order to produce a one way flow through the heat exchanger pipework. The compressors are arranged in balanced pairs, such that the exported vibrations may be cancelled. Each JT pair offers a two stage compression to give a maximum total compression ratio of around nine. For the Planck 4K-JT cooler a single pair was used to boost from 1.3bar to 10bar, and for the RAL 2K-JT cooler two pairs are used to boost from 0.15bar to 8bar. The individual mechanism modules are mass and spring balanced; the piston diameters being carefully selected to suit the operating conditions such that the rms force requirement for each motor is closely matched. In this way each mechanism operates at maximum efficiency at the common operating frequency, and this configuration also offers the best opportunity to minimise the exported vibrations under active vibration control.
Figure 54: (left) Small scale cooler motor mechanism, 165 g; (right) 2K cooler motor mechanism, 820 g
8.1.3 Gas Cleanliness and Ancillary Equipment Gas cleanliness is of paramount concern in a JT cooler because of the small size of the JT orifice (or the small size of flow paths, if a porous plug is used); any impurity that could condense at the cold end need only be present in very small quantities to cause a blockage that would either significantly reduce the cooler’s performance or block it completely. STFC’s RAL Technology Department has developed a reliable strategy for dealing with gas cleanliness based on many years of experience with these types of coolers. The following steps are taken: 1. The system is cleaned as much as possible prior to final fill using RGA analysis and contaminant limits with a bespoke fill rig as MGSE 2. An in-line getter is used for contaminants that can continuously evolve during life 3. Gas is thoroughly circulated through the closed system prior to cooldown 4. A decontamination heater is fitted as back-up
The in-line getter is part of the cooler’s permanent ancillary equipment and is located on the Ancillary Gas Panel described below. In current systems, a room temperature getter (as opposed to an actively heated one used for Planck) is used to simplify the design and reduce the overall power requirement of the cooler. The getter is required to absorb the contaminants present in the system at beginning of life, along with the predicted impurities that might evolve from the heat exchanger pipes over the mission lifetime and is usually considerably oversized. A previous 4K-JT system developed by the group has deliberately been grossly contaminated and has been shown to recover by operation of the getter alone. In addition to the getter, porous filters are fitted to each pre-cooling stage and these will trap impurities that freeze out at the temperatures of the stages. Should freezing of impurities at the orifice cause a blockage to occur, a small heater is fitted directly to the JT orifice to allow it to be warmed, releasing any impurity for removal at either the pre-cooling stage filters, or at the getter.
Page 85 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The getter is part of the Ancillary Gas Panel, which also contains the instrumentation that provides housekeeping data for monitoring the cooler’s health. Pressure sensors are fitted to both the high and low pressure lines to monitor the pressure ratio produced by the compressors. A flowmeter is fitted to the high pressure line to monitor the mass flow rate of gas. Temperature sensors are fitted at appropriate locations on both the Ancillary Panel and the compressors to monitor thermal performance of these sub-systems. The Ancillary Panel should be located directly adjacent to the compressors and there is also the possibility of making it integral with the compressor mounting bracketry.
8.1.4 Cooler Control Electronics The cooler is operated through a dedicated Control Electronics module. The Control Electronics provides all the power conditioning, control and monitoring to manage the driving of the compressors as well as monitoring and recording the cooler housekeeping data.
Figure 55: Typical cooler drive electronics for a two stage compressor set
The Control Electronics comprises the following main functions: • Command & Telemetry Interface • System Control; • Digital to Analogue Converters (DACs); • Current Control Loops; • Drive Amplifiers; • Sensor Signal Conditioning; • Sampled Data Acquisition; • Launch Lock; • EMC Filter; • Auxiliary power supply; • Active line filter; • Critical Parameter Determination; • Vibration Sampling rate; • Waveform Generation; • External Electrical interfaces.
Page 86 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The cooler drive part of the electronics generates the driving waveforms and amplifies them according to the required compressor stroke. The stroke is measured either directly using a position sensor, or indirectly from the compressor motor back-EMF, and this is adjusted to the demanded stroke through a control loop. Without filtering, the (approximately) sinusoidal driving currents required for the compressors would induce a large ripple on the current demanded from the spacecraft bus. The Control Electronics will therefore incorporate current regulation in the form of an active line filter to minimise the ripple seen by the spacecraft bus. This part of the electronics will be designed to meet the conducted emissions requirements of the spacecraft. A “launch lock” feature will be incorporated in the electronics to prevent excessive motion of the compressor pistons under launch vibrations. This can simply be achieved by shorting the compressor drive coils so that any motion of the motor coils induces a back-EMF which opposes their motion. This technique has been thoroughly qualified on previous coolers and was employed successfully on the Planck 4K-JT cooler compressors. If required, active vibration control can be incorporated in the Control Electronics; this is a control algorithm that nulls in-axis vibration of the compressors by measuring the exported vibration and correcting the driving waveforms such that it is minimised. In addition to driving the compressors, the Control Electronics will include everything necessary to power and read out the instrumentation on the cooler. This will incorporate auxiliary power supplies, analogue signal conditioning, data acquisition and telemetry compilation. A complete telemetry format can be sent immediately upon request or provided on a periodic basis, as determined by the mission requirements.
8.2 COOLER SYSTEMS ARCHITECTURE AND INTEGRATION TO SPACECRAFT One of the benefits of the JT cooler configuration is that the cooling is delivered remote from the compressor mechanisms which means that most of the cooler parts, ie almost all the mass and all those requiring electrical power (except for a few remote temperature sensors and a small heater) reside in the SVM part of the spacecraft at ambient temperatures. This greatly reduces the possibility of unwanted mechanical and electrical perturbations from the compressors being manifest at the detectors. In addition the heat exchanger pipework lends itself to easy manipulation which gives great flexibility when selecting a route through the spacecraft to the focal plane assembly. For ARIEL the pressure drop along the heat exchanger return line does not place constraints on its length. From the initial modelling results the total heat exchanger length is expected to be around 10m. The basic system layout is given in schematic form above. The cooler drive electronics, compressor mechanisms and ancillary panel should be located in close proximity in the SVM, in particular the compressors and ancillary panel should be integrated as a single item. A disadvantage of the distributed nature of the JT cooler system is that it is undesirable to integrate as a complete unit and disconnection plates are provided that allow the heat exchanger pipework to be integrated separately. The connecting pipework, which may be any reasonable length and shape, provides a useful method of routing through the SVM and connecting the cold and warm parts together after integration. The integration sequence may be envisaged as follows; 1. After full cooler acceptance tests under test configuration (expected to be with pre-cooling provided by a GM machine) the system is split into three sections; a. Heat exchanger pipework, terminating in a disconnection plate b. Compressors and ancillary panel, terminating in a disconnection plate c. Connecting pipework (this may be representative in test configuration) 2. Sections (a) and (b) are sealed shut and contain clean gas used under acceptance tests 3. The heat exchangers are manipulated into the final configuration to fit the spacecraft 4. The compressors and ancillary panel section is integrated to a panel of the SVM 5. The drive electronics are integrated to the SVM panel and electrical connections are made 6. The heat exchangers are integrated, routing through and interfacing with the v-groove radiators to the focal plane
Page 87 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
7. The SVM panel, with cooler attached, may be manoeuvred into final position as required and the connecting pipework then integrated to join the two other cooler sections together 8. The connecting pipework is pumped and purged to the correct cleanliness levels, filled to the required pressure with clean gas through a bespoke filling rig, and then both disconnection plate valves are opened to join all three sections back together in the flight configuration.
8.3 BASELINE COOLER THERMAL MODELLING AND PERFORMANCE PREDICTIONS The cooling power available is dependent upon the mass flow and the expansion ratio across the JT orifice; for ARIEL the requirement is around 50mW of JT cooling. There are very many solutions for a given criteria and the cooler compressors and heat exchanger configurations may be optimised accordingly. From a system thermal performance perspective, two considerations are very important: • heat rejection – the heat rejected at the thermal interface of each stage of the heat exchangers to the spacecraft radiators increases as the mass flow increases • temperature – the pressure in the heat exchanger return line determines the temperature of the JT thermal interface according to the vapour pressure curves for Ne
In general the JT heat exchanger will be longer at lower pressure ratios and higher mass flows for a given effectiveness (targeted to be around 98%). There is also some flexibility to distribute the total load amongst the stages by adjustment of the individual lengths. From a system budget perspective a reduction in mass flow requires an increase in pressure ratio to maintain the cooling power, leading to further considerations: • input power – the input power to the compressors is increased as more pV-work is required to provide the increased pressure ratio • mass and size – larger compressors are required to provide more p-V work and extra compressor stages may be required as the pressure ratio increases
An initial sizing exercise for the ARIEL JT cooler has been carried out to meet a cooling requirement of 50mW. JT interface temperatures of ~28K and ~32K have been considered, corresponding to heat exchanger return line pressures of ~1.5bar and ~3.5bar respectively.
The table below shows the mass flow and inlet pressure required to produce 50mW of cooling at 28K (Plow = 1.5bar). Although the heat exchangers have not been fully optimised for each scenario in the table the heat rejected for a mass flow around 25mg/s is expected to be around 100mW at 180K and 300mW at 70K with heat exchanger lengths of 2m and 3m respectively. For higher cooling power requirements, to first order, both the cooling power and heat rejected scale in direct proportion to the mass flow.
Phigh (bar) 9 11 13 15 17 19 m (mg/s) 29.8 23.6 19.6 16.7 14.6 13.0
To meet the flow and pressure conditions in the table above, several compressor arrangements can be considered; based on the RAL Small Scale Stirling Cooler (SSC) and the RAL 2K-JT cooler mechanisms shown above. Some initial modelling results are presented in the table below. An operating frequency of 60Hz has been selected which offers a reasonable compromise between swept rate and increased valve leakage at higher frequencies; this could be further optimised for each compressor arrangement. Note that the input power reported in the table is the total into the compressor mechanisms only and does not include cooler drive electronic inefficiencies. Similarly the mass reported is for the compressor mechanisms only.
Page 88 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
configuration 2 stage SSC 4 stage SSC 2 stage 2K 2 stage 2K stroke [max] (mm) 3.75 [4.25] 3.50 [4.25] 5.00 [6.5] 4.80 [6.5]
Phigh (bar) 9.4 17 9.7 16.7
Plow (bar) 1.4 1.5 1.5 1.5 m (mg/s) 30 15 30 15 fill pressure (bar) 4.5 8.0 4.5 7.0 piston sizes (mm) 14.5 / 9.5 13.25/10.25/9.5/8.5 12.5 / 9.0 14.0 / 8.5 spring rate (N/mm) 3.7 6.1 33 29 input power (W) 47 61 44 60 mass estimate (kg) 1.1 2.2 4.5 4.5
The simulations show that the requirements can be met at the upper end of the stroke and power limits of the small scale cooler capabilities. Going to a four stage compression with the SSC mechanism gives some margin on stroke, as would increasing the operating frequency, and this would also increase the required spring rate for resonance which is a little low as reported in the table. The SSC mechanism can operate well in excess of 100Hz although valve leakage limits the gain in stroke margin at these higher frequencies. The simulations also show that the requirements are comfortably met by the 2K-JT cooler mechanisms, and that reduction of the mass flow, and hence rejected heat loads, by operating at a higher pressure ratio incurs a relatively small compressor power consumption penalty. A further simulation of the two stage 2K configuration given in the table above, operating at the maximum stroke of 6.25mm [6.5mm end stop] and opening up the orifice to give a mass flow of 60mg/s, without making any other changes, results in a pressure step from 1.4bar to 8.6bar, which would produce around 100mW of cooling whilst requiring an input power of 61W into the compressors. Additional modelling has been carried out to consider providing 50mW of cooling at a higher temperature of 32K (Plow = 3.5bar). As above, the operating frequency considered is 60Hz.
configuration 2 stage SSC 4 stage SSC 2 stage 2K 2 stage 2K stroke [max] (mm) 3.60 [4.25] - 5.00 [6.5] 5.00 [6.5]
Phigh (bar) 20.5 - 20.5 14
Plow (bar) 3.5 - 3.5 3.5 m (mg/s) 10 - 10 18 fill pressure (bar) 10.5 - 10.5 8.0 piston sizes (mm) 10.0 / 7.5 - 8.5 / 6.5 8.0 / 6.25 spring rate (N/mm) 3.1 - 33 34 input power (W) 46 - 46 35 mass estimate (kg) 1.1 - 4.5 4.5
The table highlights several advantages of increasing the thermal interface to a higher temperature in that the cooling power requirement is still met by; • reduced mass flow – the heat rejected at the thermal interface of each stage of the heat exchangers may be reduced for the same compressor input power • reduced input power – the compressor input power may be reduced for the same thermal interface rejection load
8.3.1 Modelling Conclusions The models give a clear indication that the active cooling requirements for ARIEL can be met, and also show that the cooler operating parameter space to do so is extensive and has a great deal of flexibility with respect to system constraints.
Page 89 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The baseline solution for ARIEL is to use a two stage compressor pair, based on the RAL 2K-JT cooler mechanisms. Although these mechanisms are slightly oversized for this application, they offer significant margin on cooling power, whilst still satisfying the available mass and power budgets for ARIEL, and give maximum flexibility in the event of unforeseen system constraints. The baseline performance predictions are summarised in the table below for operation at 28K. A reasonable assumption for the drive electronic inefficiencies is to assume 85% for the power amplifiers and 85% for the active line filter (used to smooth current draw from the spacecraft bus) to give an overall efficiency of 72%; this has been used to calculate the total cooler input power in the table, which also includes an overall estimate for the cooler mass. Note that the mass includes the disconnection plates which are a large part of the heat exchanger mass and a baseplate that is a large part of the current ancillary panel mass, giving considerable scope for mass reduction as the design evolves. Further work should also focus on optimising the heat exchanger configuration, the operating frequency and the piston sizes to better match the system requirements as these evolve.
ARIEL two-stage Ne JT cooler baseline estimates (28K) baseline +10% margin +20% cooling +10% margin cooling power at 28K 50mW 60mW heat rejected at 180K 100mW 110mW 120mW 130mW heat rejected at 70K 300mW 330mW 360mW 400mW total input power 61W 67W 66W 73W Mass Compressors 4.5kg ancillary panel 3.5kg heat exchangers 2kg drive electronics 6.8kg Harnesses 1.2kg total mass 18kg 19.8kg 18kg 19.8kg
8.3.2 JT Cooler Development Activities Relevant to ARIEL STFC have been invited by ESA to submit a tender for the activity “Neon-Joule Thomson Cooler for ARIEL” which is currently under negotiation with an expected start date of April 2017. The activity will focus on development and verification of a heat exchanger system, using Neon gas as the working fluid, and with thermal interfaces according to the ARIEL requirements. In parallel the development of the ESA 2K-JT cooler TRP activity (ref 22634/09/NL/EM) is ongoing, with the cooler due to be delivered to the ATHENA 50mK cryochain CTP demonstrator (ref 4000117207/16/NL/HB) in autumn 2017. In addition STFC have submitted a proposal to the Agency for a “2K JT EM cooler system including cooler drive electronics” (ref AO/1-8778/16/NL/HB). This activity is intended to focus on raising the 2K-JT cooler mechanical assembly to TRL6 with a DM/EM electronics unit.
8.4 ALTERNATIVE COOLING CONFIGURATIONS There are a number of alternative active cooler technologies that may be feasible alternatives to the baseline outlined above. Although the programmatic constraints (primarily national funding priorities and capacity) restrict the choice (which is why no full trade-study has been carried out during Phase A), the alternative technologies for providing cooling are outlined here to demonstrate that there are sufficient back-up options in case these were required.
Page 90 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
8.4.1 40K Turbo-Brayton Cooler An alternative cooler solution that has been being developed under ESA TRP contract by Absolut System is a Turbo-Brayton design with Neon as the working fluid. This scheme would also use the V-Grooves to provide intermediate cooling in order to reduce the recuperator size and improve performance. In the frame of the ARIEL project a very preliminary sizing of a system has been conducted assuming Neon as the working fluid, a centrifugal compressor running at 250,000 rpm (in order to draw on similarity with on-going TRP contract), a turbine running at 150,000 rpm and a gas flow rate of ~0.5 g/s. In order to provide 65 mW of cooling at 40 K at the cold end, the sizing of the system needed (no margin) is shown in Table 25 below. The system architecture is shown in Figure 56.
Table 25: Calculated performance of Turbo-Brayton alternative cooler system
Figure 56: Schematic of Alternative Turbo-Brayton cooler system The system would consist of the compressor and drive electronics mounted within the SVM, the recuperator and expander mounted with the support struts for the PLM running from the 300K SVM to the ~55K PLM, and then cold links to connect the expander to the cold head on the detectors. The thermal links would have to be carefully designed to limit the parasitic leaks and the thermal gradient along the length. The accommodation of the system and examples of the constituent parts are shown in Figure 57 below. The Page 91 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 basic mass for the system is currently estimated as 10.5 kg and a power input of 45 W is thought to be needed (both numbers exclude margin).
Figure 57: Accommodation Scheme for Alternative Turbo-Brayton Cooler The baseline consortium configuration would not support funding of these coolers.
8.4.2 Pulse-Tube Coolers Ongoing developments of Pulse-Tube coolers at CEA may lead to a solution which could provide cooling at ~25K. The major drawbacks of such a system are the relatively large exported micro-vibrations, and the fact that the cold-head would have to be hard-mounted to the top floor of the S/C SVM. This would then require lengthy (~1m) thermal straps to connect the cold sink to the detectors. The parasitic loads on these straps would have to be carefully managed as they routed up through the V-Grooves, and they would have to be sufficiently flexible to prevent the transfer of micro-phonics from the cooler into the AIRS detector. The baseline consortium configuration would not support funding of these coolers.
8.4.3 Sorption Coolers A completely static (vibration free) solution for cooling would be the use of Sorption coolers, such as those under development at Twente. However, these coolers would require large additional heat loads on the V- Groove system (at least doubling the loads on the 3rd V-Groove radiator) which would adversely effect the temperature reached on the rest of the passively cooled payload module. These have therefore been discounted from consideration.
Page 92 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
9 ELECTRICAL SYSTEM DESIGN
9.1 OVERALL ELECTRICAL ARCHITECTURE The baseline payload architecture has three warm electronics units which interface directly to the spacecraft: • Instrument Control Unit (ICU): houses AIRS Detector Control Unit (DCU) and the Telescope Control Unit (TCU) plus incorporates a Power Supply Unit (PSU) and the Data Processing Unit (DPU). o Interfaces to the S/C with power supply (prime and redundant) and SpaceWire communications links (prime and redundant) for science data flow and commanding. • FGS Control Unit (FCU): contains the detector warm front end electronics for the FGS, plus the PSU and data processing unit for analysing and Centroiding the fine guidance data for feeding to the spacecraft AOCS system. o Interfaces to the S/C with the S/C with power supply (prime and redundant) and SpaceWire communications links (prime and redundant) for science data flow, AOCS centroid information and commanding. • Cooler Control Electronics (CCE): drives the cooler compressor, monitors cooler health and status and active control of both temperature and micro-vibration from cooler. o Interfaces to the S/C with power supply (prime and redundant) and communications links (prime and redundant) over MIL-STD-1553 (TBC) for HK data flow and commanding.
9.2 POWER BUDGET The expected power consumption of the ICU and the overall ARIEL payload is illustrated in Table 26. The reported values are well within the assigned nominal power budget including 20% of contingency (DMM). This power is estimated for a time averaged operational usage scenario, although fluctuations in this value are expected to be small during observational periods. The contamination control heater lines are not included in baseline operational power budget as well as the absorption peaks related to the M2M activation (refocusing/calibration mode).
Item CBE (W) DMM (W) MEV (W) MPV (W) Instrument Control Unit 35.4 7.1 42.5 47.0 FGS Control Unit 10.5 2.1 12.6 21.0 Cooler Electronics & 61.0 12.2 73.2 72.0 Compressors TOTAL ARIEL Payload Power 106.9 21.4 128.3 140.0 Table 26: ARIEL Payload Estimated Power Budget (CBE: Current Best Estimate; DMM: Design Maturity Margin; MEV: Maximum Expected Value; MPV: Allocation or Maximum Possible Value)
9.3 DATA RATE BUDGET The table included in this subsection displays a breakdown of the expected daily data volume. The overall data volume (25.0 Gibit/day) is dominated by the spectrometer channels, AIRS0, AIRS1 and NIR-Spec and takes into account the observing efficiency (95% - targets + calibration), as well as the required data volume margin (30% as defined by the ESA margins document). The calculation assumes on-board fitting of the ramp, with a daily average ramp length to saturation of 4 seconds for NIR-Spec and 3.26 seconds for the AIRS channels. The actual read-out mode to be used will vary between targets (dependant on their brightness) with the possibility of setting the ramp integration time in the range of 2 to 7 seconds. It is assumed that we sample up-the-ramp pixels in a non-destructive manner with a relative high sampling rate (3.5Hz for AIRS0, 9.3Hz for AIRS1 and 10Hz for FGS channels). There will be destructive readouts after some samples, depending on the brightness of the target, however the integration time for the ramps Page 93 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 downloaded, regardless of target brightness, will be 4 seconds. In case of very bright targets (saturation 0.1 sec) several CDS frames will be averaged together every 4 seconds, to form one frame for downloading. Example scenarios for the SciRD defined faintest and brightest targets are given:
Faint source (GJ 1214 [Mk = 8.5]): Saturation Time/Exposure Rate: 20 sec. Ramps will be fitted every 4 sec, resulting to 5 frames for downloading, before reset.
Mid Bright Source (HD 209458 [Mk = 6.5]): Saturation time/Exposure Rate: 4 sec. 1 ramp (4 sec long) will be fitted, resulting to one frame per reset.
Brightest Source (HD 219134 [Mk = 3.5]): Saturation time/Exposure Rate: 0.1 sec. In the duration of 0.1 sec, two samples can be used to measure slope per pixel (CDS). Every 4 sec, all samples will be averaged to form one frame.
Table 27 below, shows a breakdown of the daily data volume contributions from all channels. It should be noted that these calculations apply to all targets; all targets will have the same integration time per ramp, regardless of brightness. This concludes that with the necessary margin a data volume of 25.0 Gbit / day is produced.
Page 94 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Pixel Pixel Chan Bits per Prim. Rate Int. time per No. Bits / Total Gibits per
Spect. Spat. Total sample (Hz) ramp (sec) ramp Bits/sec day
Science Channels
FGS AOCS mode 4 21 10 840 0.07
FGS1 30 30 900 16 10 4 24 5400 0.43
FGS2 30 30 900 16 10 4 24 5400 0.43
Vis-Phot 30 30 900 16 10 4 24 5400 0.43
NIR-Spec 160 60 9600 16 10 4 24 57600 4.63
AIRS-CH0 270 64 17280 16 3.47 3.26 24 127215 10.24
AIRS-CH1 100 64 6400 16 9.35 3.26 24 47117 3.79
Total 35984 Total sci 248971 20.0
Obs. Eff 0.95 19.0
Total sci/day (Gibits) 19.0
Housekeeping Channels
Active Temps 16 16 2 512
Electronics etc 32 12 2 768
M2 actuators 8 16 0.5 64
Heaters 8 16 0.5 64
Monitoring Temps 46 16 1 736
Total HK 2144 0.2
Daily Total 19.2
Margin 0.3 5.8
Daily Data Volume 25.0
Table 27: Breakdown of daily data volume budget. These calculations apply to all targets, regardless of brightness. In the case of Brightest targets (saturation time 0.1 sec) several ramp fits (in the duration of the ramp integration time) are averaged together before downloading.
