Final Report

Home , Improved Orion, Uchinoura Space Center

QB50

Sensor Selection Working Group

(SSWG)

Final Report

19 March 2012

Editor:

Prof. Alan Smith Mullard Space Science Laboratory (MSSL) University College London (UCL) Holmbury St. Mary Dorking, Surrey RH5 6NT United Kingdom

SSWG Membership

A. Smith (Chairman, Mullard Space Science Laboratory, UK) T. Beuselinck (RedShift BVBA, Belgium) J. F. Dalsgaard Nielsen (Aalborg University, Denmark) J. De Keyser (Belgian Institute for Space Aeronomy, Belgium) A. Gregorio (University of Trieste, Italy) D. Kataria (Mullard Space Science Laboratory, UK) V. Lappas (Surrey Space Centre, UK) F. J. Lübken (Institute for Atmospheric Physics, Germany) J. Moen (University of Oslo, Norway) S. Palo (University of Colorado, USA) R. Reinhard (von Karman Institute, Belgium) A. Ridley (University of Michigan, USA) J. Rotteveel (Innovative Solutions In Space, Netherlands) G. Schmidtke (Fraunhofer Institute for Physical Measurement Techniques, Germany) T. Schmiel (TU Dresden, Germany)

Table of Contents

1. Introduction ...... 4 2. The QB50 Project ...... 5 3. Science Context ...... 10 4. In‐situ measurements in the lower thermosphere – other projects ...... 13 4.1 Past satellites in highly elliptical orbits ...... 13 4.2 The Drag and Atmospheric Neutral Density Explorer (DANDE) ...... 20 4.3 Low‐Flying Spacecraft (LFSC) Daedalus ...... 21 4.4 Sounding Rocket experiments ...... 22 4.5 Armada ...... 26 5. QB50 Candidate Sensors ...... 28 5.1 Introduction ...... 28 5.2 Ion and Neutral mass spectrometer ...... 29 5.3 FIPEX: In‐situ Atomic Oxygen Measurement in Low‐Earth Orbit ...... 32 5.4 Multi‐Needle Langmuir Probe ...... 37 5.5 Magnetoresistive magnetometer ...... 40 5.6 Accelerometer ...... 43 5.7 GPS receiver ...... 46 5.8 Laser Retroreflector ...... 52 5.9 Thermistors/thermocouples ...... 53 5.10 Q‐BOS (Bolometric Oscillation Sensor) ...... 56 5.11 Silicon Detector ...... 59 5.12 Spherical EUV and Plasma Spectrometer (SEPS) ...... 63 5.13 WINCS ...... 68 5.14 Sensor Resource Summary ...... 70 6. QB50 CubeSat physical constraints and payload architecture ...... Error! Bookmark not defined. 7. Selection ...... 71 8. Recommendations ...... 74 9. Baseline sensor package configurations ...... 75 10. References ...... 76 11. Acronyms ...... 77 12. Contributors ...... 81

1. Introduction

The QB50 Sensor Selection