Page 95 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
10 INSTRUMENT CONTROL UNIT (ICU) DESIGN
10.1 AIRS & ICU ELECTRICAL ARCHITECTURE The AIRS end-to-end general electrical architecture relies on three main blocks: the detectors proximity electronics (mainly the FPAs with their ROICs) for CH0 and CH1 science channels, the Cold Front End Electronics (CFEE) and the Instrument Control Unit (ICU), hosting the Warm Front End Electronics (WFEE) as main driver and electrical I/F to the spectrometer cold side. The ARIEL ICU (for more details refer to the dedicated ICU Design Description MSR Technical Note) implements the commanding and control of the AIRS Spectrometer and is interfaced on one side with the instrument and on the other side with the SVM Data Management System (DMS) and Power Conditioning and Distribution Unit (PCDU). Referring to Figure 58, the analogue signals from the proximity electronics are pre-processed and A/D converted in the CFEE and the output is sent to the WFEE for data processing, storing and delivering to the on-board computer (OBC), part of the S/C DMS. The WFEE sits within the ICU, which also houses the Telescope Control Unit (TCU) as well as the housekeepings (HK), telemetry & telecommands (TM/TC) management functions hosted by the Data Processing Unit (DPU) along with the necessary Power Supply Unit (PSU).
Figure 58: Block diagram of the AIRS baseline end-to-end electrical architecture The telescope M1 mirror and the detectors temperatures will be monitored and fine tuned/stabilized by means of the thermal control electronics managed by the Telescope Control Unit (TCU). The feedback loop is closed thanks to an embedded logic (mainly a FPGA hosting PID controllers) interfacing thermistors and heaters, needed to fine-stabilize the focal plane arrays temperatures in order to constraint the detectors’ thermal noise. TCU is also in charge of other tasks, as herein described. The ICU internal memories (NVM, SDRAM and SRAM) are basically conceived and adopted for temporary local buffering and to support data processing as the scientific data, once properly processed, are directly sent to the S/C Solid State Mass Memory (SSMM) to be stored before sending them to Ground. It is worth noting that the ICU is hosted inside the Service Module, where all the warm Payload Units are located.
Page 96 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
10.2 INSTRUMENT CONTROL UNIT ELECTRONICS BASELINE DESIGN The baseline ICU architecture includes 5 (active or switched-on at the same time) units: • 1 PSU - Power Supply Unit (nominal and redundant 3U boards) • 1 DPU - Data Processing Unit (nominal and redundant 6U boards) • 2 DCU - Detector Control Unit (only nominal 6U boards) • 1 TCU - Telescope Control Unit (nominal and redundant 3U and 6U boards) as represented Figure 59 along with the number and type of needed PCBs (exploiting 3U and 6U formats). Indeed, as currently foreseen, TCU will host the main logic board (TSIRC, 6U), the M2M drivers (3U) and the needed power section and points of load (3U) to properly feed its subsystems. The nominal units are indicated with blue labels while the three redundant units are highlighted by means of red labels. They are hosted inside two stacked and independent boxes including PSUs, DPUs and DCUs (ICU box) and TCU logic, M2M drivers and needed PSU/PoL (TCU box). The two boxes are electrically connected (power and TM/TC) by means of external harnessing exploiting front panel connectors (TBD/TBC) to facilitate AIV/AIT activities as the Units will be integrated and tested separately. Both ICU and TCU boxes could implement their own back panel for routing power and signals lines connecting the internal electronics boards or, alternatively, exploit external connections, but the latter solution would limit the allocated volume for the Unit. A final assessment on both solutions will be performed during the next phase, taking into account the needed resources in terms of mass, volume, power dissipation and overall complexity. At the present time, in order to minimize the length and the mass of the harnessing connecting the two boxes, we assume a stacked configuration.
Figure 59: ARIEL ICU baseline solution block diagram and electrical I/F The ICU baseline electrical architecture is based on the baseline selection of US detectors (H1RG-type) and cold front-end electronics (CFEEs) from Teledyne (SIDECAR ASIC), given their very high TRL and space heritage with respect to the present European alternative. The SIDECAR solution is the best one to drive properly the US MCT (HgCdTe) detectors and to save mass, volume and power at the same time. They can work easily down to the ARIEL required cryogenic temperatures (≤ 60 K for SIDECARs and ≤ 42 K for detectors) so that both CH0 and CH1 are fed and controlled thanks to the adoption of two DCU boards, residing in the warm part of the S/C Service Module. The two electronics sides will be connected by means of cryo-harnessing, passing through the three V-grooves (working at different temperatures) of the Telescope assembly.
Page 97 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The present ICU architecture exploits a partial cold redundancy and cross-strapping capability. In particular, both TCUs and DCUs are cross strapped and can work along with PSU and DPU boards (Nominal and Redundant) as a whole, although DCUs aren’t involved in a cold redundant configuration as no duplicated DCUs are foreseen. A very similar ICU architecture, involving DCUs as SIDECAR I/F (for biases, clocks and control signals), has been already designed and adopted for the Euclid Mission (NISP Instrument). Each DCU controls and interfaces a single SIDECAR (as well as the related detector) and, in this sense, can be considered for the ARIEL Payload a strong heritage from the Euclid project. Indeed, for the Euclid design, an overall (DPU + DCU + SIDECAR + H2RG detector) chain/system reliability figure higher than 98%, has been computed and for this reason redundancy for ARIEL DCUs, as well as in Euclid/NISP case, has not been considered, also because the related increasing complexity and needed budgets (power, mass, volume). At the present time DCU Technology Readiness Level (TRL) has been demonstrated higher than 5 (likely > 6, following the updated ESA TRL definitions) and a DCU EM as been already manufactured and fully tested as it is working properly along with the SIDECAR ASIC and the detector. A DCU EQM model (very similar to the EM one) is under development and manufacturing. Moreover a DCU/SIDECAR I/F simulator has been developed and provided to the Euclid NISP Team. The same philosophy concerning simulator is adopted for the ARIEL case. As baseline, all the ICU and TCU boards (N and R) are designed respecting both the 3U (160 mm x 100 mm) and 6U formats (230 mm x 160 mm) and stiffened by a proper mechanical frame with the external I/O connectors fixed and screwed to the board external panels. Indeed TCU logic sections (TSIRC, 6U format, N & R) will be internally interfaced with a 3U board (N & R) hosting the M2M drivers and a 3U board (N & R) hosting the power supply and points of load required to feed the TCU implemented logic and M2M driver electronics. In particular, ICU box shall host: - two (N and R) 3U format PSU boards, providing +5V to ICU's DPU and DCU boards and +28V filtered (N and R lines) to TCU box; - two (N and R) 6U format DPU boards; - two (CH0 and CH1) not-redundant 6U format DCU boards; and TCU box: - two (N and R) 3U format PSU/PoL boards, locally deriving (thanks to on-board DC/DCs and PoL) ±5V and the needed voltage levels (+20V, ±12V) from +28V filtered coming from ICU; - two (N and R) 6U format TCU boards (control FPGA, IR calibration source drivers, etc.); - two (N and R) 3U format M2M drivers boards (or a single 6U format board hosting N and R drivers); for a total of nine (9) “equivalent” (from the point of view of overall dimensions and volume allocation) 6U- format boards, fitting the ESA allocated budgets. The lateral sides of the 3U and 6U modules will be equipped with card-lock retainers, used to fix them to the unit internal frame. All the boxes panels will be manufactured in a TBD Aluminium alloy and then externally painted in black (TBC, except the bottom panel) to improve radiating exchange with the environment and assure, at the same time, a proper thermal conduction towards the SVM mounting plane. PSU - It is a standard Power Supply Unit board hosting DC/DC converters with a number of secondary sections needed to support the adopted cross-strapped and partially redundant configuration. It is in charge of collecting currents, voltages on secondary outputs and temperatures HK (A/D converted internally to the Unit, exploiting the SPI HK I/F for signals and control lines to/from the ADCs). The Unit consumption monitoring is in charge of platform as well as its switching on/off (both PSU and DPU boards thanks to a sequencing logic owning to PSU), by means of HPC commands. The PSU is mainly composed of three sections: 1. Power conditioning, hosting EMI filters and DC/DC converters; 2. Power distribution, hosting Output Power Controllers (OPC); 3. HK sampling and A/D conversion, thanks to three (TBC) 12-bits ADC. Page 98 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Each electronic board, apart TCU unit requiring +28V, is basically supplied by a main voltage level of +5V protected for overvoltage and overcurrent and locally on-board (DPU and DCU) lower voltages are derived, by means of a Point of Load (PoL), the secondary voltage levels needed by the hosted electronic components (e.g., memories, FPGA core etc.). DPU - The Data Processing Unit board hosts as baseline a CPU (the UT699E processor from Cobham Gaisler), a co-processing FPGA, memories for booting (PROM), to host the application software ASW (E2PROM and/or NVM e.g. MRAM), for data buffering (SDRAM) and to support data processing (SRAM, SDRAM) as well. The selected LEON3FT CPU is a SPARC V8 microprocessor running @ 66 MHz or @ 100 MHz, for the E version, allowing up to 140 DMIPS. The processor will run the ASW on the RTEMS OS. One of the main characteristics of the UT699E CPU is the on-board availability of 4 embedded SpW links (2 supporting the RMAP protocol) allowing to be directly interfaced to the SVM and to the DCU SpW I/F. RMAP protocol could be exploited (TBD) to read and write directly on the DCU FPGA registers. DPU, as baseline, is in charge of data processing and compression if necessary (e.g. adopting the RICE algorithm, providing a lossless compression rate CR=2 at least). It receives 16/24 bits pre-processed data (pixels/spaxels and/or ramps) from the SIDECAR ASICs and DCU and, once processed it packetizes and send them in CCSDS protocol format towards the S/C DMS (Data Management Subsystem) for storing and later downloading to Ground. Pre-processing and compression tasks can be disabled in case of raw data request from the Spacecraft/Ground. DCU - The potential baseline Detector Control Unit board design is an heritage of the design adopted for the detector control units of the NISP instrument on-board the Euclid Mission, where the same kind of detectors have been used along with the same CFEEs (SIDECARs).. The baseline DCU hosts, a FLASH-based reprogrammable FPGA to offer maximum flexibility also in case of late requirements specification (or modification) from the ARIEL Science Team. The selected FPGA is a Microsemi ProASIC3-type device offering the capability to embed a HDL FSM with some programmable Science data pre-processing tasks (e.g. pixels co-adding, spaxel pre-processing etc., ramp slopes computing etc.) by means of a flexible parameters configuration that can be reprogrammed up to the EQM/FM unit (TBC). The FPGA also hosts a SDRAM memory controller to manage 128 MB (TBC) of on-board memory used as a buffer to support the HDL-based pre-processing tasks. The DCU WFEE is in charge of the SIDECAR clocking (at least a master clock is needed for the ASIC) and feeding (secondary finely regulated voltages produced by an on-board Point of Load (PoL), and it collects digitized scientific data and HKs (currents, voltages and temperatures) describing the ASIC status. The needed enabling and control signals for SIDECAR management are fully described in the ICU DD TN [RD24]. Three different grounding references (analog and digital) are foreseen for a clean power supply SIDECAR feeding. The SIDECAR Science I/F is based on an 8 bits LVDS parallel I/F (with data buffering and packets CRC) and a TM/TC I/F running @ 2 Mbps (serial syncro) + master clock line @ 10 MHz. The DPU I/F shall be based instead, on SpW for Science data TM along with a RS485 serial I/F offering the capability to manage the DCU FPGA from DPU. Alternatively, the FPGA registers could be managed thanks to the RMAP protocol running on SpW. TCU – TCU, being hosted by a dedicated box to facilitate its own AIV/AIT activities, is described in more detail in the following paragraph.
10.3 TELESCOPE CONTROL UNIT DESIGN BASELINE As baseline, the Telescope Control Unit should be able to accomplish the following tasks in ARIEL Science mode:
● Drive the M2 refocusing mechanism ● Drive the on board calibration lamps ● Monitor the thermal state of several PLM elements
Page 99 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
● Control thermal stability of the TCS for the following PLM subsystems:
o AIRS detectors o NIR detectors o FGS detectors o M1 mirror
The Telescope Control Unit (housed with the ICU) will be composed of three (3) separated boards, as the volume of the electronics to fulfil the requirements is bigger than a standard 6U board and also to ease AIT/AIV activities in the future since the boards can operate independently with only a power supply (different for each unit) and their respective data buses. As seen in Figure 60, a 6U PCB (TSIRC) will hosts the PLM thermal monitoring and control HW, the IR calibration lamp driver and their multiplexing stages. For the driver electronics of M2 mechanism, it is foreseen an upgraded version of Euclid’s M2M, with the same driver, which will require a separated 6U board for both nominal and redundant systems (M2MD). In order to reduce M2MD modifications to fit ARIEL requirements, as well as to reduce the number of I/F from ICU’s PSU, a dedicated 3U board PSU is foreseen (TCU-PSU), which will generate (from the main power line of +28V coming from ICU) all the voltage levels required by the M2MD as well as TSIRC boards. The system will be based on a cold redundancy, with all boards housed inside a dedicated box on top of ICU’s. The digital system of TCU will consist of a FPGA with a HDL FSM, to control all TCU boards as a slave system of ICU, in order to simplify the SW architecture of ICU’s DPU. The UT6325 FPGA will be hosted in TSIRC board as well as its PoL converters to generate the proper voltages for GPIO interfaces and internal cores. The FPGA will host two Digital Signal Processing Modules (DSPM, one for the thermal monitoring subsystem and the other one for the calibration lamp driver), five PID controllers, GPIO interface management to generate multiplexers addresses and select the proper voltage and gain for a given thermistor, as well as to control OPCs of TCU-PSU. It will also include a memory bank, two I2C (or SpW, selected in the next B1 phase of the mission) links to communicate with DPU and one MIL-STD-1553 (or SpW as well) link to communicate with M2M Driver. The telescope thermal monitoring will be performed by means of two types of sensors: Cernox thermistors for precise readings (detectors, M1, optical elements, etc.) and DT-670 diodes for housekeeping TM of other elements of the PLM (V-grooves, OB, baffle, etc.). These sensors will be driven and read thanks to the Thermal Stabilizer & IR Calibrator (TSIRC) board electronics. Sensors will be driven by an adjustable current source (one for each type of sensor, with 4 to 6 selectable levels) and thanks to their 4-wire configuration, read by means of an instrumental amplifier and a 16-bit ADC. All 46 PLM sensors will be sequentially powered and read with a multiplexing stage with no cross-strapping: nominal sensors will be connected to nominal TSIRC and backup ones to redundant TSIRC. Thermal perturbations in the PLM are expected have time periods much longer than 1 second, therefore, the system will be designed so that each second, all Cernox thermistors (approx. 60%) plus 4% of diodes will be read so the temperature controller can be updated with the proper feedback, and every 10 seconds housekeeping TM can be generated and sent to the DPU.
Page 100 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 60: TCU block diagram and electrical I/F
The thermal control of the TCS subsystems (which are placed between critical detectors/mirrors and their thermal sinks) will be carried out by monitoring their temperature, as explained previously, and activating their heaters once the correction has been calculated by FPGA logic (a PID control type loop). The heaters power will be supplied by its driver stage, which consists of a DAC and a buffer carefully designed to supply a constant current for detector heaters (avoiding EMI with the detectors) as well as a buffered PWM signal for M1. Detectors will have a single (redundant) heater to stabilize its temperature, but M1 requires several (3 to 5) to help distributing the heat. Survival heaters and thermistors might be installed in each TCS as well, but are assumed to be completely in charge of S/C. The IR calibration lamp will be based on a thermal source to generate the proper light spectrum for the detectors. Thermal source consists of a 4-wire tungsten filament in order to power and read its voltage at the same time. ARIEL requirements foresee a 24-bit DAC to control the filament current as well as a 24-bit DAC to read its tension. Due to the fact that space qualified DAC of 24 bits are not easily available (low TRL) an alternative solution has been designed, although Texas Instruments has recently released a space rated 24-bit ADC which we could implement. The alternative solution will be based on the same architecture as the thermal monitoring: a controlled current source (overclocked 16-bit DAC with a current buffer) to power the lamp, an instrumental amplifier to read its differential tension, an oversampled 16-bit DAC (to achieve 24-bit resolution), and the control logic inside the FPGA (a PID controller with a Delta-Sigma modulator to reduce bits count at a higher rate). Frequency increase (~65MHz) is expected to be achievable since thermal response of tungsten filaments is higher than few milliseconds and the UT6325 FPGA can operate at 120 MHz. The M2M Driver and mechanism are based on an inherited design from Euclid’s and GAIA’s M2MM. The 4.45 (TBC) kg mechanism will have 3 DOF (Tip/Tilt and piston) controlled by a dedicated driver hosted inside TCU box. For details of the design of the M2M please refer to section 11.1 and [RD25]. The system will rely on a single 6U board with a cold redundancy configuration, where nominal coils of the stepper motors are connected to the nominal section of the driver, and backup coils are connected to the redundant section. The maximum power consumption is expected to be approximately 10W. Drivers and mechanism voltages (±5V, ±12V and +20V) will be supplied by a 3U board PSU included in the TCU box (TCU-PSU). Communication with the driver board is achieved by a MIL-STD-1553 bus (or SpW), where each section is
Page 101 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 provided with a logic capable of decoding all telecommands received in order to generate the switching sequences required by the motors, and encoding the status information to provide serial telemetry.
10.4 COLD FRONT END ELECTRONICS SYSTEM DESIGN The ICU baseline architecture is driven by the adoption of US detectors and CFEEs from Teledyne, given their very high TRL and space heritage with respect to the present European alternative. Nevertheless, a possible ICU alternative architecture (refer to the ICU Detailed Design document, [RD24]) is based on the adoption of two European detectors (CH0 and CH1 channels) developed within a joint collaboration between CEA-LETI and Sofradir (ROIC development) and characterised by different pixel architectures and readout modes. See section 12.3.3.2 for details. The two pixel alternatives are SFD (Source Follower per Detector) and CTIA (Capacitance Trans Impedance Amplifier), respectively requiring the following CFEEs (presently in charge of SRON, The Netherlands): • In the SFD case, an ASIC would be required; • In the CTIA case the baseline choice would be to adopt an amplification and A/D conversion stage located on an intermediate 110 K thermal interface. The preferred option relies on the adoption of a CTIA-based pixel readout architecture plus a 110 K readout electronics stage (with performance expected to be in the 30 to 50 e- rms as readout noise) for the European detectors CFEE, as shown in Figure 61. In this configuration, the AIRS architectural blocks of the electronics design are the following: • Detector Sensor Chip Assembly + Read Out Integrated Circuit (AIRS FPA) • Cold Front End Electronics (Payload side) • Warm Front End Electronics (DCU on SVM side) • Instrument Control Unit (and TCU unit on SVM side)
Page 102 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 61 CFEE alternative solution simplified block diagrams. In the lower figure the numbers give the order in which the functions could possibly be transferred to the WFEE depending on the EMC analysis results. Provided that in case of adoption of European detectors and ASICs/CFEEs they shall have the required TRL before freezing the adopted technology for the electronics design (2019 - TBC), the justification of this alternative architecture is based on the following items: • The proposed electronics chain, as stated in the CEA-LETI design description document for the ESA MCR and its update for MSR, ensures that critical functions, especially the management of electrical interfaces (power supply, polarisation, grounding) to the detector is internal to AIRS. This is the main key technical driver for this implementation based on return of experience which lessons learned are that electrical interface (especially clean power supply and polarization, adequate EMC filter, optimized grounding scheme) are integrally part of the detection chain performance. In order to ensure capability for AIRS to meet performance requirements on noise levels, the full contributors shall be under AIRS responsibility. • Given the optional approach for the detectors procurement (US or EU detectors + CFEE), this architecture ensures that the selection process can be an AIRS internal process with controlled impact on external interface. This is a major system and programmatic element for supporting the proposed architecture. • From design point of view, this architecture -as interface with the proposed ICU- allows correct coupling of non-redundant detectors with redundant ICU subsystems while preventing fault propagation from the ICU to the detectors. This is also proposing adequate power supply handling and grounding for detectors lines.
Page 103 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
10.5 ICU ELECTRICAL INTERFACES AND BUDGETS
10.5.1 ICU Interface Definitions Three different sets of harnesses are foreseen: internal to the ICU and TCU subsystems (internal harnessing), cryoharnesses towards the Payload module and towards the S/C (or external harnessing for the SVM-Payload linking). The SIDECAR ASIC – DCU electrical I/F have been already explicated in the DCU board paragraph as well as the adopted kind of harnessing, while the main electrical I/Fs internally connecting the Unit subsystems are the following (refer to Figure 59): • DPU Power I/F: +5V (from PSU to DPU) • DCU Power I/F: +5V (from PSU to DCU-0 and DCU-1) • TCU Power I/F: +28V (from PSU to TCU N and R PSU boards)
• DPU TM/TC I/F: GPIO and SPI (from DPU to PSU) • DCU TM/TC I/F: RS485, SPI and SpW (from DPU to DCU-0 and DCU-1) • TCU TM/TC I/F: I2C, SPI or SpW (from DPU to TCU N and R TSIRC boards)
In case a back panel were adopted for signal and power lines routing inside the ICU and/or TCU boxes no flying harness would be foreseen between their own boards. Note that the ICU-TCU harnessing will be implemented by means of “external” (i.e. outside the ICU and TCU boxes) connections. This choice would facilitate the AIV and AIT activities. Concerning the selected (electrically and thermally) conductive materials, the harnessing towards the Payload (detectors, telescope mirrors and mechanisms) has to be carefully considered as it plays a fundamental role in thermal linking the warm electronics side to the cold electronics part. An initial assessment of the electrical and thermal design of these harnesses has been made in phase A, the conclusions are summarised in section 6.1 and given in detail in [RD26]). The foreseen harnessing towards the S/C is hereunder itemized. Nominal Power Supply and control IFs (the same are adopted for the Redundant I/F): • Power line: +28V + RTN (From S/C PCDU to ICU PSU) • Switch On HPC (Signal + RTN) (From S/C DMS to ICU PSU) • Switch Off HPC (Signal + RTN) (From S/C DMS to ICU PSU) • Switch Status BSM (Signal + RTN) (From S/C DMS to ICU PSU) • Sync (Signal + RTN) (From S/C DMS to ICU DPU) – TBD/TBC Nominal I/O digital TM/TC IFs (the same are adopted for the Redundant I/F): • 1 or 2 (baseline 1 TM + 1 TC) Standard Spacewire link(s) (configured @ 10 Mbit/s – TBC) (From S/C DMS to the ICU DPU). The MIL1553 BUS use is not foreseen at this stage of the design, except for the M2M driver TM/TC towards TSIRC board. Concerning the required harnessing and I/F connecting TCU and the Payload (telescope mirrors, mechanisms, calibration lamp), refer to the TCU Design Description [RD27].
10.6 ICU MECHANICAL DESIGN The ICU will be built in Aluminium alloy, with a stacked configuration. The TCU is in a separate housing to facilitate ease of AIV, but is mechanically coupled to the ICU such that only a single interface to the S/C is required. The design is shown in detail in the ICU MICD [RD28] and is illustrate below.
Page 104 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 62: ICU and TCU, 3D view of the overall assembly with rendering
Page 105 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 63: ICU and TCU, 3D view of the boxes wireframe The ICU actual mass estimation including TCU is 12.65 kg including 20% contingency. The CFEE mass budget is included in the Spectrometer mass budget.
10.7 ELECTRICAL GROUND SUPPORT EQUIPMENT The ICU Short-functional, Full-functional and Performance tests on the overall Unit assembly shall be performed using an appropriate Electrical Ground Support Equipment (EGSE). The EGSE shall support both the ICU test/verification and the AIRS Spectrometer end-to-end test at different stages of the AIV flow, e.g.:
• ICU Integration and testing (using additional test equipment to simulate the instrument I/Fs); • Integration and testing of the WFEE (DCU) + CFEE + FPA (AIRS-level); • Overall ARIEL payload Integration and Verification (TCU functionalities and telescope monitoring included).
At the first stage, the EGSE will include additional HW and SW Test Equipment (TE) simulating the relevant I/F and functionalities of the missing payload segments. These additional TE can be removed from the EGSE as soon as the related units can be added to the test configuration.
Page 106 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The ICU subsystems simulators and EGSE are needed to perform the preliminary tests (Short Functional Tests, Full Functional Tests) on the ICU Engineering Model (EM) and, later on the FM/PFM Model (Performance Tests). The chosen baseline model philosophy for the ICU subsystem is the Proto Flight approach, as fully described in the ICU detailed design document, [RD24].
10.7.1 Overview The aim of the ICU EGSE is to support testing and operations on both the ICU and the AIRS Spectrometer. A block diagram of the needed ICU subsystems simulators and EGSE is illustrated in Figure 64. The represented scheme fits both the baseline and alternative solution design.
Figure 64: ARIEL ICU simulators and EGSE; the green arrows represent the electrical I/Fs The main function performed by the ICU EGSE may be summarized as follows:
• TM/TC S/C interface exercising; • Storage of scientific data (S/C science data interface to mass memory simulator); • Power generation; • Payload instrument simulation with possibility to simulate instrument HK data; • ROICs/ASICs/CFEEs simulation with possibility to simulate Scientific data; • Data packets acquisition for storage and following monitoring & processing.
The Instrument Control Unit needed GSE (Electrical and SW GSE) is hereunder itemised: • ICU EGSE: • Workstations (desktops and/or laptops PC) • SCOS-2000/SCOE (ESA provided) • Harnessing (to the S/C and to the Payload) • Software (S/C and ICU subsystems simulators) • Instrument Database (IDB hosting TCs, TMs and related parameters) • Instrument Databank/Datapool
10.7.2 Functional description The ICU EGSE shall allow the test operator or Test Conductor to:
Page 107 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• Fully control the ICU via a unique control station; • Command & monitor the ICU via the adopted S/C communication protocol (SpW/CCSDS/RMAP) through a DMS simulator (SIS); • Generate primary power bus and interface all HK lines; • Send simulated scientific data via ROICs/ASICs/CFEEs Data & HKs Simulators; • Acquire the ICU scientific packets via the Data Acquisition link (SSMM/OBC or DMS simulator); • Compare the acquired data versus the simulated data to check for ICU correct data handling.
The best candidate for the SW platform to be adopted for the EGSE is the standard ESA SCOS-2000 (Spacecraft Control & Operations System). SCOS-2000 is the generic mission control system software of ESA, which supports CCSDS TM and TC packet standards and the ESA Packet Utilization Standard (PUS). It has been proven by recent ESA missions (e.g. Herschel/Planck) that, using SCOS-2000 and its add-ons, it can be extended to cover the on-ground testing phase and work as a proper EGSE. The use of SCOS-2000 will guarantee a very smooth transition from the ICU subsystem AIV to the ARIEL Payload AIV and the in-orbit operation phases (assuming that SCOS-2000 will be adopted by ESA for the EGSE and the Ground Segment, respectively). In particular, this smooth transition will concern the SCOS- 2000 instrument Data Base (MIB tables), which describes the TM and TC packets structure.
10.7.3 The S/C DMS (OBC and SSMM) and PDCU simulator (SIS) The S/C Interface Simulator (SIS, expected to be provided by ESA/Prime) interfaces the unit under test (ICU) by means of the S/C TM/TC SpW links and power lines, acquiring and recording data flows and forwarding them to the PI EGSE for further processing (e.g. thanks to an Ethernet connection towards an EGSE computer) and data archive. The SIS simulator shall be a full-function test and simulation module for the S/C TM/TC interfaces (SpaceWire CCSDS and/or RMAP protocols) and for Power I/Fs (+28 V) as well, in charge to adequately feed the ICU PSU and to allow simulation of missing electrical and power interfaces between the instrument and the ARIEL SVM. It shall have SVM DMS representative characteristics (e.g. Latching Current Limiters -LCL- class), including redundant I/Fs. It is expected to receive from ESA or from the S/C Prime Contractor at least 4 SIS: • one to be used by the Institute in charge of ASW development (INAF/IAPS) • on to be used by the industrial prim in charge of the ICU design and manufacturing • one to be delivered to the AIRS team for AIV and AIT activities at AIRS level. • one to be delivered to the ARIEL Consortium for AIV and AIT activities at system level.
Additionally the harnessing between SVM and the ICU is expected to be provided by ESA/Prime, both for the Engineering Model(s) and for the Proto Flight Model. This applies also to the harnesses for the SIS.
10.7.4 Payload Simulator A plan for an integrated system simulator for ICU, AIRS warm and cold FEEs, telescope mirrors temperatures + mechanism HKs and S/C (refer to Figure 64) is under assessment by the ARIEL Consortium in order to allow for independent Research Institutes/Team studies and development works. The Spectrometer simulator is in charge of simulating the behaviour of the AIRS instrument interfaces towards the WFEEs/ICU. It shall be possible to log the traffic on the links, to configure/monitor the HK production and to set up a set of procedures able to adequately react to commands provided by the ICU. This is to allow the execution of ICU acceptance (SFT, FFT, PT) and end-to-end tests in an integrated environment reproducing the real one.
10.7.5 Detectors and ROIC Simulators The detectors and ROIC simulators, as well as ASICs and CFEEs simulators, are in charge of commands acquisition/reception and to send simulated scientific and HK data to the ICU through the science data links towards CFEEs and WFEEs/DCUs, simulating the actual behaviour of the detectors readout electronics and the adopted data sampling procedures (i.e. non destructive readout and sampling up-the-ramp). Page 108 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
10.8 ARIEL ICU ON-BOARD SOFTWARE The ARIEL ICU On Board Software (OBS) will be composed by the following three main components: 1. Basic Software: - Boot software: it is installed on the PROMs of the ICU DPU board and allows loading the ICU Application Software. It contains all the low level drivers for the CPU board and its related interfaces. - Basic I/O SW, Service SW & Peripheral Drivers: it is a HW-dependent Software including the Software Drivers for all the internal and external ICU digital interfaces. This SW is used by the Application SW and can depend on the selected operating system. 2. Application Software: - Instrument Control & Configuration Software: it implements the ARIEL scientific payload handling. It controls the spectrometer, implements the operating modes, monitors the instrument health and runs FDIR procedures. It implements the interface layer between the S/C and the instrument. - Data Processing and Compression Software: it implements all the necessary on-board processing functionalities, included the (if implemented) on-board lossless compression. After the processing the SW prepares CCSDS packets for the transmission to the S/C Mass Memory (SSMM). 3. Real-Time Operating System: the selected baseline operating system is RTEMS. The role and the interconnections between the three listed components can be clearly identified in the layered representation reported in Figure 65.
10.8.1 ICU on-board software description With reference to the following figure, the physical layer includes all ICU HW components with a direct level of interaction with the on board software. The Runtime environment includes the Real Time Operating System (RTOS) layer, necessary to provide multi-tasking support. In case the baseline architecture based on the LEON processor will be confirmed, the RTEMS operating system is a good RTOS candidate, being already used for applications on board ESA satellites. The other indicated system services are those not directly provided with the OS kernel, but included in the Basic Software Component mentioned above. An OS abstraction layer has then been included in the layered structure for the ARIEL OBS, in which all middleware libraries have been considered. The middleware services are based on the use of RTOS function calls. They include all library functions dedicated to the low level handling of the ICU HW devices/interfaces. All the middleware libraries will be developed in house and will provide a mean for developing an Application Software virtually independent from the HW and OS below it. This layer is very important and will ease the testing activities. The Application Layer includes both the ICU Instrument Control software and the Data Processing software. The ICU Instrument Control SW will implement the TM/TC S/C interface handling, the payload housekeeping data acquisition and monitoring, the instruments operating modes management and the autonomous function execution. The software will be written in C, though some functions may need to be coded in assembly to optimise their performance.
Page 109 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Data Ramp fitting Data Data Data reordering functions deglitching Compression packing
TC TM HK Memory File OBP FDIR Handler handler monitor handler mngmt interpreter handler
Application Layer Boot Software Time mgmt TM/TC SpW Data SpW Tasks handl. 1553 I/F Library I/F library I/F library libraries library
Middleware – OS Abstraction Layer
ISR Timer Non-Volatile Serial I/O SpW 1553 driver driver driver driver driver
Operating System
Operating system services Runtime Environment
CPU Volatile Non-Volatile processor Timers memory memory Serial I/O Data bus 1553 I/F SpW I/F
Physical Layer
Figure 65 - SW layers structure of the ARIEL On Board Software
In case stringent timing requirements have to be met for subsystem commanding, an interrupt-driven command sequencer (On Board procedures, OBP interpreter) can be included into the ICU on board software. Based on the experience of Herschel’s HIFI and SPIRE instrument control software, this is a flexible and effective solution to implement time-critical commanding procedures. The Data Processing SW implements all the necessary on board processing functionalities, included (if implemented) the on-board lossless compression (i.e. RICE). After the processing the SW prepares CCSDS packets for the transmission to the S/C Solid State Mass Memory.
10.8.2 Data processing It is desirable to minimise the level of on-board data processing taking place – the more processing is possible to be done on the ground gives the maximum flexibility in the algorithms and the chance to improve the processing during the mission to allow the science team the optimum chance to extract the best SNR from the available data. The actual processing will be finalized during Phase B study exploiting simulated data flows to verify the effectiveness of the adopted data reduction steps for the selected detectors. In particular the deglitching algorithm performances shall be verified e.g. against the expected data redundancy (spectra overlapping), the data acquisition rate and the spaxels dimensions as well. Finally, the need to implement an on-board effective lossless compression is strictly related to the results of the on-board deglitching algorithm. If required, a dedicated trade off activity to evaluate the performances of different standard lossless compression algorithms on the on-board CPU processor shall be planned.
Page 110 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
11 ARIEL TELESCOPE SYSTEM DESIGN
11.1 TELESCOPE ARCHITECTURE TRADE-OFF During the EChO phase A study, an extensive trade-off of telescope configuration was undertaken jointly by ESA, the industrial primes and the instrument consortia. The telescope for ARIEL has broadly similar requirements (passively cooled to <70K, diffraction limit at ~3 microns, collecting area >0.6 m2 (was ~0.9 m2 for EChO), and the accommodation constraints from the S/C are thought to be similar. In preparing the proposal for ARIEL for M4 we scaled the telescope design from EChO and used the same horizontal launch, V-Groove architecture that was determined to be the optimal configuration for EChO, see [RD16]. During the ARIEL phase A there has been no fundamental changes to the design requirements on the telescope or to the accommodation constraints from the S/C. Therefore it is assumed that the conclusions of the architecture trade-off conducted for EChO hold true and the horizontal launch, off-axis telescope is assumed as the baseline.
11.2 TELESCOPE MATERIAL TRADE-OFF A trade-off of the material to be used for the telescope mirrors and structures has been carried out during the first part of phase A. This trade is documented in detail in ARIEL-INAF-PL-TN-004 [RD6]. The conclusion, as shown in Table 28 below, is that for the consortium provision of the telescope the optimum solution is a telescope with mirror and structures made from Aluminium 6061 T651 alloy. The demonstration of the feasibility of this solution has been undertaken through phase A and is covered in section 11.7. We note that the assessment and assumptions that go into the trade-off are in some part specific to the consortium organisation and industrial capacity of the consortium partners involved. Therefore independent assessment of the same trade by others may reach a different conclusion due to the programmatic constraints.
Consortium Thermo- Machinability Thermo- Specific Ability ability to Material Availability mechanical for Light- optical Stiffness Score to polish produce Design weighting stability (/strength) the mirror Al Difficult Good Bad Good Acceptable Acceptable Good 93 (RSA443) Al 6061 Good Good V. Good V. Good Good Bad Good 119 Zerodur Good V. Good Difficult Good V. Bad Acceptable V. Good 102 SiC Difficult Difficult Difficult Bad V. Good V. Good Good 96 Be V. Bad Bad Good Difficult V. Good V. Good Bad 77 Table 28: Conclusion of Telescope Material Trade-Off.
11.3 TELESCOPE BASELINE DESIGN The baseline telescope design is an afocal unobscured off-axis Cassegrain telescope (M1 and M2) with a recollimating off-axis parabolic tertiary mirror (M3) [RD5]. The system aperture stop is located at the M1 surface. M1 aperture is an ellipse with major axis dimensions of 1100 mm x 730 mm. M4 is a folding mirror. See Figure 66 for the telescope layout.
Page 111 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 66: Scale drawing of the telescope – view in Y-Z plane
11.4 TELESCOPE SCIENTIFIC REQUIREMENTS As derived in the preliminary telescope requirements, [AD5], the telescope has to provide the optical performance reported in the following table. Parameter Value Collecting area >0.6 m2 FoV 30” with diffraction limited performance 41” with optical quality TBD allowing FGS centroiding 50” unvignetted WFE Diffraction limited @ 3 µm Wavelength range 0.55-8 µm Throughput Minimum >0.78 Average >0.82 Output beam dimension 20.0 x 13.3 mm elliptical Table 29: Summary of the telescope optical requirements The most critical aspect of the telescope performance is likely to be the WFE. The requirement is to achieve diffraction limited image quality at 3 um, which enables the instrument to meet all of its scientific objectives. There is a derived requirement on the distribution of the WFE terms on the telescope since the coherence of the signal on the VISPhot channel is dependant on the distribution of the terms of the WFE adding up to a 200nm rms WF (equivalent to 3 um diffraction limit). See [RD30] for full details of this analysis.
11.4.1 Telescope FoV Requirements There are three aspects to consider: the FOV required for the FGS, the FOV required for the IR spectrometer (AIRS) and the FOV required for the telescope. These are all defined by requirements in the MRD, [AD1].
11.4.1.1 Assumptions • The FOV is defined as the full angle unless explicitly stated as a half-angle. E.g. a FOV of 20” implies +/- 10” total coverage.
Page 112 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• The APE is defined as the half angle. E.g. an APE of 10” implies a maximum pointing error of +/- 10” from nominal at 99.7% confidence level. • When the term diffraction limited is used in this discussion, it should be taken to mean diffraction limited at 3 um (in line with the telescope requirements).
11.4.1.2 Required FOV for the FGS The FGS FOV is defined by the following requirements from the MRD [AD1].
The inputs to the equation of R-PERF-040 are: λmax = 1.95 um Dtel = 0.9 m (taking the equivalent circular aperture of the baseline telescope) APEcoarse = 8” (R-AOCS-020) This gives: FoV(FGS) = 17” This is illustrated in Figure 67 below.
FGS FOV
Φ 17“
FGS Detector (or larger)
Page 113 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 67: FGS FOV
11.4.1.3 Required FOV for the spectrometer The spectrometer FOV is defined by the following requirements from the MRD.
The inputs to the equation of R-PERF-050 are: λmax = 7.8 um Dtel = 0.9 m (taking the equivalent circular aperture of the baseline telescope) APEfine = 1” (R-AOCS-030) This gives: FoV(science) = 6.4”
To this, one needs to add R-PERF-060 to allow sufficient sampling of the backgrounds in the cross-dispersion direction. Therefore the total spectrometer FOV is: FOV(science) = 6.4” x 26.4” So the slit footprint should be at least 26.4” long and 6.4” wide.10 This is illustrated in Figure 68.
10 Section Error! Reference source not found. gives different values (4.7” x 20” Ch0, 7.4” x 20” Ch1) due to a slightly different interpretation of the MRD requirements. The difference is not significant because the calculation in this section is only used to size the telescope FOV and the values used here give a larger overall FOV, thus building in some margin. The values in section Error! Reference source not found. are the dimensions used to size the AIRS slits.
Page 114 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
PSF AT LAMBDA MAX + FINE APE 4 . 6
AIRS SLIT
26.4
ALL DIMENSIONS IN ARC SEC
Figure 68: AIRS FoV
11.4.1.4 Required FOV for the Telescope The total FoV for the telescope is defined by the FoV requirements for the FGS and AIRS, plus R-PERF-070, shown below.
This additional FoV requirement is intended to allow for one-off movements of the telescope with respect to the instrument optical bench (for example due to launch loads, settling of the structure post-launch and dimensional changes on cool-down). To understand how this affects the total telescope FoV in practice it is necessary to consider a number of possible misalignment cases. Figure 67 shows the situation when the telescope, AIRS and FGS are all perfectly aligned.
Page 115 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
PERFECT ALIGNMENT
FGS FOV
PSF AT LAMBDA MAX + FINE APE
Ø17.0 6 . 4
AIRS SLIT
26.4 CENTRE OF TELESCOPE FOV
ALL DIMENSIONS IN ARC SEC
Figure 69: Perfect alignment of the telescope, AIRS and FGS. All dimensions are to scale. Figure 69 shows a 10” offset between the telescope and the instrument optical bench (R-PERF-070), applied along the length of the slit. In this situation we require a 26.4” diffraction limited telescope FoV to ensure there is a well resolved PSF at the centre of the slit. A larger 37” telescope FoV is required to allow the FGS to acquire a star and centre it on the slit. The image quality level over this 37” annulus is TBD but it can be of lower quality, sufficient to allow the star centroid to be well enough resolved to be initially located and then brought to the centre of the FGS FoV, where the telescope image quality is better. We refer to this annulus as the ‘FGS acquisition’ FoV. A still larger telescope FoV, extending to 46”, is required to capture the slit background. There are no image quality requirements over this additional FoV; the only real requirement is for the FoV to be unvignetted so that background photons reach the slit. We refer to this annulus as the ‘background’ FoV.
Page 116 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
AIRS AND FGS IN PERFECT ALIGMENT 10" OFFSET BETWEEN TEL AND OB
Ø46.4 TELESCOPE DIFFRACTION LIMITED FOV Ø37.0
Ø26.4
TELESCOPE ‘BACKGROUND’ FOV
TELESCOPE ‘FGS ACQUISITION’ FOV 10.0
ALL DIMENSIONS IN ARC SEC
Figure 70: 10” offset between the telescope and the instrument optical bench. Perfect alignment is assumed between AIRS and the FGS and between AIRS/FGS and the optical bench. All dimensions are to scale. However, this situation does not allow any margin for misalignment of the FGS and AIRS. They must be co- aligned to better than ±5” in any case, or else the PSF cannot be located on the centre of the slit and in the FGS FoV at the same time. We also need to allow some margin for the fact that the telescope to OB alignment will be set to some datum on the optical bench, and there may be some residual misalignment between that datum and the individual FGS and AIRS instruments. It is reasonable to suppose that these misalignments will be small in comparison to the offset between telescope and OB, given that AIRS, FGS and telescope will be integrated on a single optical bench. For the moment we assume that both the FGS and AIRS will be aligned to a datum on the optical bench (for example an optical reference cube) to within ± 2”, implying a maximum co-alignment error between the two them of 4”. We then assume that this reference cube is the datum to which the telescope is aligned. This implies that we must add a 4” margin on all of the telescope FoVs in Figure 70 to allow for these alignment errors. This gives the final telescope FoVs shown in Figure 71.
Page 117 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
10" OFFSET BETWEEN TEL AND OB 2" OFFSET BETWEEN AIRS/FGS AND OB ALIGNMENT DATUM
Ø50.4
Ø41.0
OPTICAL BENCH ALIGNMENT DATUM Ø30.4
10.0 2.0 CENTRE OF TELESCOPE FOV
ALL DIMENSIONS IN ARC SEC
Figure 71: Final telescope FoV. All dimensions are to scale. This gives the following FoVs (telescope values rounded to the nearest arcsec).
Designation FoV (arcsec) AIRS 6.4 x 26.4 FGS 17 Telescope: Diffraction limited 30 Telescope: FGS acquisition 41 Telescope: Background 50 Table 30: Summary of FoV requirements. Telescope FoV values are rounded to the nearest arcsec
11.5 TELESCOPE CHARACTERISTICS The characteristics of the optical design and of the optical components are summarised in the following tables, see [RD5]. Parameter Values Optical concept Eccentric pupil Cassegrain telescope plus off- axis paraboloidal mirror and a plane folding Afocal design Focal length (M1&M2) 14.17 m FoV center 0.1°- Off-axis YZ plane Pupil size Ellipse with major axis 1.1 m x 0.73 m Focal ratio 13 and 18.9 Angular magnification 55 Exit pupil size Ellipse: 20.0 mm x 13.3 mm Table 31: Summary of the optical design characteristics.
Page 118 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Optical element M1 M2 M3 R (mm) 2319.5 239.0 491.5 k -1 -1.4 -1 Off-axis (mm) 500 50 20 (y direction) Elliptical, 550 (x) Elliptical, 56 (x) Elliptical, 15 (x) Clear Aperture Radius (mm) by 365 (y) by 40 (y) by 11 (y) Type Concave mirror Convex mirror Concave mirror Table 32: Summary of the optical elements characteristics
11.5.1 Telescope Optical Performance
0.9958 3.0000 OBJ: 0.0000, 0.0000 (deg) OBJ: 0.0000, 0.0042 (deg) OBJ: 0.0000, 0.0057 (deg) OBJ: 0.0000, -0.0042 (deg) 0.8962 0.40 0.7966
IMA: -0.0000, 0.0000 rad IMA: -0.0000, -0.0042 rad IMA: -0.0000, -0.0057 rad IMA: -0.0000, 0.0042 rad 0.6971
0.5975 OBJ: 0.0000, -0.0057 (deg) OBJ: 0.0042, 0.0000 (deg) OBJ: 0.0057, 0.0000 (deg) OBJ: 0.0070, 0.0000 (deg)
0.4979
0.3983 IMA: -0.0000, 0.0057 rad IMA: -0.0042, 0.0000 rad IMA: -0.0057, 0.0000 rad IMA: -0.0070, 0.0000 rad 0.2987
OBJ: 0.0000, 0.0070 (deg) OBJ: 0.0000, -0.0070 (deg) OBJ: -0.0070, 0.0000 (deg) OBJ: -0.0042, 0.0000 (deg) 0.1992
0.0996
0.0000 IMA: -0.0000, -0.0070 rad IMA: -0.0000, 0.0070 rad IMA: 0.0070, 0.0000 rad IMA: 0.0042, 0.0000 rad Surface IMA: IMAGE SPACE Huygens PSF Spot Diagram
Figure 72 : Left: PSF calculated at the telescope FoV centre for a wavelength of 3 microns depicted over a 1 mrad
square box. Right: Spot Diagrams in the afocal space, the scale (box) is 0.4 mrad. Figure 72 shows the expected PSF of the telescope at the centre of the FoV for the 3 micron reference wavelength and calculated over a 1 mrad square box. Also the spot diagrams in the afocal space are shown, they are all well within the Airy disk diameter for the 3 micron wavelength. The RMS wavefront error over the whole telescope FoV is shown in Figure 73, the vertical lines are highlighting the diffraction limited field of 30”.
35.0
32.5
30.0
27.5
25.0
22.5 RMS_wv_error (nm) RMS_wv_error
20.0
17.5 -30 -20 -10 0 10 20 30 FoV(")
Figure 73: RMS_wv_error in the y direction of the FoV. In the x direction the wavefront error is constant. Figure 74 and Figure 75 show the Strehl ratio and rms wavefront error over the 30 arcsec diffraction limited field of view in the collimated space after M4. This shows the nominal image quality of the telescope. Page 119 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 74: Strehl ratio over a 30 arcsec square field of view at 3 um (nominal image quality)
Page 120 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 75: rms wavefront error over a 30 arcsec square field of view at 3 um (nominal image quality) Full details of the Telescope Assembly optical analysis (including tolerance analysis) are contained in [RD34]. The telescope RMS wavefront error is always less than 25 nm over the 30” nominal telescope FoV (see Figure 73 and Figure 75), that is well below the diffraction limit at 3 micron (220 nm). In fact the Strehl ratio is approximately 1 all over the FoV (see Figure 74). Theoretically the telescope is diffraction limited also for the minimum wavelength of the FGS, i.e. 0.55 nm.
11.5.2 Preliminary Mechanical Design The ARIEL Telescope is composed by three mirrors: M1 is elliptical with major axis 1.1m x 0.73m and M2 and M3 of smaller diameter (respectively about 110 mm and 30 mm), there is also a 30 mm diameter plane mirror (M4) for folding the beam exiting M3. All mirrors are made of Al6061, an alloy that, together with optical quality specifications, ensures a very good thermal uniformity. At present, the light-weighting level of the primary mirror is on the order of 50% for a mass around 90 kg. M1 is supported directly from the Optical Bench via a 9-point whiffletree structure that connects to the optical bench by 3 triangular mountings. The telescope mechanical configuration is shown in Figure 76. Also M2 and M3 are thermally decoupled from the supporting structures for stability purposes. The M2 is mounted on the telescope beam, an Aluminium arm rigidly connected to the OB, and is provided with a refocus mechanism (see dedicated chapter). The Telescope Beam is supported, on M2 side, by the central bipod of the ARIEL PLM. To increase the rigidity of the structure the optical bench, the rear bipods and the beam have been redesigned. The bipods legs have been re-arranged in position and angle (see mechanical analysis chapter) and their thickness has been increased. The optical bench stiffness has been enhanced by including the instrument box and adding stiffening blades. Two Al alloy bars from the OB to the M2 structure have been
Page 121 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 designed to increase the rigidity of the whole telescope structure. The baffle surface is mechanically supported on these two arms in a position close to the midpoint of its main axis. In this way thermo-elastic contraction will not apply additional stress to the other units. The baffle is also thermally connected with high conductance straps to the Optical Bench. This thermo-mechanical design can guarantee at the same time rigidity and a very good thermal conductance in all directions.
Figure 76: ARIEL telescope mechanical configuration.
11.5.3 Preliminary Thermal Design and Modelling The telescope is passively cooled to ≤70K and thermal control is based on a passive/active approach. A high level scheme of the Telescope Assembly is shown in Figure 77.
Figure 77: Simple schemes of the Telescope Assembly thermal configuration. The telescope baffle provides a large radiator area with a direct view to deep space; this provides sufficient radiative cooling to dump the parasitic loads from the PLM support struts, cryo-harnesses and radiative load from the final V-Groove. Temperature control of the mirrors is achieved by partial thermal decoupling from PLM units: each mirror is mounted on its supporting structure by insulating struts with a total conductance of less than 0.1 W/K (GM1-3, GM2, GM3 and GMech Figure 77). This configuration can filter out all potential instabilities with periods on the order of 10 – 100 s originated in the PLM. For the primary mirror, the high thermal capacitance, due to its mass, will allow a higher level of passive filtering, damping instabilities at lower frequencies (with periods on the order of few hours). The slower fluctuations (with periods on the order of several hours or longer) that could be transmitted to the optics will be smoothed by the active control system (see Chapter 7).
Page 122 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The telescope will also incorporate contamination control heaters on the M1, M2 and M3 mirrors. These heaters will be active during the early orbit operations to ensure that the sensitive optical surfaces remain warmer than the support structure through the critical parts of cooldown. Once the telescope assembly reaches 170K the decontamination heaters will be activated to maintain, by a proportional control loop, the mirrors at that temperature while the rest of the structure and baffle fall to lower temperatures. An initial calculation of the power required to maintain this temperature gradient shows that approximately 30 W (no margin) of heater power is required during this phase. This would hold the sensitive surfaces at 170K while the baffle and structure cool below 160K where they will act as a contamination getter for water and other contaminants being off-gassed by the PLM and optical surfaces. The telescope decontamination process and hardware is under the responsibility of the S/C thermal control. At the moment the telescope thermal analysis is performed with the general PLM TMM/GMM, as the resolution of the present model is considered detailed enough for the Phase A study of the mirrors and structure thermal behaviour and to identify the critical issues. In any case, the implementation of a detailed ESATAN TMM/GMM for the whole Telescope Assembly is already planned for future stages of the design activities. The results of the thermal analysis are reported in Chapter 7. In routine science operations the M1 operates at a temperature around 50K with a very high thermal uniformity, better than 10 mK. Even the whole Telescope Assembly shows a limited gradient, on the order of 2K, between the units. Transient simulations show that the telescope design, with the present level of expected fluctuations on the SVM interface and on the optical bench, is already capable of filtering out most of the instabilities down to induced oscillations of the M1 on the order of few mK. Anyway the implementation of a fully redundant thermal control system (electrical resistances + thermistors), driven by the TCU, is at present planned on M1 (the need on M2 is TBC) to ensure thermal stability in presence of oscillations with a wider frequency spectrum.
11.6 TELESCOPE ASSEMBLY DESIGN & ACCOMMODATION The optical and mechanical configuration of the preliminary design phase has been modified in order to satisfy the static, modal and thermo-elastic requirements. The design has been updated taking into account the allocated envelope for the cold payload module.
Figure 78 Telescope allocated envelope The optical bench, all the structural elements and the telescope mirrors are assumed to be made of Al 6061. The optical bench thickness has been increased to include the instrument box as part of the structure (Figure 79) by means of stiffening ribs on the front and rear surface (Figure 80). This H section gives at the bench a high structural rigidity to OUT-OF-PLANE load (high mass of objects).
Page 123 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The rear side instrument box cavity is closed with an Aluminum Alloy 6061 honeycomb panel which works as radiator (green transparent color in Figure 79).
Figure 79 Optical bench configuration
Figure 80 Optical bench ribs - front and rear Primary mirror M1 in Al 6061 is light-weighted with a pattern studied in order to give high bending stiffness both during manufacturing and in operating condition. The mounting is designed on the rear side with 9 holes on which three brackets (in 6061) are fixed. In this way the 9 points of M1 are reported to 3 points on the optical bench, ensuring the isostatic thermal fixation of the primary mirror.
Figure 81 Primary mirror M1 – rear side
Page 124 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The structure has been reinforced by the insertion of 4 lateral struts in Al 6061. These stiffening arms guarantee small deformations and high bending stiffness between M1 and M2 during application of the in-plane and out-of-plane loads.
Figure 82 Lateral struts The struts work also as support of the baffle which is an Aluminum Alloy 6061 open honeycomb. With this mounting the baffle will not have any structural function, but it will protect the optical beams from stray light and it contribute to the thermal stability of the telescope.
Figure 83 Baffle The rear bipods are subject to a high load (due to the mass of optical bench and M1). In order to be compliant to the static and modal requirements, their geometrical position and size had to be optimized, with respect to the optical bench, by rotating the angle to telescope z-axis and increasing the legs opening angle and the section.
Page 125 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 84 Rear bipods
Figure 85 Rear bipods section
These adjustments require an update to the PLM mounting interface ring at the SVM: the diameter needs to increase to 1850 mm.
Page 126 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 86 PLM attachment points envelope
11.7 TELESCOPE ASSEMBLY PRELIMINARY MECHANICAL ANALYSIS SUMMARY
11.7.1 Analysis at M1 Mirror Level A first static analysis has been performed during the design of the M1 mirror in order to verify that the implemented lightening were stiffness enough to avoid deformation while moving the mirror from the manufacturing position to the operative one. During manufacturing and final polishing the mirror will be placed in horizontal plane, while during integration and on ground test the mirror will be mounted on the telescope in vertical position, with the fixation on the rear face to the optical bench.
Figure 87 M1 Mirror in the manufacturing position
Page 127 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 88 M1 Mirror in operative position The static analysis has been executed on the stand alone mirror without the optical bench but only the mounting plates. The following figures report the distortion map for mirror in Al 6061 (m = 85 kg) and mounting plates in 6061 Al (m = 15 kg).
Figure 89 Distortion map – Manufacturing position
Page 128 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 90 Distortion map – Operative position, front size
Figure 91 Distortion map – Operative position, rear size
11.7.2 Analysis at Telescope Level – Alignment Position The static analysis at telescope level has been executed in order to verify the deformation of the M1 mirror due to the gravity effect. The analysis has been executed at room temperature (293,15K).
Figure 92 Telescope static analysis system reference
Page 129 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
M1 is mounted with an angle with respect the vertical axis of the spacecraft: a first result of the analysis is that the telescope shall be tested and calibrated on ground tilted of 12.191° in order to avoid deformation due to gravity effect.
Figure 93 Tilt for on ground test and calibration The results of the analysis during test and calibration phases are reported hereafter:
ΔxM1-M2 38,4nm
ΔyM1-M2 44,3µm
ΔzM1-M2 659nm
Figure 94 Telescope Static Analysis – Deformation on XA_TEL axis
Page 130 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 95 Telescope Static Analysis – Deformation on YA_TEL axis
Figure 96 Telescope Static Analysis – Deformation on ZA_TEL axis
11.7.3 Design limit loads (DLL) at Telescope Level Stress analysis has been performed on the basis of the design limit loads reported in the up to date issue of the PID-A document applied to the CoG of the telescope without baffle Results of the analysis are reported hereafter in the table: Ariel Al 6061 Titanium-Alloy G10 Optical reference Req Remarks Axis Yield Strenght Yield Strenght Yield Strenght frame <276MPa <750MPa <310MPa
1MPa x 7 = 53MPa x 7 = XA_TEL -ZARIEL 7g 1.6MPa x 7 = 11,2MPa In plane 7MPa 371MPa
1MPa x 15 = 6,7MPa x 15 = 0.94MPa x 15 = YA_TEL YARIEL 15g out-of-plane 15MPa 100MPa 12,6MPa
ZA_TEL XARIEL 7g 1MPa x 7 = 11,6MPa x 7 = 1.2MPa x 7 = 8.82MPa In plane
Page 131 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
7MPa 81,2MPa
Table 33 DLL analysis results
Figure 97 DLL – Stress on XA_TEL axis
Figure 98 DLL – Stress on YA_TEL axis
Figure 99 DLL – Stress on ZA_TEL axis
Page 132 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
11.7.4 Modal Analysis The dynamic load analysis has been performed in launch conditions (except for the M1 baffle which has not been considered in the FEM) with the telescope bolted at its flight interface to a rigid structure (“hard mounted”). The following first mode frequencies result for the analysis (for the telescope assembly alone). The full PLM mechanical analysis is reported in section 6 and in detail in [RD3].
Ariel reference Eigen Optical Axis Requirement Margin Remarks frame frequency
XA_TEL -ZARIEL >30 Hz 34,6Hz 15,3% In plane
YA_TEL YARIEL >50 Hz 74,2Hz 48,4% out-of-plane
ZA_TEL XARIEL >30 Hz 52,2Hz 74% In plane
Table 34: Telescope Assembly First mode Frequencies
Figure 100 Modal Analysis – First mode XA_TEL axis
Figure 101 Modal Analysis – First mode ZA_TEL axis
Page 133 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 102 Modal Analysis – First mode YA_TEL axis
11.7.5 Preliminary Thermo-Elastic Analysis The thermo-elastic analysis has been performed in order to verify the distortion on the primary mirror and between M1 and M2 during cooling from room temperature (293K on each element) to the operative condition as result of the Thermal Analysis. The S/C plane where the PLM is fixed has been assumed as a fixed rigid boundary made of Carbon Fiber at a uniform temperature of 270K. The thermal map assumed for the cold condition of the telescope is summarized in the following figures.
Figure 103 Thermal map on rear and front bipods
Page 134 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 104 Thermal map on M1 and structure elements
Figure 105 Thermal map on optical bench The directional deformation X, Y, Z (optical reference system) between M1 and M2 (centers of the mirrors) is reported in the following table
ΔxM1-M2 0,11 mm
ΔyM1-M2 0,65 mm
ΔzM1-M2 6,65 mm Table 35 Deformation between M1 and M2 The M1 deformation map for the three axes is shown in the figures below.
Page 135 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 106 Directional deformation - XA_TEL axis
Figure 107 Directional deformation - YA_TEL axis
Figure 108 Directional deformation - ZA_TEL axis
Page 136 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
11.7.5.1 Stress on the bipods In this section is reported the stress on the bipods, as they are the only struts of the Telescope Assembly that operate at ambient and cryogenic temperatures at their ends. All other TA units gradient is within few K. The stress due to thermo-elastic contraction is compared to launch loads. Figure 109 shows the stress induced by the cryogenic temperature. Figure 110 and Figure 111 report the details of the stress on the bipods due to the loads at launch.
Figure 109 Stress on the bipods due to cryogenic temperature
Figure 110 Stress on the bipods due to in plane loads at launch (ZA_TEL and XA_TEL)
Figure 111 Stress on the bipods due to out of plane loads at launch (YA_TEL)
Page 137 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The thermo-elastic stresses result higher than the structural loads: launch loads touch levels on the order of magnitude of 1 MPa. Cryogenic stresses may reach up to more than 20 MPa on the bipod legs head, while the foot experiences loads of 5 MPa (see Figure 109). In any case, all these loads are well below the Yield strength of the materials considered (hundreds of MPa). The safety margin for the softer materials, Al 6061 and G10, is around 10 on the legs head and more than 50 on the foot. On Ti alloy elements the margin is a factor of 2.5 higher. The present thermo-mechanical design seems compliant to the required level of performances under load and stability.
11.8 M2 MIRROR MECHANISM (M2M) DESIGN BASELINE The mechanism which is necessary to ensure that the telescope is in best focus and meets it’s WFE requirements when in operations is located on the M2 mirror. This mechanism builds on the heritage from other similar M2 mechanisms which has been developed at Sener, ES. Specifically it builds on the design heritage from the Gaia and Euclid M2 mechanisms. During the phase A stud an initial assessment of the possible requirements on the M2 mechanism has been made. Some key requirements will require revisiting in the early part of phase B with the knowledge that has been derived from the final round of Telescope assembly thermo-elastic analysis; specifically the range capability in both translation and rotation will be reassessed. Details of the Sener evaluation of the requirements against the design heritage can be found in [RD25]. A summary of the comparison between the ARIEL requirements and those on Gaia and Euclid can be seen in the table below.
Table 36: Comparison of Gaia, Euclid and ARIEL M2M Requirements and Design
The actuator used for these mechanisms is illustrated in Figure 112 below.
Page 138 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 112: Actuator used in Gaia and Euclid M2M
A development plan has been outlined in [RD25] with the activities that will be required to bring this mechanism to TRL6 by the time of mission adoption at the end of phase B1. This has been incorporated in to the Spanish consortium planning for the next phase.
11.9 PATHFINDER TELESCOPE MIRROR PROGRAM
11.9.1 PTM Plan The viability of Aluminium as the baseline material for the telescope mirrors has been assessed during the phase A by producing a pathfinder mirror from the baseline material. The planned baseline scope for the PTM program during phase A was agreed (at the MCR) to be: Procurement of a Al 6061 T651 blank Design of flight primary mirror and translation/simplification to a Pathfinder build standard (spherical vs off-axis parabola). Rough machining and some light-weighting (TBC) of blank to form 1.1 x 0.7 m elliptical primary mirror with a spherical surface form with equivalent sag to the flight design. Diamond turning of PTM mirror with target of 1 µm shape accuracy. Measure shape accuracy and roughness achieved on PTM after diamond turning. Analyse, then measure, the effect of gravity offloading on the mirror when mounted in a representative fashion. Further details of the plans for the PTM mirror can be found in [RD10].
Page 139 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Further use of the PTM mirror in phase B (if mission is selected) may include: the polishing of sub-areas to demonstrate full performance to the WFE and surface roughness requirements; cryogenic testing of the PTM to measure ambient to operational changes in WFE; use in the STM of the PLM.
11.9.2 PTM Progress
11.9.2.1 Design and Analysis Details of the PTM program are given in [RD33] and summarized here. The baseline PTM mirror has the same form (elliptical surface 1110 x 768 mm) as the primary, but with a spherical surface in order to ease manufacturing and testing. The RoC of the spherical surface is chosen in order to match the sag of the planned primary mirror. The PTM mirror has a coarse lightweighting implemented corresponding to 36% lightweighting – this is studied further in the in-flight mirror design above. A simplified drawing of the M1 mirror demonstrator is shown below.
Figure 113: Illustration of M1 PTM
Page 140 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
A set of analyses has been carried out to show the mirror deformation under gravity loading corresponding to the manufacturing and measurement conditions vs the operational condition. These are shown in detail in [RD33] and an example output for the baseline wiffle-tree mounting scheme is shown below.
Figure 114: Deformation load case LC7 FEM output for PTM mirror
11.9.2.2 Fabrication The PTM mirror has undergone procurement and fabrication during phase A, see [RD33] for details. The milling activities have been completed, and at the time of writing the PTM is undergoing the final stages of diamond turning – the outputs from this will be provided to the ESA MSR review team as soon as they are available (expected by end Feb 2017). The PTM mirror after milling is shown in the figure below, also shown in the surface error on the milled surface with respect to the theoretical surface as measured by Media Lario on the CMM.
Figure 115: PTM mirror after milling process completed
Page 141 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 116: Deviation from theoretical surface of the PTM after milling
11.9.2.3 Polishing Trials In addition to the work on the main PTM mirror, trials have been undertaken during phase A to demonstrate the surface roughness that is achievable on the Al6061-T651 samples using the polishing processes and machines foreseen for the M1 production. Details are again contained in [RD33]. A number of different pads and slurry combinations in order to optimise the figuring process. The final roughness polishing was also trialed, deomstrating the capability to achieve areas with a surface roughness below 10 nm rms as shown in the figure below.
Figure 117: Talysurf CCI surface roughness measurements (50X) after a Zeeko polishing run on Al6061-T651 using aluminium oxide slurry and a non-woven synthetic pad
Page 142 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
12 ARIEL IR SPECTROMETER (AIRS) DESIGN
12.1 AIRS DESIGN ARCHITECTURE AND FUNCTIONAL ANALYSIS AIRS is the ARIEL scientific instrument providing low-resolution spectroscopy in 2-IR channels (called Channel 0 CH0 for the [1.95-3.90] µm band and CH1 for the [3.90-7.80] µm band). It is located at the intermediate focal plane of the telescope + common optical system. AIRS instrument is composed of three main architectural blocks: • AIRS Optical Bench (AIRS-OB) • AIRS Focal Plane Assembly for Channel 0 and 1 (AIRS-FPA-0 and AIRS-FPA-1) • AIRS Detector Control Unit (AIRS-DCU) AIRS-OB and AIRS-FPA 1 & 2 are located on the cold section of the PLM, while the AIRS-DCU is located in the warm part of the SVM.
FGS COLD PLM Telescope WARM SVM
AIRS OB AIRS FPA
Common Optics
AIRS ICU AIRS DCU
Figure 118: Architectural block diagram of the ARIEL Infra-Red Spectrometer.
The AIRS Optical Bench is the main structural element of the AIRS, providing the following functions • function-AIRS-OB-01: Mechanical I/F to the ARIEL Thermal Optical Bench (localisation and alignment of the AIRS mechanical references with respect to TOB references) • function-AIRS-OB-02: Thermal I/F to the ARIEL Thermal Optical Bench at 55 K ensuring passive cooling of the Optical elements and of the FPA). • function-AIRS-OB-03: Structure support for the CH0 and CH1 optical elements (sub-systems for each channel composed of: slit, collimator, prism, folding mirror, camera). This includes implementation of the optical prescription for spectroscopy (with Resolution R ≥ 100 in CH0 and R ≥ 30 in CH1) and alignment of the optical elements with respect to the mechanical references of AIRS-OB. • function-AIRS-OB-04: Light tightness, contamination control and baffling to limit in field and out of field straylight. • function-AIRS-OB-05: Mechanical I/F to support 2 Focal Pane Assemblies (FPA-CH0 and FPA- CH1). • function-AIRS-OB-06: Thermal I/F for passive cooling of the 2 Focal Pane Assemblies (FPA-CH0 and FPA-CH1) at 55 K. The AIRS Focal Plane assembly of Channel CH0 and Channel CH1 is composed of a thermal- mechanical structure interfacing with the AIRS-OB and containing the detector Sensor Chip Assembly and its Cold Front end electronics with the associated harness. For each Channel, the FPA are providing the following functions:
Page 143 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• function-AIRS-FPA-01: Mechanical I/F to the AIRS-OB (localization and alignment of the AIRS- FPA mechanical references with respect to AIRS-OB references). • function-AIRS-FPA-02: Thermal I/F to the AIRS-OB at 55 K ensuring passive cooling of the FPA housing and of the FPA Cold Front End Electronics). • function-AIRS-FPA-03: Light tightness, contamination control and baffling to limit out of field straylight. • function-AIRS-FPA-04: Structure support for the Detector Cold Front End Electronics. • function-AIRS-FPA-05: Routing of the Cryo Flex Harness from the Cold Front End Electronics to the Detector Sensor Chip Assembly. • function-AIRS-FPA-06: Structure Support for the Detector Sensor Chip Assembly containing the Photon Sensitive Device (including localization and alignment of the Sensitive Area with respect to FPA mechanical reference). • function-AIRS-FPA-07: thermal decoupling of the SCA from the rest of the environment (including CFEE through CFC and FPA housing through mechanical mounting) in order to ensure functioning temperature of <42 K for the detector. • function-AIRS-FPA-08: Thermal and mechanical interface to the Cryo-cooler Cold finger as defined with external interfaces. • function-AIRS-FPA-09: Active Thermal control of the temperature of the detector with heaters and temperature probes located on the SCA to +/- 0.05 K over 10 hours. • function-AIRS-FPA-09: Electronics control of the detector through a dedicated Cold Front end electronics. The AIRS Detector Control Unit (AIRS-DCU) is a Warm Front End Electronics unit located on the warm part of the SVM, which interfaces internally to the FPA CFEE and the detector in order to control the detection process, and externally to the ICU to transfer the Science Data Packet. The main function of the DCU are: • function-AIRS-DCU-01: provide Housekeeping for AIRS-OB and AIRS-FPA. • function-AIRS-DCU-02: control data acquisition at detector level through the CFEE from users entry. • function-AIRS-DCU-03: process the data from the detector to create the Science Data Packet. • function-AIRS-DCU-04: Control the thermal stabilization of the Detector through temperature probes and heaters. • function-AIRS-DCU-05: Ensure interface with the ICU for TC reception, HK and Science Data Packet transmission. The design of the FPA to the DCU harness at will be performed at PLM level, especially for the introduction of bracket to guide the harness towards the SVM. The AIRS will be specifying the electrical requirements on the CFEE to DCU harness but the mechanical and thermal implementation shall be under PLM/SVM responsibility as it depends on the actual routing and specific design of the PLM/SVM.
12.1.1 Spectrometer Architecture Trade-Offs Document AIRS optical design trade-off analysis ARIEL-CEA-INST-DD-001 [RD2] contains an analysis of sensitivity and performance comparison of prism and grating option for the ARIEL spectrometer. The recommendation from this trade off analysis on the optical design is to implement as baseline a prism system with 2 independent optical path CH0 [1.95-3.90] µm and CH1 [3.90-7.80] µm with the following configuration: • an input image of the target in the entrance focal plane of the spectrometer including a large slit which function is to limit the background flux on the detector and act as a field stop
Page 144 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• A collimator that collimate the incoming f#=18 beam on the prism • A prism which is the dispersive element • A camera which function is to re-image the beam on the detector • A focal plane array detector with reference pixel pitch is 18 µm (sensitivity to pixel size is given section 3.3.1 of AIRS optical design trade-off analysis ARIEL-CEA-INST-DD-001). The key parameters of the system are given in Table 37: Channel 0 Channel 1 Waveband [1,95-3,90]µm [3,90-7,80]µm Material CaF2 CaF2 Incidence angle 72,05° 61,15° Prism apex 79,5° 17,50° Transmission (Fresnel) 75,20% 89,30% ∅ Collimator 6 mm 6 mm Collimator focale length 108 mm 108 mm ∅ Camera λmin 9,1 mm 9,0 mm ∅ Camera λmax 9,5 mm 8,8 mm Camera focale length 137,3 mm 68,4 mm PSF sampling λmin 2 2 PSF sampling λmax 3,83 4,1 Spectrum Length 267 pixels 94 pixels Table 37: Table of characteristics of the baseline prism design in CH0 and CH1 (scaled from the trade off analysis for telescope f# and 18 µm baseline pixels). The main expected advantage of this baseline would be to have a better-expected SNR because of the capability to circularise the PSF through the prism anamorphose, and to have an enhanced transmission compared to estimated efficiencies of the grating, despite the variable resolution inherent to the prism. The prism design with relatively modest size elements could be a safe approach for the current phase.
12.2 AIRS OPTICAL MODULE DESIGN BASELINE
12.2.1 Optical Design Following the trade-off analysis and its outcome recommendation for a prism system baseline, a detailed implementation of the design has been performed under Zemax for further analysis. The implementation is made at operational temperature of 55 K and the indexes of refraction are defined at this temperature. The selected material for the prism is CaF2, which has heritage from previous IR space missions.
Figure 119: Zemax implementation of Channel 0 spectrometer.
Page 145 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 120: Zemax implementation of Channel 0 and Channel 1 spectrometer. Zemax implementation of the baseline recommended in the trade-off analysis report confirms the performance of the predicted system in terms of resolution and PSF. The detailed Zemax model introduces doublets systems for the Camera (CaF2/Sapphire) and the Collimator (CaF2/ZnSe) in order to control the chromatic aberrations. With this correction the system is diffraction limited over the useful wavelength range. A fold mirror is inserted on the optical path following between the collimator and the prism in order to allow having both channel entrance planes and exit plane (detector plane) collocated and to optimize the location of the exit focal plane above the entrance slit. This solution improves the volume implementation of the overall instrument on PLM Thermal Optical Bench (TOB) and limit the distance from the AIRS-FPA to the SVM for cryo-cooler harnesses.
CH0 CH1
Figure 121: Detailed Optimized folded Zemax implementation of Channel 0 and Channel 1 spectrometer. The field of the spectrometer is modelled with the assumption that the ellipticity and f# of the input beam is oriented with respect to the AIRS entrance focal plane according to agreed interface with the Telescope and COB (AIRS Requirement Engineering Document ARIEL-CEA-INST-RS-002 [AD2]).
Page 146 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
CH0 CH1 AIRS Slit Spectral direction 4,7 7,4 arcsec AIRS Slit Spectral direction 0,296 0,465 mm AIRS Slit Spatial direction 20,0 20,0 arcsec AIRS Slit Spatial direction 1,26 1,26 mm λmax 3,9 7,8 µm Dtel 0,73 0,73 m APEfine(3s) 1 1 arcsec
Table 38: AIRS slit dimensions for CH0 and CH1. The magnification of the system is 0.785 in CH0 and 1.57 in CH1 (1 µm in entrance focal plane of CH0/CH1 = 0.785 / 1.57 µm in image focal plane), leading to a platescale 0.224 arcsec/pixel in CH0 and 0.448 arcsec/pixel in CH1. From the size of the spectrum and the size of the slit, we can derive the size of the useful area in the detector focal plane for imaging the spectrum and the associated flux calibration area: • CH0 window: 270 spectral pixels x 64 spatial pixels • CH1 window: 100 spectral pixels x 64 spatial pixels
The size in the spectral direction is justified by the need to have the full spectrum in the window. The size in the spatial dimension (64 pixels) is justified by the need to have sufficient pixels on the border of the spectrum to monitor the dark current and the background.
CH0 spectrum window: 270 x 64 pixels CH1 spectrum window: 100 x 64 pixels
Figure 122: AIRS CH0 spectrum window 270 x 64 pixels and CH1 spectrum window 100 x 64 pixels.
12.2.2 Optical Performance Predictions The circularisation of the PSF using the prism anamorphosis is observed on the implemented design and the size of the PSF first dark ring at λmin is 36 µm providing 2 pixels of 18 µm sampling (see Figure 123)
Figure 123: PSF of the CH0 at λmin = 1.95 µm.
Page 147 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 124: PSF of the CH1 at λmin = 3.90 µm. The current budget allocation for the AIRS Photon Conversion Efficiency is summarised in Figure 125. The budget for the slit is set at the maximum λ of each waveband as the slit width is set according to R- AIRS-P-0610 and R-AIRS-P-0612 to fit the first Airy ring of the PSF at λmax (a maximum loss of 0.85 is allocated to this contribution which is conservative as I would be the loss corresponding to a circular slit sized at first Airy ring, in the case of the rectangular slit, the transmission loss is most likely to be >95%). For λ below λmax, transmission of the slit increase as a function of the variation of the PSF with λ for the fix size slit. AIRS Optical elements Photon conversion Efficiency CH0 0,33 Slit (λmax) 0,85 Collimator doublet 0,97 Prism (CaF2) 0,75 Fold Mirror 0,99 Camera doublet 0,98 Detector QE 0,55
CH1 0,39 Slit (λmax) 0,85 Collimator doublet 0,97 Prism (CaF2) 0,89 Fold Mirror 0,99 Camera doublet 0,98 Detector QE 0,55
Figure 125: Budget allocation for CH0 and CH1 photon conversion efficiency. The current design allows meeting Photon conversion efficiency with margin. Transmissions of the prisms are evaluated here under Fresnel assumption. Anti-reflecting coating will be implemented in order to improve transmission.
Page 148 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Sub-System Transmission CH0 Sub-System Transmission CH1 0,8 0,9
0,7 0,8
0,7 0,6 0,6 0,5 0,5 0,4 AIRS-OB AIRS-OB 0,4 AIRS AIRS 0,3 Transmission Transmission AIRS Req 0,3 AIRS-Req 0,2 0,2
0,1 0,1
0 0 1,5 2 2,5 3 3,5 4 4,5 3,5 4,5 5,5 6,5 7,5 8,5 Lambda (µm) Lambda (µm)
Figure 126: Photon conversion Efficiency of the full spectrometer channel 0 and channel 1 (including allocation to telescope and common optics) and associated requirement on FoM. Resolution R>100 in CH0 and R>30 in CH1 is met by this prism based. The gain in Transmission is used to compensate for the variable resolution of the prism which tend to dilute the resolution element corresponding to the required resolution over more pixels (see Figure 127).
Figure 127: Variable resolution of the prism design with lambda for CH0 and CH1. Relative performance analysis in AIRS optical design trade-off analysis ARIEL-CEA-INST-DD-001 shows that the effect of increasing the transmission and circularising the PSF overweight the loss due to resolution increase. Based on the baseline optical design and flux at input of the telescope from R-AIRS-P-0350 and R-AIRS- P-0360 the following flux at detector focal plane can be estimated and used for detector and DCU requirements:
Page 149 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Baseline Optical Design
Spectrum size CH0: 270 pixels x 64 pixels CH1: 100 pixels x 64 pixels CH0: Maximum flux: 1100 e-/s/pix Faint target GJ 1214 flux CH0: Minimum flux: 110 e-/s/pix CH1: Maximum flux: 2400 e-/s/pix CH1: Minimum flux: 60 e-/s/pix CH0: Maximum flux: 225000 e-/s/pix Bright target HD 219134 flux CH0: Minimum flux is 12500 e-/s/pix CH1: Maximum flux: 270000 e-/s/pix CH1: Minimum flux: 10000 e-/s/pix Dynamics Faint target CH0: 10 CH1: 40 Dynamics Bright target CH0: 18 CH1: 27
Figure 128: Table of Flux at detector level (e-/s/pix) for brightest and faintest targets in the baseline design case.
12.3 THERMAL AND MECHANICAL DESIGN
12.3.1 AIRS Global Thermal and Mechanical Design The approach for the mechanical design at that stage is to have two independent spectrometer half- boxes, each containing one channel, that are ultimately assembled together. The folding of the two channels in the volume is described in Figure 129.
Detector
Camera
Slit
Collimator Prism
Fold Mirror
Figure 129: Implementation of optical design into the allocated volume: baseline case with 2 detectors. The overall maximum size of the volume is set by the CH0 longest arm. As the beam is collimated for both channels between the collimator and the camera, the design can easily be adapted for folding and matching length and alignment of both arms to a single plane for the entrance focal plane and for the detector focal plane with a single folding mirror.
Page 150 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The support holders for all the optical elements will ensure decoupling of CTE mismatch during thermal cool down and stiffness for vibration. The two CaF2 prisms are cylindrical with the faces cut. The prisms are glued with 2216 on Stainless Steel 304L mounting system pads (ensuring correct CTE match through close CTE material selection). Mirrors are fly-cutted on the mount in RSP Aluminium 6061-T651. The upper part of the mount can be adjusted in rotation while the lower part can be adjusted in translation. The use of specific tools allows for alignment after integration on the bench through the cover. The collimator and camera doublets lenses are hold in compression using copper-beryllium spring washer to maintain in location the optics against location embossment. An internal shim allows adjustment of distance between the two lenses of the doublet.
Prism Fold Doublet
Figure 130: Support holders for the Prism / Fold Mirror / and Doublet (Camera and Collimator). The full structure for the AIRS Optical Bench will be in aluminium 6061 to provide homothetic shrinking design at cold and good thermal coupling with the ARIEL PLM Thermal Optical Bench. AIRS assembly is integrated on the TOB via 3 points I/F and 2 pins system for accurate alignment. Adjustment without mechanism will be the baseline for optical alignment of AIRS to ARIEL telescope. Detailed design will be provided in the coming work but alignment procedure could rely on: • Optical alignment at AIRS instrument level through as-measured shimming process • Optical alignment at optical component level through as-measured shimming process • Best focus detector shimming
Figure 131: Accommodation of the AIRS on the PLM Thermal Optical Bench.
Page 151 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The AIRS-OB cools-down passively to 55-K through thermal conductance at interface with the ARIEL Thermal Optical Bench. This ensures minimal thermal noise on the Detection System. The overall assembly matched the 300 mm x 300 mm x 300 mm = 0,027 m3 volume with a current estimated volume of [AIRS-OB + 2 AIRS-FPA] = 482 mm x 184 mm x 206 mm=0,018 m3 (see Figure 132) and the global envelope of AIRS fits onto the TOB allowed volume. mm 184 mm distance = distance 90 Slit to Slit
H = 206 mm = H
Slit Height = 40 mm
L = 482 mm
Figure 132: Expected volume of the instrument including opto-mechanical bench and detectors boxes. The whole system fits with 25% margins in the 15 kg budget with the following mass break down:
Page 152 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
AIRS Products Mass (g) CH0 3766 Structure 2796 Optics and Mounts 220 Screws 100 AIRS-FPA CH0 650
CH1 3669 Structure 2713 Optics and Mounts 206 Screws 100 AIRS-FPA CH1 650
Common parts 63 Foot 63
Thermal Control 200 MLI 200
Total (no margin) 7698
Total (25% margin) 9623 Figure 133: Expected mass of the instrument including opto-mechanical bench and detectors boxes. This allocation does not take into the following elements: • Thermal truss and cold finger from the cryocooler to cooldown the detector. • Harness between the FPA-CFEE and the DCU located on the SVM (shared design PLM/AIRS).
12.3.2 AIRS-Focal Plane Assembly Specific Thermal and Mechanical Design The 2 AIRS Focal Plane Assemblies (AIRS-FPA) are thermal mechanical boxes containing for each channel (CH0 and CH1): • Sensor Chip Assembly (AIRS-FPA-SCA) with the detector and associated Read Out Integrated Circuit, and the thermal control (2 temperature probes + 2 heaters per SCA). The AIRS-FPA-SCA is thermally decoupled from the mechanical structure and regulated through a thermal control at 42 K+/-0.05K. • Cold Front End Electronics (AIRS-FPA-CFEE), linked to the Detector ROIC with a dedicated harness (Cryo Flex Cable CFC), and passively cooled to 55 K through the AIRS-Optical Bench.
Connector Tprobe to DCU Mechanical Housing AIRS - FPA - CFEE (SIDECAR) CFEE
Temperature probes Cable CFEE to DCU Detector Cryo Flex Flex Cryo Connector SCA
Heater Connector Heater to DCU I/F to the Cold finger
Figure 134: Design of the AIRS Focal Plane Assembly with the detailed units (one FPA per channel).
Page 153 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The AIRS-FPA-CFEE of each channel is interfacing to the AIRS-DCU located on the SVM through dedicated harness. The thermal power load of the AIRS-FPA-SCA has been optimised in order to remain below 15 mW at 42-K. In particular: • AIRS-FPA-SCA has been thermally decoupled from the AIRs Optical Bench at 55-K • AIRS-FPA-SCA has been thermally decoupled from the AIRS-FPA-CFEE at 55-K • Contribution from the wires for the thermal control Temperature probes and heater has been accounted • Joule contribution from the detector has been accounted.
Heat load from Thermal control to SCA (through wires) 42 K Flex) through 55 K
Heat load from Optical Bench to SCA (through pads) CFEE to SCA ( from 55 K load Heat
Heat Sink Cold Finger I/F <40 K
Figure 135: Thermal Description of the AIRS Focal Plane Assembly with parasitic flux contributors and Cold finger. One critical element of the FPA Thermal design is the Cryo Flex Cable in order to limit the parasitic leakage from 55 K SIDECARs toward the 42 K SCA. Analysis was performed to optimize material / track and shielding design in order to limit the flux toward detector. The harness track design is based on assumption of 37 tracks to read the H1RG (2 outputs max: baseline 1 output) with 3 metallic layers (2 shielding / 1 track). The Harness length is set to 100 mm.
Integral Thermal Integral Electrical Number Fill Factor Conductivity Total Section Flex Layer Material Function Resistivity Section (m²) per Length (m) Thermal Losses (W) (Unitless) (42K --> 55K) (m²) Ohm .m layer W/m Layer 1 Kapton/Adh. Coverlay 1 1,258E+0 W/m 1,170E-6 m² 1 1,170E-6 m² 1,000E-1 m 1,472E-5 W Layer 2 Constantan Elec. Shield 0,65 2,273E+2 W/m 4,642E-7 Ohm.m 1,268E-7 m² 1 1,268E-7 m² 1,000E-1 m 2,881E-4 W FLEX Layer 3 Adh./Kapton/Adh.Insulator 1 1,258E+0 W/m 1,170E-6 m² 1 1,170E-6 m² 1,000E-1 m 1,472E-5 W Layer 4 Constantan Elec. Tracks 1 2,273E+2 W/m 4,642E-7 Ohm.m 5,000E-9 m² 37 1,850E-7 m² 1,000E-1 m 4,205E-4 W
CONSTANTAN CONSTANTAN Layer 5 Adh./Kapton/Adh.Insulator 1 1,258E+0 W/m 1,170E-6 m² 1 1,170E-6 m² 1,000E-1 m 1,472E-5 W Layer 6 Constantan Elec. Shield 0,65 2,273E+2 W/m 4,642E-7 Ohm.m 1,268E-7 m² 1 1,268E-7 m² 1,000E-1 m 2,881E-4 W Layer 7 Adh./Kapton Coverlay 1 1,258E+0 W/m 1,170E-6 m² 1 1,170E-6 m² 1,000E-1 m 1,472E-5 W Kapton Losses 5,887E-5 W Constantan Losses 9,967E-4 W Flex Total Losses 1,056E-3 W Figure 136: Thermal Description of the AIRS-FPA Cryo Flex Cable (CFC) with estimation of parasitic flux contributors in the baseline Constantan and track design. A trade off analysis on material (Manganin / Constantan / CuBe / Copper) has been performed showing that Copper alloy shall be preferred to pure metal as they are not sensitive to Residual Resistance Ratio effect. Design baseline will be based on Constantan Harness leading to Flex loss of ~1 mW from the SIDECAR to the SCA.
Page 154 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The support structure holding the SCA on the FPA housing is based on hexapods in order to allow good structural behavior in vibration and accurate positioning while ensuring correct thermal decoupling of the 42-K stage from the 55-K main bench. Current baseline design relies on 6 in Titanium leading to a total loss from the FPA structure toward the SCA of 6x 1 mW = 6 mW Thermal loss of the thermal control is estimated under assumption of 2 heaters and 2 temperature probes (with 4 manganin wires per devices for hot redundancy design) leading to power load of < 0,5mW The consolidated dissipated power budget at the FPA-SCA 42-K stage is estimated to < 15 mW including detector internal power and thermal control dissipation. Contributors to Thermal Budget SCA Estimated allocation
Thermal Parasitic through Cryo Flex 1 mW
Thermal Parasitic through Hexapods Structure 6 mW
Thermal Parasitic through thermal Control wires 0,5 mW
SCA electrical dissipation 3 mW
SCA thermal control 2 mW
TOTAL 12,5 mW
Figure 137: Consolidated Power budget at SCA level.
12.3.3 AIRS-Focal Plane Assembly Detector The detector requirements can be derived from the range of object described in requirements R-AIRS-P- 0350 and R-AIRS-P-0360, and from the baseline assumption on telescope and AIRS optical design. In particular the minimum and maximum fluxes for the faintest and brightest objects summarized on Figure 128 are used to size the window frame rate and the linearity of the Full Well range.
Page 155 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Parameter Requirement Comment Spectral Coverage CH0 : [1.95-3.90] µm [λmin and λmax] for which QE performance are required. CH1 : [3.90-7.80] µm Format > 320 x 256 pixels The minimum format required from the window is compatible with this format. Pixel pitch 15 µm to 18 µm This can be adapted by adjustement of the Camera focal length Min signal dynamics CH0 : 100-e-/s/pixel Minimum star in peak PSF CH1 : 60 e-/s/pixel Max signal dynamics CH0 : 225000 e-/s/pixel Maximum star in peak PSF CH1 : 270000 e-/s/pixel Qsat >75 ke- QEmin > 0.5 QEmean > 0.55 [Max/Nominal] Pixel read out rate [100 / 60] kpix/s Max flux integration time 10-s in Nominal Science Observation Nominal range of Science Data Rate is [3,5-7,0] s. Up to 1000-s for calibration mode (TBC) Calibration modes are TBD. RON <20 e- rms In CDS mode at Nominal Pixel read out rate and nominal output number. Iobs max CH0: 30 e-/s/pixel CH1: 50 e-/s/pixel DSNU 60% (pixel to pixel Iobsc dispersion) Flux linearity < 3% Over range CH0 : 0,5 ke- to 70 ke- Over range CH1 : 0,5 ke- to 70 ke- Number of output (operational) 1 Pending on the window mode capability that allow reading any adressed pixel window on the array. Additional output for redundancy shall be available. Cross-talk (inter-pixel capacitance) < 2% (goal 1%) To each neighbouring pixel. Operability >95% in an operationnal window See size of Spectrum Window in Window Mode. Window Mode CH0 270 x 64 pixels CH1 100 x 64 pixels Power Consumption < 2 mW At nominal pixel rate and nominal output number. Detector operating temperature 42-K
Figure 138: AIRS-FPA-Detector Requirements. The main justification for the detector number and format is directly derived from the optical design where the 2 channels are using 2 independent optical paths with spectral length of CH0: 270 pixels and CH1: 100 pixels (with positioning margins). The 64 pixels in spatial dimension are justified by the science performance simulations showing that more than 50 pixels (+/-25 pixels around the spectrum) is sufficient to monitor the background and the dark current. The choice to go for 2 detectors is mainly justified by the broad range of wavebands covered, having 2 detectors allows for detector optimization per Channel in terms of coating, FWC and dark current (which varies strongly with λ). AIRS detector noise budget allocation is derived through System Performance analysis (EXOSIM simulated post processing) leading to the following requirements: • Detector Read out noise CDS of 20 e-rms (15 e-rms as a goal). • Detector Dark Current CH0 = 30 e-/s/pixel • Detector Dark Current CH1 = 50 e-/s/pixel For the baseline design flux and spectrum window size the following parameters are allowing to read in each channel (0 and 1) the spectrum window before the brightest pixels saturate: • FWC(@3% linearity) = 70 000 e- • Pixel read out rate: 60 kpix/s
Page 156 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• Windowing mode feasible through 1 output • Read / Reset / Read mode feasible for read out strategy (in order not to lose half of the cycle of observation) The functioning point is selected to allow the time to read a full frame with 2 samples (CDS) in 288 ms (CH0) and 107 ms in CH1. This read time is less than the time to saturates the pixels (respectively 311 ms in CH0 with max flux of 225 000 e-/s/pix and 259 ms in CH1 with max flux of 270 000 e-/s/pixel). The detector type for the range of waveband and performance (QE, low dark current, low flux linearity, readout noise and format) described in the requirements Figure 138 is restrained to HgCdTe (Mercure Cadmium Tellurium) photosensitive material hybridized by Indium bumps on a Read-Out Integrated circuit. Two options are considered for the detector implementation: • Baseline using Teledyne H1RG (CH0 Standard / CH1 NEOCam-type) detector • Option for a European MCT detector
12.3.3.1 Baseline US AIRS-Focal Plane Assembly Detector Currently the baseline detector for CH1 is Teledyne Imaging Sensors (TIS) H1RG with 10.5um cut-off developed in the frame of the NEOCam program. The baseline CH0 detector is a standard H1RG product with cut-off at 5.3 µm. McMurthy et al. provides a detailed summary of this device’s performance and a summary is presented in Table 39.
Parameter CH1-NEOCam CH0-Standard Comment (H1RG) (H1RG 5,3 µm)
Wavelength of 10.6um 5,3 µm Cut-off from 9.3 µm to 10.6 µm operation demonstrated at low temperature (30- 35 K).
Array format 1024x1024 1024x1024 H1RG ROIC used
Pixel size 18 µm 18 µm H1RG ROIC used
Linear well capacity > 55ke > 65ke 55ke to 75ke demonstrated
Dark current <1e/s/pixel@40K <0,1e/s/pixel@40K For operability of more than 90% pixels.
Read noise (CDS) 20 e rms 15 erms CDS
Quantum efficiency >65% >70% Demonstrated on process evaluation chips without any AR coatings
Operating temperature <42K <42K Defect number strongly related to temperature
Power dissipation <3mW <2 mW
Max. frame rate (NDR) >100 kpix/s >100 kpix/s Window readouts
Cross-talk (IPC) <2% <2% To each of the neighbouring pixels Table 39: Baseline AIRS Detector Requirements Currently tested devices have a cut-off wavelength between 9.3 and 10.6 µm. When operated at a temperature of 42K (Figure 139) the device has an operability of 94%, with a dark current of 16 e/s for 90% of detectors pixels. The Quantum Efficiency without AR coating is measured in excess of 60%.
Page 157 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 139: Dark current (left), Readout noise (Fowler-1) (central) for Teledyne MCT detector selected for ARIEL baseline (from McMurty et al., Optical Engineering 52,9,2013). Device is presented on the right panel. The device Technology Readiness is estimated to be at level 9 for Chanel 0 and 6 for Channel 1, while the control-electronics SIDECAR ASIC and ROIC being at TRL 9, as these are effectively the same technology used on NIRSPEC on JWST. For ARIEL, the cut-off wavelength of the photosensitive layer can be adapted by modifying the stoichiometry of the Hg1-xCdXTe Cadmium X proportion during the epitaxy layer growth. Decreasing the lambda cut-off leads to decrease in the dark-current and increase Full Well Capacitance for SFD ROIC. The US detectors with 9.3 to 10.6 µm lambda cut-off (around 35-40K) represent state of the art technology with low dark current and Capacity well > 70 ke- SFD ROIC.
12.3.3.2 Option European AIRS-Focal Plane Assembly Detector There are specific efforts underway within Europe (CEA/LETI) to develop a detector (640x512 or 320x256, 15um pixels) with cut-off at 8.2um @45K, with both source follower per unit cell (SFD) and CTIA based ROIC aimed at meeting the requirements for ARIEL IR Spectrograph within Europe. Using two different AR coatings for the two channels is being explored. CEA (LETI+SAp) and Sofradir are undertaking activities under CNES funding to produce Photo-sensitive detectors applicable to ARIEL needs. The different steps of this works covers 1. Material: Liquid Phase Epitaxy MCT on CZT substrate (with expected Cut Off Wavelength Wco= 8,2 µm @ T=45K) [Performed in November 2016] 2. Design and production of Photovoltaics layers: Design of PV map for PV technology of FPA TV 640x512 @15µm pitch with standard n/p PV MCT technology. 2 types of detectors: compatibility for SFD or CTIA CMOS ROIC [Performed in January 2017] 3. TAP (Tip Auto tests @ wafer level) at 40-K [Performed in February 2017] 4. Hybridization on SFD and CTIA ROIC [To be performed in April 2017] 5. Characterization of hybridized detectors at CEA/Sap [To be performed in May 2017]
The photosensitive selected Material is Hg1-xCdxTe with Cadmium composition x selected to have lambda cut-off λc = 8,2 μm at 45K. The diode technology used is n/p diode, as this is a mature technology in Europe with space qualification heritage, with a format of 2D planar detector array of 640x512 pixels, and pixel pitch 15 μm. The objective of the study is to demonstrate maturity and performance of such a detector in terms of Quantum efficiency, Read out noise and low dark current with n/p technology. The dark current performances with n/p technology are known to be less good than p/n diode used in the US detectors, but they are expected (from models based on actual measurements) to be in-line with the performance needs. A set of Liquid Phase Epitaxy growth HgCdTe over CZT substrates were achieved to the required lambda cut in November 2016. The detector hybridation and AR coating implementation is planned to happen in April 2017 and the performance tests are planned in May 2017.
Page 158 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Several batch have been realised with lambda cut-off around the required λc = 8,2 μm at 45K, that will allow characterizing the dark current as a function of lambda cut-off (open trade off for science, the various detector will allow probing from 7.6 to 8.5 µm lambda-cut off). The different batches available are illustrated in Figure 140 showing lambda cut-off at ambient (300 K) and at operational temperature. The Red curve 16980 correspond to the required λc = 8,2 μm at 45K case.
Figure 140: Lambda cut-off of manufactured batch of European n/p MCT material The Quantum efficiency is expected, with a lambda cut off of 8.2 µm at 45 K, to be >0.50 over the full waveband. The dark current depends on two parameters: diode technology (n/p vs. p/n) and dark current regime (diffusion vs. generation-recombination GR). For a 8.2µm cutoff wavelength, n/p diode and diffusion regime, the expected dark current at 45K is 7.8 e-/s per 15µm pixel. One can typically expect a diffusion dark current 10 times lower in p/n technology (as can be seen from H1RG performance). A purpose of the CNES R&D activity is also to evaluate the potential limitation, if any, due to generation-recombination regime. The CEA 12.5µm cut-off devices (developed under ESA contract 4000106410/12/NL/HB and CNES contracts 127094/00 and 130731/00) are diffusion limited down to about 42K. The CEA 2.1µm cut-off devices (ESA contract 4200022949/10/NL/CP) start being dominated by GR regime at temperatures below 150K, so the CEA Ariel devices will be GR dominated in between. The GR dark current will degrade dark current performance compared to the diffusion dark current. Additional developments are currently addressing this issue (ESA contract 4000113066/15/NL/RA and CNES contract 160314/00). Moreover, ESA will place a contract with CEA that will specifically address the development of detectors for Ariel. This contract is currently in negotiation phase (AO/1-8584/16/NL/BW, negotiation to take place 16/02/2017).
Relevant performances, especially Dark Current, Read out noise and Quantum Efficiency will be characterized in 2017 at CEA-SAP on hybridized circuit of both types: SFD and CTIA. ROIC SFD type for 2Kx2K with pixel pitch 15µm for space application is currently in development at SOFRADIR (contract ESA NIRLFSA 2Kx2K 4000118961/16/NL/BJ). Current performance of this ROIC reach a readout noise of 11 e-rms CDS for a FWC@3%lin of 60 000 e-. CTIA ROIC for space application if retained would be developed under CNES funding to reach TRL 6 in 2019.
12.4 AIRS OBSERVATION STRATEGY The observation strategy and its impact on the design of the AIRS electronics system (AIRS-FPA-CFEE and AIRS-DCU) have been analysed from a top-down approach from the science needs currently
Page 159 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 expressed. A summary of those needs are recalled here and the baseline Data Processing requirements used to size the full AIRS detection chain are derived. R-AIRS-P-490 requirement defines the time scale of the physical events that are being measured ranging from 90-s for the bright targets to 300-s on faint targets. Those timescales are required to be oversampled from the Science Data provided from the Spacecraft. We define here the notion of Science Data Period: the period at which the Data from the Spacecraft are sampled and sent to the Science Ground Segment in order to: • Sample correctly the event (>10 samples per period of 90-s) • Provide reliable samples (the quality of the sample shall be adequate with required science needs, mainly less than x% of samples shall be corrupted by glitches) • Fits within the data volume allocated to the AIRS according to R-AIRS-D-1750 (14 Gbits/day). For AIRS this will be justified in the current section and translated into: • Science Data Packet shall be produced at a rate of 1 per period of [3.5 – 7.0] s. • Science Data Packet shall contain the optimal information of all the intermediate samples collected between the production of 2 Science Data Packet. • For all pixels of the spectral windows for CH0 and CH1 Science Data Packet shall contain at minimum pixel information (on 16-bits) and quality criteria of this information (on 8-bits for baseline, could be downgraded to 1-bit flag for above threshold samples). This means that all the intermediate data taken at higher rates need to be processed on-boad in order to provide synthetic and reliable information on the timescale below the Science Data Period (<[3.5-7.0]s). Depending on the brightness of object, during one Science Data Period, multiple detector data need to be merged: • In the Case of Brigthest objects, only Correlated Double Sampling (CDS) read out can be performed before saturating, and the multiple CDS (typically taken at 0.300-s in CH0 and 0.100 s in CH1) needs to be merged over the Science Data Period of [3.5-7.0] s • In the case of Faint to Bright objects multiple Slopes with several Sample Up The Ramp (UTR) can be acquired during one Science Data Period. Data processing must ensure processing of the slope from the UTR and then merging of the numerous slopes in 1 slope information + quality criteria. • In the case of the Faintest objects, the single slope without reset and with multiple SUR can be acquired over several Science Data Period. The on board processing shall ensure processing of the slope for each Science Data Period with a quality criteria. • In each case as the mean or chi² of the data are computed, deglitching by rejecting samples which have a value above a threshold is an option (though not implemented currently, TBD with science SGS team) Figure 141 summarizes the data processing steps implemented in current AIRS baseline.
Page 160 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
AIRS Data Flow
Multi-CDS Multi-Slope Single-Slope
Brigthest Brigth Normal faint faintest
225000 e/s/pix 120000 e/s/pix 20000 e/s/pix 10000 e/s/pix 1000 e/s/pix
Frame Rate [0,100 to 0,300] s
AIRS-FPA: Depending on the object, the detector is read in CDS or Sample Up the Ramp mode to maximise dynamics of detector.
AIRS-DCU AIRS-DCU Case 1: Multiple CDS are merged Case 2: Slopes are estimated from over 1 Science Data Period (1 SUR and then the multiple slopes mean CDS and 1 σ² per pixel for are merged over 1 Science Data CH0 and for CH1). Period (1 slope and 1 χ² per pixel for CH0 and for CH1)
Frame Rate [3,5 to 7] s
Science Data Period CH0 Science Data Period CH1
CH0 spectrum window: 270 x 64 pixels CH1 spectrum window: 100 x 64 pixels
Data are transfered with time sampling SDP-CH0 and SDP-CH1 from AIRS-DCU to ARIEL –ICU. Data are images of 270x64 pixels for CH0 and 100x64 pixels in CH1 with 1 value per pixel (16 bit) and 1 Quality criteria per pixel (8-bits)
Figure 141: Summary of AIRS Data Flow and Data Processing Steps for full range of objects.
Page 161 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Multiple simulations (summary in this section) have been performed in order to set the requirements on the AIRS-FPA-CFEE and the AIRS-DCU by looking at the full range of objects brightness in both Channels and the optimal data processing implementation over the full AIRS electronics chain. The criteria for optimal design have been: • Maximise the efficiency of the photon collection (meaning that we want to maximize the number of photon converted into electrons during observations to be collected into information sent to the ground) • Maximise the dynamics of the Full Well Capacity of the detector used for data collection • Fits within data volume allocation • Electronics implementation: the 2 detectors shall have the same pixel read out frequency (60 kpix/s from detection design section). A 3.5-s science data period is the maximum allowed to fit within the 14 Gbits/day of data volume. Once the minimum Science data Period is fixed from the Data volume allocation, we can estimate the range of Science Data Period needed to be able to observe the full range of objects by defining in each case a read out strategy maximising the use of the detector dynamics and the observation efficiency (i.e. matching an integer number of CDS or slopes per Science Data Period to ensure having always the same amount of information per Science Data Period). Figure 142 to Figure 146 present results from optimisation made over the full range of objects showing for each objects flux what is the Science Data Period required, to implement a read strategy with N slope per Science Data Period and M read per slope (M = number of window read time +1) and the associated use of dynamics of the FWC and efficiency of the strategy (the last 2 being the criteria of optimisation).
Figure 142: Science Data Period as a function of Object flux in CH0 (optimal strategy).
Page 162 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 143: N# of slopes per Science Data Period as a function of Object flux in CH0 (optimal strategy).
Figure 144: N# of read per slopes as a function of Object flux in CH0 (optimal strategy).
Figure 145: Optimal use of detector dynamics as a function of Object flux in CH0 (optimal strategy).
Page 163 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 146: Efficiency of detection process as a function of Object flux in CH0 (optimal strategy). Those analysis (similar performed in CH1 not reproduced here) show that with a minimum Science Data Period of 3.5-s, a maximum Science Data Period of 7.0-s is required. The top-down approach evaluation of the science needs leads to the following design selected for the on-board processing: • For each targeted observation, DCU shall control a specific read out strategy at AIRS-FPA-CFEE level that is defined on ground by users and provided by TC through the ICU. • AIRS-DCU shall produce Science Data Packet at a period ranging from 3.5-s (maximum to reach the data volume allocation) to 7.0-s (needed to cover the full dynamics of targets). • This strategy can be: Correlated Double Sampling (CDS) or Sample Up The Ramp (UTR) • The optimal strategy, depending on the object flux can go from • CH0: 14 CDS per Science Data Period (CH0) for brightest to 25 Sample Up the Ramp for the faintest objects • CH1: 19 Slopes (3 samples per per slope) Science Data Period for brightest to 65 Sample Up the Ramp for the faintest objects • DCU shall merge all information collected during a Science Data Period for each channel and produce a pixel level information (data on 16-bits + quality 8-bits) • Merging of data shall be optimum in terms of the following constraint: optimal merging of information in data and quality at individual sample level / minimize DCU complexity for on-board processing. • Transparent mode (for calibration or debugging) providing full data visibility shall be available. • De-glitching can be implemented (optional for the moment) as all information (mean and sigma²) are available to identify samples that are beyond a given threshold (for sigma clipping). • At that stage, data compression is not required. • It shall be possible to control the detectors in Read/Reset/Read per pixel mode in order not to lose half of observation cycles (see [RD37] section 4.3). This design implementation is presented in section 12.5 AIRS Detection Chain Electronics and in the ICU Design Description [RD24].
Page 164 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
12.5 AIRS DETECTION CHAIN ELECTRONICS
12.5.1 Electrical System Overview The AIRS detection chain is defined as the functional sub-assembly of the AIRS that is necessary to detect the AIRS spectrometer images and to pre-process scientific data to fit with the AIRS telemetry allocation. The AIRS detection chain electrical system encompasses various sub-systems: • The two detector array assemblies each being composed of a detector circuit for infrared light to current conversion hybridized on top of a Silicon readout circuit for current to time multiplexed video signal conversion (the ROIC for ReadOut Integrated Circuit) • The AIRS-Cold Front-End Electronics that is in charge of the Sensor Chip Assembly interface signals handling whose function is to provide it with an electrical interface that is compliant with the instrument environment (harness equivalent load, EMC perturbations, …). For optimum performance the AIRS-FPA-CFEE shall be located close enough to the detector assembly in order to limit harness length between the two sub-systems. CFEE will therefore be operated at a low temperature. (55-K) in the baseline case. • The AIRS-DCU (warm front-end electronics) that is interfacing with the detector through the AIRS-FPA-CFEE implements all the functions that are necessary to operate the detector and that are complementary to functions implemented in the AIRS-FPA-CFEE. The complexity of the AIRS- DCU will depend on detector and AIRS-FPA-CFEE types. While mainly based on analog electronic functions the AIRS-DCU relies as well on low-level digital functions i.e. for detector and analog electronics timing generation and scientific data pre-processing before transmission of scientific data frames to the instrument data processing unit for data post-processing such as loss less compression. Additionally the AIRS-DCU provides closed loop thermal control functions for detector fine temperature monitoring and regulation. An overview of such a detection chain is illustrated in the next figure as defined for the AIRS instrument.
Figure 147: Architecture blocks of the AIRS detection chain.
12.5.2 Electrical system description The full electrical architecture is depicted on Figure 148 with the AIRS detector assembly on top, the AIRS-CFEE just below and at the bottom the AIRS-DCU. The AIRS-DCU, though a sub-system of the AIRS is presented in the current baseline as part integrated on a dedicated slot of the AIRS ICU. This baseline implementation is fully described in the ICU Design Description [RD24].
Page 165 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The description in the AIRS DD is focusing on alternative DCU design for both options US or European detectors. The AIRS-DCU is partially internally redundant in order to maximize the reliability figure by providing to each detector a redundant interface to both main and redundant part of the ICU. This is achieved by implementing non-redundant to redundant low-level functions (to keep reliability as high as possible) for interfacing with redundant functions with lower intrinsic reliability figures such as the ‘detector data pre- processing’ function that required a FPGA associated to a memory device. The AIRS-DCU is interfacing to the DPU function of the ICU for scientific & housekeeping data transmission and configuration command reception. It interfaces with the unit power supply (PSU) that is providing secondary voltage lines including a low noise line for detector and AIRS-FPA-CFEE feeding which is mandatory for adequate control of performance at detector level. Reliability of the full Detection chain is estimated to be >97% (see detailed reliability analysis in the AIRS Detailed Design Description [RD41]).
Figure 148: Full electrical Architecture of the AIRS detection chain.
12.5.3 AIRS-FPA Cold Front End Electronics
12.5.3.1 Baseline Detector CFEE In the detector baseline option the AIRS-CFEE is naturally based on the Teledyne detector companion Sidecar ASIC. This device encompasses most of the functions that are required to handle the electrical interface signals of the H1RG detector. It implements in particularly digitization functions (over 12 or 16 bits) and the clock sequencer function. The interface to the warm front end electronics is therefore mainly digital: serialized scientific data & ASIC configuration command. Other signals are the master clock that is used by the ASIC for timing generation and voltage lines that are used internally to bias the devices and provide the detector with adjustable low noise biases. In the selected baseline with two
Page 166 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 detectors (1 per channel), two Sidecar ASICs are required to handle the electrical interface of both CH0 and CH1 detector assembly. The sidecar ASIC is a very versatile device that has to be configured according to the specific need. For AIRS the considered configuration is a follow (per detection chain): - 1x preamplifier - 1x 16-bit ADC - Microcontroller / clocking - 3x LVDS outputs (1x serial data / 1x serial clock / 1x serial data strobe) - 2x LVDS for master clock input & data input
12.5.3.2 Option European Detector CFEE In the European detector alternative the AIRS-FPA-CFEE has much limited functions since no ASIC currently exists with a complexity and technical maturity as high as the Sidecar. The implemented functions are mainly the detector signal buffering since the internal amplifier is not designed to drive long harness stray capacitance in order to limit power dissipation. To limit noise pick-up along the harness the cold front-end electronics provides differential signal outputs (1 per detector assembly). Additionally passive filtering is implemented in order to feed the detector with clean enough bias lines. The block diagram for both options of ROIC, SFD or CTIA are given in the Figure 149.
Figure 149: CFEE function for the European option in the SFD case (left) and CTIA case (right).
12.5.3.3 CFEE High level Requirements The Table 40 is a preliminary list of AIRS-FPA-CFEE design requirements from detector specifications. Even if numbers are subject to variation this table is useful to assess criticality of this sub-system: Detector specification Parameter Value Comment R-DET-001 Pixel readout rate 60 kpixels/s R-DET-002 Readout noise 15 e- rms R-DET-003 Full Well Capacity 80 ke- @ 1 V (TBC) Readout electronics requirements R-CFEE-001 Bandwidth 1 MHz From (R-DET-001) R-CFEE-002 Noise (e-) ≤ 5 e- rms From (R-DET-002)
Page 167 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
R-CFEE-003 Noise (µV rms) ≤ 62 µV rms From (R-CFEE-002) & (R-DET-003) R-CFEE-004 Noise density ≤ 62 nV rms From (R-CFEE-001) & (R-CFEE-003) R-CFEE-005 A to D dynamic range (if applicable) Table 40: CFEE preliminary design requirements.
12.5.4 On-Board Data Processing As discussed in section 12.4, due to large observation photon flux dynamic range two different detection modes of operation shall be implemented. The on-board data processing is in turn depending on the readout mode and thus three modes are defined (multiple CDS / Multiple slopes SUR / Single slope SUR). The Figure 150 illustrates these three modes of operation by representing the pixel level signal.
Figure 150: 3 different modes of detector read out depending on the object flux. In order to maximise the efficiency of the observation while in operation, it is mandatory that the detector can be operated: • In Read/Reset/Rest per pixel for the CDS mode (not to lose half observation cycle) • In Non Destructive Read Out mode for the Sample Up the Ramp
The Table 41 summarizes the applicable requirements for the different modes (it is supposed the processing function is following a de-multiplexing function): Detector sampling Observation Rate Processing mode - Reference to Signal samplings subtraction CH0: 280 ms (CDS) Ultra-bright sources Multiple CDS CH1: 100 ms - CDS averaging over SDP - Variance estimator of averaged CDS - Ramp fitting (Least Square Fit) - Fit estimator (chi squared) Bright sources Multiple up-the-ramp - Valid ramp slops averaging over SDP (TBC) - Variance estimator of averaged ramps - Ramp fitting (Least Square Fit) Faint sources Up-the-ramp - Fit estimator (chi squared) Table 41: Processing function for the 3 modes of observation. An estimation of the number of operation to be performed while observing faintest and brightest sources is given in the Table 42 for both CH0 and CH1 detectors.
Page 168 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Up-the- Source Slope Operations / SDP Operations / Detector ramp type average detector second samples (s) Faintest 25 1 5 339 520 7 762 788 CH0 Brightest 2 (CDS) 14 1 296 000 4 324 000 Faintest 65 1 5 049 600 7 721 371 CH1 Brightest 3 19 0 652 800 4 163 200 Table 42: Sizing case (brightest / faintest objects) in terms of operations. These number of operations show that the pre-processing could be implemented either as a software running in a standard LEON processor for better flexibility or a FPGA (ACTEL RTAX2000D – featuring 64x DSP Mathblocks) for better efficiency but with less flexibility. The AIRS data rates are given in the Table 43 for both raw (output of analogue to digital converters) and pre-processed scientific data: Raw data rate
Detector Tframe (ms) Coding (bits) Pixel number (bits/s)
CH0 280 16 17 280 0 987 428 CH1 100 16 06 400 1 024 000 Total 2 011 428 Processed data rate SDP min (s) (bits / s) CH0 4 16+8 17 280 103 680 CH1 3.5 16+8 06 400 043 886 Total 147 566
Table 43: AIRS Data rate. In this table it is assumed the pixel signal and the scientific data are 16-bit codded while the fit estimator for scientific data is 8-bit codded.
12.6 AIRS AIV/AIT PHILOSOPHY The AIRS instrument philosophy is defined in order to be in-line with the higher level Ariel schedule and philosophy, while enabling instrument verification and qualification.
The following models will be produced:
AIRS STM : Structural and Mechanical Model of AIRS (+ S/W Models) AIRS AVM : Avionics Model (limited to electrical sub-unit / wire-table model) AIRS PVM : Performance and Verification Model (equivalent in as-built to QM) AIRS EQM : Engineering and Qualification Model (not delivered) AIRS PFM : Proto Flight Model
One of the key needs identified from the AIRS philosophy is the number of detectors that would be needed for the current model approach.
Page 169 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The number of detectors derived from this philosophy is: • AIRS STM detectors chain: ≥ 2 Mechanical samples • AIRS AVM detectors chain: ≥ TBD (1 Bare ROIC? Depending of AVM standard) • AIRS PVM detectors chain: ≥ 2 Engineering Grade samples. • AIRS EQM detectors chain: ≥ 2 Engineering Grade samples. • AIRS PFM detectors chain: ≥ 2 Flight Grade samples (+TBD for the spare philosophy).
Page 170 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
13 FINE GUIDANCE SYSTEM
13.1 FGS KEY REQUIREMENTS The Fine Guidance System (FGS) main task is to ensure the centering, focusing and guiding of the satellite, but it will also provide high precision astrometry and photometry of the target in the visible and additionally a low resolution near-IR spectrometer. The term “FGS instrument” will be used to refer to this combined functionality of both guidance and science channels. In particular, the data from the FGS will be used for de-trending and data analysis on ground. The FGS function uses star light coming through the optical path of the telescope to determine the changes in the line of sight of the ARIEL instrument. The attitude measurement is then fused with the rate information from the star tracker, and used as input for the control loop stabilizing the spacecraft through the high performance gyros. The Fine Guidance Sensor is a critical piece of equipment as it is an important contributor for the AOCS RPE performance in terms of the achievable single-star centroiding accuracy. To meet the goals for guiding and science, four spectral bands are defined: • FGS – 1 : 0.8-1.0 µm, • FGS – 2 : 1.0-1.2 µm, • VISPhot: 0.50-0.55 µm and • NIRSpec: 1.25-1.95 µm spectrometer, R>10.
Two FGS channels (FGS-1 and FGS-2) deliver guiding information to the S/C . In case of failure in one channel the remaining channel provides redundancy. The spectral bands are selected using a series of dichroic mirrors. Data from FGS1 and FGS2 will be used for photometric analysis too. The pointing stability between the instrument LoS and the science target with the FGS in the control loop shall be controlled to ensure compliance with the photometric stability requirements. Further details of the functional, performance, interface and environmental requirements for the FGS are contained in the FGS detailed design description document, [RD29].
13.2 FGS DESIGN ARCHITECTURE The system is composed of the optics box and the electronics box. The optics box is situated at the instrument optical bench (IOB) containing cryogenic optics with two detector modules at ~50 K and the cold front-end electronics (CFEE). The electronics part - FGS Control electronics Unit (FCU) is accommodated in S/C service module at a temperature of 270-300 K. The modules of the FCU control and read the detectors and carry out the data processing for a centroid and photometric calculation. FGS systems are independent from the spectrometer instrument, thus have their own power and data interfaces with the spacecraft. The FGS is also involved in the focusing of the main telescope. This will be done using images from the two detector arrays. The procedure will be controlled interactively from ground.
Figure 151 below depicts the overall FGS system layout.
Page 171 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
FGS@IOB
Figure 151: FGS Top-level architecture Two architectural options were considered, one with the CFEE placed in an independent box and one with the CFEE placed in the optical box with the detector. Choosing Teledyne detector as a baseline design, the second option is baselined and is presented in this document
13.3 FGS OPTICAL MODULE DESIGN BASELINE
13.3.1 Optical Design The FGS optical module has been designed with the following basic assumptions: • FoV – max usable on sky FGS FoV is 25.2 arc sec, what correspond with ±0.19 deg (internal FGS’s FoV)11, • Spectral bandwidth: 0.5-1.95 µm is split into four bands, • Detector: MCT FPA with minimum array and pixel size (18 µm for MCT) and ~1024x1024, H1RG TELEDYNE with its SIDECAR is baseline, European detectors with pixel size 15 µm are optional, • Low distortion (< 1% level over FoV), • Minimum bin/star image spread FWHM: 6x6 pixels • Able to achieve centroiding to 1/10th of a pixel level • Input WFE: 250 nm rms = telescope diffraction limit @ 3 µm + allocation for dichroics
As a baseline telescope system for the FGS , the off–axis Gregorian mirror telescope is proposed (see Figure 12-3). The main parameters of the telescope are: focal length = 500 mm, F-number = 25.
Figure 152: Scheme of the baseline 0ff-axis Gregorian telescope
11 This is oversized on the required FOV of 17” (see section Error! Reference source not found.), providing some margin for beam clearances at components.
Page 172 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
13.3.2 Dichroic System Definition The four-channel FGS configuration needs to use three dichroic beam-splitters and band pass filters for VIS-Phot and FGS-1 channels. The beam folding geometry allows the detection of the signal from two channels on one 1024x1024 pixel detector. The first detector works with FGS 1 and VIS-Phot channels. The second detector works with FGS 2 and NIRSpec channels. In the beam of NIRSpec channel a prism is placed to form a Low Resolution Spectrometer (LRS). This concept of the dichroic system with the prism is shown in Figure 153.
Figure 153: Scheme of FGS Dichroics After analysis of polarisation effects was stated that the dichroics should work for incident angle 200 or less to avoid polarisation effect. The scheme below presents the baseline optical design solution allowing for this assumption.
Page 173 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 154: Baseline optical design for FGS To obtain the equal optical path for the Detector 1 in two parallel channels VISPhot and FGS-1, a negative lens for VIS-Phot and positive lens for FGS-1 are used. The flat plane of the negative lens in the VISPhot channel is used for coating of the long pass filter > 0,5 µm. On the back side of dichroic D4 is long pass filter >0,8 µm The dichroic mirror D4 is combined with positive lens in one holder. Materials: the prism is made of SF11 , dichroic and lenses are made of BK7. The telescope mirrors will be made of aluminium. To obtain the equal optical path for the Detector 2 a lens is incorporated on back side of the dichroic D5. Additional optical elements in the photometer channels direct the FoV in FGS channels and photometry channels onto the two detectors. It must be taken into account as a design restriction that applying two long detectors limits the parallel channels separation to 11.5 mm between their optical axes. However, the chosen fold geometry allows two channels to fit on one detector.
13.3.3 FGS Stand-alone Optical Performance Predictions The main parameters for every channel are presented in the table below. VIS Phot FGS-1 FGS-2 NIR Spec Spectral range [um] 0,5 - 0,55 0,8 - 1 1,05 - 1,2 1,25-1,95 Focal length [mm] -630,40 -393,30 -500,00 -386,75 FGS entrance pupil dimensions 13,30 20,00 13,30 20,00 13,30 20,00 13,30 20,00 F# -47,4 -31,52 -29,57 -19,67 -37,59 -25,00 -29,08 -19,34 PSF dimensions [um] *1 - FWHM 26,5 19 30 21 46 32 48 40 PSF dimensions [um] *1 – 10% 44 34 52 36 80 53 90 75 PSF dimensions with introduced Wave front 28 83 24 56 - - 48 40 error – FWHM level PSF dimensions with introduced Wave front 113 146 106 105 - - 96 70 error – 10% level Pixel pitch[um] 18,00 18,00 18,00 18,00
Page 174 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
PSF dimensions in pixels 3,44 2,22 3,89 3,89 6,11 3,89 4,4-7,2 4,4-6,6 FoV [deg] *2 0,38 0,38 0,38 0,38 Full FoV illuminated area [mm] 4,21 4,24 2,61 2,56 3,32 3,32 2,59 2,94 Full FoV illuminated area [px] 234,11 235,56 145,11 142,22 184,51 184,51 143,89 163,33 0,0016 0,0016 0,0026 0,0026 0,0020 0,0020 0,0026 0,0023 Plate scale [deg/px] 32 22 32 86 70 70 55 39 Spectral resolution 18,5-25 Table 44: The main parameters for every channel *1 - Estimated from the Zemax plot - dimension refer to PSF first zero *2 - Input angle 25,2 arcsec , Output angle (input for FGS) 0,382 deg , Angular magnification: -54,6, data base on ARIEL-RAL-PL-TN-001_Baseline Telescope Prescription_Iss 2.0 (16 Jul. 2016)
PSF dimensions are estimated based on Zemax model – after normalization - FWHM and width at 10% of the maximum intensity. The data presented in the table above are calculated for elliptical aperture and for both major and minor axes the dimensions figures are shown. Spot diagrams (without inclusion of the input telescope assembly Wave Front Error) for every channel of the FGS are shown below.
Figure 155: FGS system spot diagreams: VISPhot (upper left), FGS-1 (upper right), FGS-2 (lower left), NIRSpec (lower right).
Page 175 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
In case of the FGS with the spectrometric channel LRS, the main design parameter is the requested spectral resolution which shall be R>10. A first approximation of the spectral resolution based on the spot diagrams shows that R=25 for 1,25 µm and R=18,5 for 1,95µm.
13.3.4 NIR-SPEC resolution comparison for FGS with and without 200 nm WFE.
. Figure 12-10 Spot diagrams for the wavelength 1,25um and 1,3 um without (left picture) and with (right picture) 200 nm WFE at the input
Figure 156: PSF cross-section for the wavelength 1,25um and 1,3 um without (left picture) and with (right picture) 200 nm WFE. As it is visible in both cases two spectral lines are resolved. WFE change the shape of the PSF and may slightly change the resolution from nominal 25.
Page 176 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 157: Spot diagrams for the wavelength 1,95um and 1,85 um without (left picture) and with (right picture) 200 nm WFE.
Figure 158: PSF cross-section for the wavelength 1,95um and 1,85 um without (left picture) and with (right picture) 200 nm WFE For the long-wavelength border of the NIR-SPEC channel, WFE does not have a substantial influence on the spectral resolution.
13.3.5 Complete Optical Performance Predictions To obtain a final optical performance of the FGS the two sources of aberration have to be taken into account: the Telescope and optical system of the FGS and (substantially mode influential) the main ARIEL Telescope assembly. Predictions of the FGS optical performance have been based on wave front analysis and PSF spot diagrams calculated by ZEMAX. Calculations were obtained for each of the four channels. In all cases, the preliminary design and optimisation of only the FGS optical system allows to obtain the diffraction limited performance in every channel, but when optical quality analysis with the telescope Wave Front Error (WFE) was done the quality of the image dropped significantly (as expected). The Telescope WFE was implemented in ZEMAX using Zernike polynomials based on the scaling up of the STREGO primary mirror as detailed in full in [RD30]. The wavefront maps assumed from the telescope assembly (and common optics) at the pupil entrance of FGS is shown in the figure below. Calculation for a reference wavelength 0,55 um was done. After introduction of the wavefront distortion of the telescope, the Strehl ratio is of the combined FGS telescope system is 0.12. For comparison, the wavefront error for the FGS without introduction of the
Page 177 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 telescope distortion the Strehl ratio was 0.99. However, it has been shown that the FGS can still function correctly with such low Strehl ratios [RD30].
. Figure 159: Wave front maps with the Telescope Wave Front Error with elliptical aperture 13.5 x 20mm
Spot diagram dimensions are drawn against a reference square of 108x108 µm (6x6 pixels) and the Airy disc diameter (depicted as a black circle). Results of analysis are presented below for three wavelengths 0,55 µm, 1.0 µm and 1,95 µm.
Figure 160: Spot diagram (left) and PSF (right) for λ=0,55 µm including telescope assembly WFE
Page 178 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 161: Spot diagram (left) and PSF (right) for λ=1.0 µm including telescope assembly WFE
Figure 162: Spot diagram (left) and PSF (right) for λ=1.95 µm including telescope assembly WFE
13.4 MECHANICAL DESIGN Based on the optical design, mechanical dimensions of the dichroic beam splitters, the telescope mirrors and the MCT detector with accompanying electronics, the design of the FGS cold unit is shown in the figures below.
Page 179 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 163: FGS Optical Module mechanical design
The mirrors of the Gregorian off axis telescope are fixed in mounts which give the possibility to align the optical axis of the FGS telescope to the instrument telescope. The distance between mirrors is adjusted by the mount of first mirror. At the entrance of the FGS telescope the set of baffles are placed. Behind the telescope, the set of dichroic mirrors is placed. On one baseplate 3 holders of the dichroics are situated. The first holder (1) (see Figure 164) contains the plate with dichroic D3 and two mirrors. One mirror is at the same side like D3, second one on the opposite side of the plate of first mirror. On the back side of the first holder there is mounted also the D5 holder. The second holder (2) contains the dichroic D4 and lens for a compensation of the optical path. The third holder (3) contains the mirror and the prism. The beams are focused on two detectors H1RG Teledyne (4) with the SIDECAR ASICs (5).
Page 180 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 164: Top view of FGS with lid removed
The mass prediction for the whole optical unit of the FGS is of the unit is 1,6 kg ±20%, max with margin 2kg.
13.5 FGS DETECTOR SYSTEM
13.5.1 Baseline Detector Performance Parameters Two detectors are baselined for the FGS module. First detector is to cover the FGS-1 and VIS-Phot1 channels and the second one to cover FGS 2 and NIR-Phot2 channels. The range of wavelength is from 0.5um to 1.95um. The baseline detectors for the FGS module are Teledyne H1RG already developed for the NASA missions (such as JWST) and CFEEs from Teledyne (SIDECAR ASIC), given their very high TRL (9) and space heritage w.r.t. the present European alternative. The SIDECAR solution is the best one to drive properly the US MCT (HgCdTe) detectors and to save mass, volume and power at the same time. They can work easily down to the ARIEL required cryogenic temperatures (~60 K for SIDECARs and ~42 K for detectors). However, there are various development activities such as Large Format Near Infrared detectors for astronomy in Europe funded by ESA. The presented table below contains the FGS detector system requirements.
Page 181 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Value Parameter FGS1 + VIS- FGS2 + NIR- Comment Phot Spec Minimum format required. If independent detectors were to No. Pixels 1024 x 256 1024 x 256 be used, the minimum format of each would be 256x256 pixels ≥ 15 x 15 µm Other sizes might be acceptable but would require redesign in Pixel size ≤ 20 x 20 µm phase B. 1.05 µm to 1.95 Channel wavelength 0.5 µm to 1 µm µm VIS-Phot: 30x30 FGS2: 180x180 pixels pixels Window mode Two windows in each FPA in normal operation FGS1: 150x150 NIR-Spec: pixels 160x60 pixels Global reset, line by line reset, reset-read-read modes in full Reset and read modes - - frame and window readout modes Linear well capacity defined as the point at which the signal > 100 ke- (to be revisited) Linear well capacity response non-linearity exceeds 5%. This Allows 1s long integrations on bright targets ≤ 20 e- rms Read noise For a single CDS measurement
Dark current ≤ 1 e- /s/pixel At specified operating temperature (see below)
≥ 50% towards blue Over the entire operating wavelength range with suitable AR Quantum efficiency end, ≥ 60% towards ≥ 60% (goal 70%) red end (goal > coatings including fill-factor 70%) Operating temperature > 55 K (goal >65K) FGS arrays are passively cooled
Maximum readout speed ≥ 10 non-destructive full frame reads per Requirement is actually on each window – i.e. 10 Hz, for (frame rate) second each window
A minimum number of outputs required Number of outputs to meet the frame rate, read noise and power requirements
Maximum power ≤ 10 mW Detector and ROIC at the maximum readout speed consumption
Integration time 0 < t ≤ 1000 s No upper limit on integration time preferred.
Cross-talk (inter-pixel < 2% (goal 1%) To each neighbouring pixel capacitance) Operable pixels meet minimum QE of 45%, dark < 15e/s/pixel and read noise < 25e- (CDS) simultaneously. Any large clusters, bad columns or rows etc. should be avoided. Pixel operability > 95% (goal 99%) Arrays might be acceptable if the pixel operability is met over contiguous minimum number of pixels in a required pixel format (refer ‘array format specification’ above) Pixel response non- < 5% (sigma / mean) Over the operable pixels uniformity < 3% variation from a linear response Non-linearity (to be revisited) Up to 80% linear full well capacity
Table 45: FGS Detector Requirements and Baseline detector performance
Page 182 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
13.5.2 Other requirements The FGS 1 and FGS 2 windows are required to be readout at 10Hz window rate when they are in guiding mode, whilst the Photometric and spectroscopic channels (VISPhot1 and NIRSpec) are expected to integrate longer on the fainter science objects before reading out the regions destructively. Hence this will place a requirement on the ROIC of these devices to support multiple window readout architecture. The baselined H1RG detectors already support such readout schemes. A larger window size is required at the beginning mode of each observation when searching predefined star and for beginning of guiding procedure . When star reach the optical axis of the AIRS Instrument (defined in a specific location on the FGS detector), the window can be reduced and then a smaller window is sampled much more rapidly for the tight guidance control at 10Hz and to reduce the overall data volume.
13.5.2.1 Modes of operation of the detectors More detail is given in [RD29] on the different operational modes envisaged for the FGS. In summary these are: 1. Auto-calibration mode – The mode is run during commissioning phase or on request. 2. Searching target mode – The mode is run before an observation to find, located and stabilise the pointing onto the target star. 3. Measurement mode – operational science gathering and pointing stabilisation measurements . 4. Failure mode – in case when of failure of one detector to use alternate for guidance information.
13.5.3 Detectors available and performance parameters Apart from the Teledyne H1RG devices, there are some initial results available in the public domain from the detector development programs initiated and funded by ESA (LFNIR programs) within Europe. Both LETI/SOFRADIR and Leonardo, UK have shown progress towards developing good performing detectors. However, there are areas which require improvements such as QE, read noise, dark currents etc.12,13.
13.6 FGS CONTROL UNIT (FCU) HARDWARE
13.6.1 FCU Design Architecture The FGS has its own control electronics in the SVM to carry out all necessary communication, control and data processing tasks. It will drive and read the FGS detector electronics, establish a control loop with the spacecraft and deliver scientific data products. The FGS provides photometry measurements of the science target for its two detector arrays and it will support the star trackers with precise measurements @10 Hz of the target in its field of view. These data products will also be sent as science data products to ground. On-board compression will be used to reduce the telemetry. The FGS has few sensors and HK values. The precise overall telemetry contribution of the FGS depends on the parameter configuration, which is < 4 kbit/s if no imagettes are sent and < 10 kbit/s if e.g. 10x10 pixel imagettes are included.
12 Naidu Bezawada et al. Characterisation activities of new NIR to VLWIR detectors from Selex at the UKATC, Proc. of SPIE vol. 9154, 91540O, 2014 13 P-E Crouzet et al. First characterisation of the NIR European large format array detectors at ESTEC, Proc. of SPIE Vol. 9639, 9639OX-1, 2015
Page 183 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The FGS control electronics in the service module are independent from the spectrometer channels and the spectrometer ICU. They will consist of the following sub-units: • mechanical chassis: typical warm electronics box, with a total mass estimation of 5.5 kg • digital electronic boards for control and data processing • SpaceWire interfaces are used where applicable • power supply unit provides secondary voltages to the FGS components The overall scheme of the FGS control electronics unit is presented in the Figure 12-23.Two main blocks are depicted DPU (digital Processing Unit) and PSU (Power Supply Unit)
1
2
Figure 165: Scheme of the FGS control unit The FGS DPU is foreseen to work in cold redundancy concept as show in Figure 12-23 The SpW is redundant as well, but works in cold redundancy scheme. It means each SpW link (nominal and redundant) will be accessible from both main and redundant FGS DPUs.
13.6.2 DPU Block Scheme The Figure 12-24 shows the main concept of the FGS DPU subsystem. The central element of the DPU is Aeroflex GR712RC chip that implements Leon3FT processor. If special functions would be required the GR712RC could be replace by RTAX2000S/SL anti-fuse FPGA from Microsemi with synthetized Leon3FT processor.
Page 184 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 166: The FGS DPU block scheme – main concept The CPU will be clocked by 25MHz (TBC) oscillator. The 25MHz is proposed for CPU in order to minimize the power consumption, but can be increase in case the algorithms that will analyze images would require more computational power. The proposed FGS DPU is equipped with 16MB of SRAM operating memory. The memory comes from Aeroflex and if necessary its size might be increased. A 8MB MRAM memory from Aeroflex is proposed to store the Application SW. The MRAM was chosen mainly due to its very good radiation immunity. It size can be increased or reduced depending on flight software requirements. The PROM memory from Aeroflex is proposed to store the Start-up Software (Boot SW) for FGS DPU. The component was selected due to its good radiation immunity and the chip’s supplied by voltage 3.3V. Thanks that the number of power lines for DPU can be minimized. The assumed capacity of the PROM is 32kB but can be extended by adding additional chips if needed. The communication with the s/c will be realized via SpW using following drivers and receivers: UT54LVDS031LVUCC and UT54LVDS032LVUCC - both from Aeroflex. As mentioned the Leon3FT processor is proposed for the FGS DPU. The processor is widely used in the ESA space projects. There are several real time operating systems that are qualified for space and that runs on this processor – i.e. RTEMS OS. Although at the current stage the FGS DPU computation requirements are not well defined, it can be assumed that the above architecture shall be sufficient for this type of application. The concept of FGS DPU based on GR712RC is used in different space applications. The design or technological risks associated to this concept is very low. The design concept presented in the chapter 13.6.2 is similar to development made by CBK for SWI instrument for JUICE mission, where CBK is responsible for the DPU and PSU units. Similar architecture was also used by CBK in Proba3 project.
13.6.2.1 DPU voltage requirements For FGS DPU architecture three voltage lines are necessary: 1.8, 2.5 and 3.3V. An additional SpW voltage might be required.
Page 185 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
13.6.3 FGS DPU Power and Mass Budget
13.6.3.1 FGS DPU Power Budget Since the project is at a very early state an experience from previous space project (f. e. STIX/Solar Orbiter, ASM, or SWI/JUICE) has to be taken into account for calculating the FGS DPU power consumption. The worst case operational average power is assumed to be 3650mW, see [RD29] for details.
13.6.3.2 DPU Mass Budget The draft of the FGS DPU mass can be estimated based on the architecture proposed above. The result of these estimations is given in Table 12-5. Since the size and number of PCB layers are not known it was assumed that its mass will be around 70g. The mass of PROM, MRAM and SRAM memories are not given in the datasheets thus it was estimated based on other components with similar package. Finally the mass of connectors, analogue elements and other necessary components is uncertain as well, but it was assumed to be within 40g. The mass of the FGS DPU shall be around 157g. Details can be found in [RD29].
13.6.4 PSU design It is a standard Power Supply Unit board hosting DC/DC converters with a number of secondary sections needed to support the adopted cross-strapped and redundant configuration. It is in charge of collecting currents, voltages on secondary outputs and temperatures HK (internally to the Unit, exploiting the SPI HK I/F). The Unit consumption monitoring and switching on/off is in charge of platform by means of HPC commands. The PSU is mainly composed of three sections (refer to Figure 12-25): 1. Power conditioning section, performing the following tasks:
- DC/DC conversion: main DC/DC for the generation of the +5V to be distributed to the other boards, Aux DC/DC for internal logic powering, HK DC/DC for powering the HK section for the acquisition of voltages/currents/temperatures HK; - Inrush current limitation; - Polarity inversion protection; - Power-on sequence generation; - Unit power-on reset generation; - EMI (Electro-Magnetic Interference) filtering.
2. Power distribution section hosting Output Power Controllers (OPC), which implement overcurrent and overvoltage protection capabilities on the +5V and +28V voltage/current distribution EMI- filtered lines (+28V only to the Telescope Control Units, N and R);
3. HK acquisition section with three 12 bits ADCs for voltages, currents and temperatures measurements, controlled by the processor via SPI (Serial Peripheral IF).
Each electronic board, is basically supplied by a main voltage level of +5V protected for overvoltage and overcurrent and locally on-board (DPU and DCU) are derived, by means of a Point of Load (PoL), the secondary voltage levels needed by the hosted electronic components (e.g., memories, FPGA core etc.), as illustrated in the single electronics boards block diagrams (see Figure 12-25).
Page 186 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 167: Power Supply Unit block diagram
13.6.5 DCU design The design is the same what is proposed by ICU team in document ARIEL-RAL-PL-DD-001. The description is repeated with a minor correction needed for working with FGS DPU. See document ARIEL- CBK-PL-DD-001 v.1.6
13.6.6 FGS FCU Mechanical Description and power consumption
13.6.6.1 FGS FCU general requirement The FGS FCU overall size is 212(180) x155 x140 mm for main and redundant blocks together. The mass of the electronic box should be around 5,5 kg with contingency and with harness ≤ 6kg
13.6.6.2 FGS FCU mechanical information The preliminary analysis the FGS FCU bases on the design DPU and PSU for JUICE/ SWI experiment. Position of trays with PCBs and dimensions are taken from this instrument.
Page 187 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 168: DPU PSU design in JUICE/SWI experiment
13.6.7 Overall FCU Power Budget Max Power Unit [mW] FGS DPU 3800 DCU 3300 Detectors & sidecare 250 Total 7350 PSU efficiency 70% Total power 10500 Margin 20% Total power with margin 12600 Table 46: FCU Power consumption
13.7 FGS ALGORITHMS AND SOFTWARE DESIGN The FGS will have to carry out and support a number of different tasks through its application software. There will be functions to control the FGS subsystems, process the detector data and communicate with the spacecraft. The FGS application software will offer its functionality in the form of ECSS service commands and reports. The software includes FDIR functionality and offers several modes of operation to support maintenance and calibration activities.
13.7.1 Centroiding Algorithm The main requirement of the FGS is the centroiding performance of 10 milli-arcsec at 10 Hz. The FGS will also be used for focusing the main telescope and support this activity through a dedicated imaging
Page 188 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017 mode. For the best support of the operating modes, several centroiding and data extraction algorithms will be implemented, fully configurable by parameter and command. In the warm FGS control electronics the data will be processed in real-time. Output data products are reformatted images, centroid coordinates, dimensions and errors in both axes, photometry, glitch count and housekeeping. On-board compression will be used to reduce the telemetry. Additional data processing capabilities include frame stacking for PSF measurements. A number of calibration steps need to be carried out before the centroiding can be applied, most importantly bias and flat field correction. For the purpose of glitch detection and correction, several sub- frames will be buffered.
13.7.2 Photometer Channel Data Processing The detectors will be used in windowing mode to optimize integration time for collecting science data. The main detector contains FGS prime and the NIR/Phot1 channel. The FGS will be operated at 10Hz and science data coming from the FGS in this frame rate will be collected. The VIS/Phot1 channel will operate with a frame rate defined for the best SNR for this channel. The FGS2 detector will only be used for photometric purposes. The frame rate for both spectral channels will be defined to maximize SNR. In the case of failure of the FGS prime, the FGS redundant will take over the guiding function. Science data products that will be downlinked will be compressed in a lossless manner. Images will be compressed using an integer wavelet transform with an arithmetic compression backend. This will yield a factor 3, depending on the noise. For the centroids, compression will be much less efficient, as most parameters will be floats. If only the centroids using 30x30 pxs window and no images are transmitted 5,4kbit/s.
13.7.3 Data Products and Telemetry The FGS will deliver centroid data products and images to the spacecraft. In addition, these data products will be sent as science data products to ground. On top of that, HK are generated and sent. All rates and window sizes are configurable. A typical centroid dataset will consist of astrometric and photometric measures, plus several bits of status information. CONTENTS SIZE (BYTES) Astrometry X/Y position, X/Y FWHM (if applicable), variances 24 Photometry Integrated flux, FWHM flux, background flux, variances 24 Status Operating mode, time tag, glitch count, star count, validity flags 6 Table 47: Centroid data product. Floats are used where applicable During guiding, a typically 70x70 pixel window will be analyzed at 10 Hz. Assuming 16 bit words, the raw ROI data rate will be 78,4 kbit/s. In case 150x150 pixels should be transmitted at 10 Hz, a data rate of 360kbit/sec would be achieved. Thus, a SpaceWire link would be preferred over MIL-STD-41553. The centroid data will be 54 bytes per frame, or 4.32 kbit/s.
13.7.3.1 Data rate in measurement mode The data rates from the different functions of the FGS are shown in the table below.
Page 189 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Pixels in Pixels in Channel Bits per Freq. Time / No.bits / Total Gbits spectrum spatial total sample rate ramp ramp bits/sec per day direction direction FGS 4 21 10 840 0,07 AOCS FGS1 30 30 900 16 10 4 24 5400 0,43 FGS2 30 30 900 16 10 4 24 5400 0,43 VISPhot 30 30 900 16 10 4 24 5400 0,43 NIRSpec 160 60 16200 16 10 4 24 57600 4,63 TOTAL 73800 5,92 Table 48: FGS data rate in measurement mode In the case of very bright sources (0.1s saturation time), several CDS reads will be averaged together to make one frame every 4 sec
13.7.4 FGS Predicted Performance The figure below presents the results from a centroiding study demonstrating that the precision requirement can be met for the whole range of targets. A sample of 864 simulated images was used as input for the algorithm under test. The dots represent the computed centroid estimation errors. The colored solid line represent the mean true CEEs. The vertical dashed lines indicate the expected electron fluxes on the detector of faint and bright targets.
Figure 169: Centroiding performance as given by the centroid estimation error depending on source signal as estimated from simulations for the intensity weighted iterative centre of gravity algorithm. Note the unit of the ordinate is 0.1”, so the 0.01” limit is the horizontal dotted line at 0.1.
Page 190 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
14 IN-FLIGHT CALIBRATION AND DATA PROCESSING
14.1 IN-FLIGHT CALIBRATION
14.1.1 The Smooth Transition Philosophy In order to achieve the requirements set out in [AD2], the ARIEL instrument payload must be calibrated both before launch, during the ground testing and development phase and in-flight. The ARIEL Instrument Operations Science Data Centre (IOSDC) together with, where appropriate, with the ESA Science Operations Centre, SOC [AD6] will have responsibility for the calibration and operations of ARIEL throughout the mission. The IOSDC will also interact with the payload team during mission development. A smooth transition philosophy will be adopted to ensure a similar environment throughout all ARIEL Mission calibration phases from ground level channel tests, end-to-end ground testing, through in-orbit PV to the routine calibration phase. Under the smooth transition philosophy, the initial ground testing and calibration should closely resemble, and smoothly adapt/evolve, to the final operational environment. This smooth transition philosophy will facilitate the transfer of knowledge and procedures within the IOSDC as described in [RD21] and reduce conversion efforts / costs.
14.1.2 Ground Test Calibration Plan The ground test and calibration campaign will be the responsibility of the consortium and is described in [RD39]. It is desirable to implement as many tests as possible at the lower levels (Component and Unit Level) to avoid complication further down the signal chain at instrument (subsystem) or spacecraft (payload) level. The test plan will follow the methodology given below and will follow the development schedule outlined in [AD4]: • Component Level Test: Low level testing of components such as Detector SCA, CFEE, Dichroics, mechanism actuators, etc. • Unit Level (parts of a subsystem) Test: Unit level tests are those carried out on individual components of the system without inter-dependencies / influences from other components. This would include the FGS and AIRS detector systems, the FGS Beam splitter assembly, Cooler compressor, etc. • Susbsystem Level Test: This is the Instrument Level Testing (ILT) and would include the entire AIRS instrument, FGS instrument (including FCE), the Telescope Assembly (including the Bipods and the V-grooves), Instrument Control Unit (ICU) and the Cooler unit. Instrument level tests are those carried out on the entire instrument including the optical paths and inter-dependencies / influences. For the instrument level testing there will be a need for a telescope simulator or source collimator with same F number as “real” telescope. • Payload Module Level Test: This level comprises the integrated instruments and telescope payload and includes Payload Functional Tests, Test Facility Functional Tests, EGSE integration testing and Ground Segment End- to-end testing. It is expected that there will be the following payload models [RD4];
o STM: Structural and Thermal Model o PVM: Performance Verification Model (optical/electrical calibration) o PfM : Proto-flight model
A test matrix is prepared for each block of the ARIEL Payload. These blocks are defined as;
Page 191 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
• AIRS Instrument • FGS Instrument • Telescope Assembly • Instrument Control Unit • Cooler Unit • Payload System It is assumed that a test will comprise of one of four types • Measurement: To check against design intent at one or more operating conditions. • Verification: Check against requirements over all operational conditions. • Check: Single (or a few) point measurements covering a restricted set of operational conditions to confirm previous measurements have remained valid. • Calibration: Gather multiple measurements to allow future understanding of returned data. • Analyse: Analyse at a low level based partially on test data. Within each test matrix, each block is divided into a set of parameters to be tested, the level where the test takes place (Component, Unit, Subsystem, Payload Module) and the nature of each test (Measurement, Verification, Check, Calibration). The tests will require a suitable Electronic and Optical Ground Segment Equipment environment [RD39, RD40].
14.1.3 In-Flight Calibration Plan There will be a 6 month PV/commission phase characterisation of satellite performance and verify if performances characterised on the ground are still valid. Following this there will be regular calibration observations during routine operations (i.e. routine calibration phase). It is expected that calibration products and tables will be periodically updated during the operational phase. The preliminary set of planning for the in-flight calibration of the ARIEL payload is contained in the Operations and Calibration Plan, [RD20]. This combines use of the on-board source with use of well characterised target stars to achieve the necessary calibration accuracy over the timescales from seconds (for a single ramp) to minutes/hours (for a single transit) to years (for the mission where transits from the same target may be stacked). There will be an internal calibrator, situated behind an aperture in the ARIEL M5 mirror (see Section 5.4) covering the entire wavelength range of the AIRS instrument (via a thermal filament) and the FGS (via an array of LEDs). The main function of the calibrator is that of monitoring the flat field calibration. The flashes produced by the calibration are spatially reproducible in time and will be used during commissioning. The calibrator might be used during science operation if required. A standard set of stellar calibrators will encompass a large sample of G stars to monitor the time evolution response and stability of the system. They will be observed just like the science target and then fed through the pipeline [RD38]. Both short calibration and long calibration observations of G stars will be made, observed exactly like the target star. A list of suitable G-stars are given in [RD20]. Planetary Nebulae may be used for both wavelength calibration (to verify pre-launch ground testing, and to monitor throughout the mission) and if suffiently uniformly extended, for flat fielding. Suitable targets for these observations are provided from the Infrared Space Observatory – Short Wavelength Spectrometer (ISO-SWS) catalogue and the brighter targets observed by the Spitzer IRS. Any source seen with ISO-SWS is likely to be suitable for ARIEL to observe with high SNR. A list of suitable planetary nebulae are given in [RD20].
Page 192 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
An overview of the major calibration observations to be made during PV/commissioning phase is given in Table below. Calibration Calibrator Detector Dark Current Detector dark pixels or dark sky Pixel Deglitching Detector dark pixels or dark sky Detector Gain Variation On-board calibration source Non-linearity Calibration stars Afterglow Calibration stars and on-board calibration source Pixel Crosstalk Detector dark pixels or dark sky (glitches) Flat Fielding On-board calibration source or PNe Variations in Thermal Background Dark Sky Instrument Performance Aging Calibration stars Straylight variation Calibration stars Absolute Pointing Performance Calibration stars Pointing Stability Calibration stars Optical Distortion Calibration stars Optimisation of PSF Calibration stars Absolute Photometric Calibration Calibration stars Relative Photometric Calibration Calibration stars Wavelength Calibration PNe Table 49: Calibration observations during PV/commissioning phase
14.1.4 Routine Calibration Phase Calibration acquisitions in routine phase will be as follows; • A network of G stars (selected during PV phase) will be used for both the absolute flux calibration and stability monitoring • Whenever necessary via use of the internal calibration source • Wavelength calibration (PNe) • Pointing offset between FGS and AIRS (month to 6 month timescale by scanning slit across standard G star) • Background acquisition (dark sky observation) Calibration observations to monitor the stability of observations will be of two types in order to characterize all instrument drifts: • Short Calibration : Duration: 1 h typically every 36h, within a window of 24-48 h to monitor systematics (e.g. wavelength variation, drifts, etc) • Long Calibration : Duration: 6 h typically every 30d, within a window of 20-40 days to monitor stability on the timescale of a typical transit. The frequency and duration of the calibration observations will be refined and finalised during PV Phase.
Page 193 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
14.2 ON-BOARD DATA PROCESSING REQUIREMENTS The preliminary set of on-board data processing steps are discussed in section 10.8 above and in detail in [RD24].
14.3 GROUND DATA PROCESSING
14.3.1 The ARIEL Science Ground Segment The ARIEL Science Ground Segment consists of the consortium Instrument Operations Science Data Centre (IOSDC) and the ESA Science Operations Centre (SOC). The preliminary plan for the processing of the ARIEL data is contained in the ARIEL IOSDC Organisation Plan [RD21] and the Pipeline Description [RD38]. An overview of the entire ARIEL Ground Segment is shown in Figure 170 with the assumed division of responsibilities described in [AD6]. The IOSDC will be functionally organised into specialised teams with responsibility for specific work packages defined in [RD21]. The IOSDC teams are; Management Team: Overall control and organization of IOSDC including tasks assignment, interface / communication between teams and contact point between Consortium IOSDC and ESA/SOC; Operations Team: Operational procedures and interactions for instrument including: instrument operation logging, health monitoring and trend analysis test support; Software Team: data analysis pipelines and QLA tools production and softaware updates. Observations Team: interface between instrument/software and science side, including observation scheduling; science validation of observations during commissioning / PV phase. Editorial Team: Oversee all documentation for ARIEL IOSDC and interaction with instrument and the SOC. Production of the ARIEL data reduction guides. Instrument Team: Ground testing and instrument OBS; interacts closely with other teams during PV/commissioning; responsible for the IWS; Calibration Team: Initial definition of AOTs and Calibration plan for flight operations. Should also support testing. Following the smooth transition philosophy, the instrument and calibration teams will be independent during the early development phase but will be slowly integrated into the IOSDC in the final ground test / launch / commissioning phases.
Page 194 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Figure 170: ARIEL Science Ground Segment Overview with inputs/outputs shown as blue/red arrows respectively. The ESA responsibilities are in the blue boxes and the IOSDC in the green boxes.
14.3.2 ARIEL Science Data Levels and Products The overall science data processing is broken down into different pipelines leading to a specific data processing level and associated data product. ARIEL science Data Products are defined in Table 50. All Data Level products are ingested into the ARIEL archive, and made available to the science community. The Consortium/IOSDC will be responsible for the definition and algorithmic description of the pipelines, the software development, validation, delivery and update of the pipeline. Raw data will be delivered to the ESA SOC from the ESA Mission Operations Centre (MOC). The data analysis pipeline (Level 0 to Level 2 processing) will be delivered to ESA SOC by the IOSDC and the SOC will run the pipelines and populate the ARIEL data archive at the SOC.
Page 195 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
Produced Level Description by AIRS and FGS: 0 • raw telemetry packets from spacecraft, delivered from MOC to SOC.; MOC • unpacked and decompressed raw data;
AIRS and FGS/NIR-Spec: • raw spectral cubes of individual time frames; • formatted spectral cube images (cubes of raw spectral detector images (ADU - Analogue Detector Units) of all time frames, for each target observation); • spatial dispersion, spectral dispersion, intensity; • re-ordered, uncompressed, meta-data enriched, raw data of target (star 1 SOC + planet(s)) in ADU;
FGS1, FGS2, FGS/Vis-Phot: • raw images of individual time frames; • formatted images (ADU) of all time frames, for each target observation; • spatial dispersion, spectral dispersion, intensity; • re-ordered, uncompressed, meta-data enriched, raw data of target (star + planet(s)) in ADU; AIRS and FGS: • calibrated target (star + planet(s)) spectra and light-curves; 2 • Level 1 data converted to target flux and spectra vs time in physical SOC units; • science data after removal of instrument signatures; AIRS and FGS: • stellar variability decorrelated Level 2 data • fitted light-curves; 3 • wavelebgth (band) resolved image maps; IOSDC AIRS: • rebinned / stacked spectra • individual planet spectra; Table 50: ARIEL Science Data Levels and Products
14.3.3 ARIEL Science Data Processing Pipeline The ARIEL data products listed in Table 50 are created by passing the science data through an automated pipeline. The raw telemetry data is delivered from MOC to SOC and must be unpacked and decompressed. Data is likely to be re-organised into objects suitable for the data processing system as Level 0 data products. The Level 0 to Level 1 processing is carried out at ESA-SOC and will include time conversion, masking and unit/engineering conversions to produce formatted (but not calibrated) spectral cubes and images as the Level 1 products. The Level 1 to Level 2 processing is carried out at ESA-SOC and will remove various instrument signatures and detector effects (dark current, linearization, flat fielding, crosstalk) as well as the flux and wavelength calibration. The Level 2 product will be science ready target (star + planet) spectra and images. The Level 0, Level 1 and Level 2 products will be fed into the ARIEL data science archive at SOC. The Level 2 to Level 3 data processing will require sophisticated procedures and manual intervention in order to extract waveband resolved/integrated image maps (photometers) and the planet spectrum (spectrometers), e.g. [RD19]. The associated pipelines will not run automatically at SOC. Level 3 data products will be produced by IOSDC/Consortium and subsequently, with a delivery scheme to be agreed based also on quality criteria, ingested into the SOC archive.
Page 196 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
15 PAYLOAD PERFORMANCE MODELLING
15.1 EXOSIM OVERVIEW The payload performance is studied using ExoSim, an end-to-end simulator implementing a parametric description of the ARIEL payload design. ExoSim implements a realistic, time-domain simulation modelling the current baseline instrument. This allows a quick and cost-effective evaluation of the proposed design in delivering the ARIEL science as well as a tool to optimize the instrument, or conduct trade-off studies when required. ExoSim is different from the radiometric models used during the study the ARIEL mission, as it allows a thorough time-domain assessment of several systematics which would not be possible, or very difficult to study in static simulations. ExoSim allows both the consortium and ESA to ensure that the payload and spacecraft designs are optimised to enable the science of the mission without unnecessary costs and risks. This simulation tool has been used throughout the preparation of this report to define the expected instrument performance and demonstrate the mission science capabilities. Our time-domain simulator is a second-generation software, and inherit many of the algorithms developed for EChOSim14, a similar simulator developed for the M3 candidate mission EChO. However, ExoSim has been completely rewritten to make it more efficient, improving and extending the algorithms implemented. For instance, the new simulator is capable of implementing pointing jitter in two dimensions and to simulate the effect of a non-stable star on the timelines. It can simulate photometers, prims and grating based spectrometers. All relevant instrument effects are implemented, from wave front aberrations to intra-pixel responses and QE variations in the focal plane. ExoSim is open software, and a guide to the installation and operation is provided in the performance model document, [RD12]. In [RD11] a detailed description is provided of how simulations are used to study the performance of ARIEL. A thorough description of the algorithms implemented can be found in Sarkar et al15 and Sarkar et al16.
15.2 EXOSIM STRUCTURE The general philosophy behind the simulator is a central engine running several modules. Each module is specialized to address some aspect of the simulation. The outputs of the simulator can be computed thanks to dynamical parameters estimated within these modules or defined from data considered as inputs and computed by other means. With reference to Figure 171 the first module in the ExoSim logical flow simulates the Astroscene, providing a description of the astrophysical scene (star and planet) and zodiacal light. The Instrument module provides a description of the ARIEL instrument. Comprehends a description of the telescope and common optics and a description of all optical elements in each photometric and spectroscopic channel. Each of these optical elements are characterised by their transmission, emissivity and temperature. The PSF can be simulated using Gaussians, Airy functions or user provided in the form of data cubes (images at each wavelengths). The latter option is particularly useful to include in the simulations of the WFE. The Focal plane array is characterised by defining the detector pixel response function (intra-pixel response) and variations in QE across the focal plane. For spectrometers, light is dispersed and convolved with the focal plane using a user defined linear dispersion, characteristic of the the prism used, but gratings can be simulated as well, if needed. The Timeline Generator synthetizes the clock lines required to operate the detectors (frame clocks). In this way, it is possible to simulate up-the-ramp sampling with arbitrary reset times. Lightcurves vs wavelength are generated within this module. Options available are transit, eclipses and phase curves. For software architecture convenience, stellar effects (spots, granulation and pulsation) are estimated and applied to the lightcurves at this stage of the simulation.
14 Pascale, E., et al., ExA 40, 601, 2015 15 Sarkar, S. et al, 2016, SPIE 9904 16 Sarkar, S., et al, 2017, In preparation
Page 197 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The noise module applies stochastic processes to the timelines such as read noise, dark current and photon noise. Spatial and spectral jitter effects are simulated here having computed a realization of jitter timeline (displacement of boresight in two dimension) from a jitter PSD, or used a jitter timeline as input. The output of the simulations is stored in FITS files data cubes, i.e. raw NDR vs time. Because the simulations are realistic, an advanced reduction pipeline is required. Details of this pipeline are provided in [RD11].
Figure 171: Logical description of the ExoSim simulator. The instrument is defined in a parametric form which also contains all quantities required to define the target observed.
15.3 REFERENCE CASES AND BENCHMARKING ExoSim has been extensively validated using existing radiometric models (including that developed by ESA) as well as other space instrumentation, such as Hubble Space Telescope. A description of this validation work is provided in [RD17].
Page 198 Doc Ref: ARIEL-RAL-PL-DD-001 ARIEL Payload ARIEL Payload Design Issue: 2.0 Consortium Description Date: 15 February 2017
The comparison with the ESA radiometric model demonstrated that there is excellent agreement between the dynamic and static simulations (Figure 172).
Figure 172: Top panel - ExoSim focal plane signal compared to the signal from the ESA radiometric model for GJ 1214 (Left), and 55 Cancri (Right). Bottom panel – Comparison between noise estimates from the two simulators. The ExoSim noise results show the average noise variance/time for 50 realizations with the standard deviation shown as error bars. There is excellent agreement between the two simulator as shown by the residuals estimated from the differences between the two.
15.4 RESULTS DEMONSTRATING COMPLIANCE TO KEY SCIENCE PERFORMANCE REQUIREMENTS ExoSim was used throughout this study phase to assess the performance of instrument design developed and allowing iterations between the different study teams. The performance of the baseline design reported in this document is discussed in Section 4.4 and [RD11] where a comparison with requirements is shown to be fully compliant. We have used ExoSim to calibrate the noise model used by the ESA radiometric model [RD17] to estimate the predicted performance of the ARIEL baseline design [RD18]. We further used ExoSim to generate the simulations used as input in the spectral retrieval simulations to of [RD19].
Page 199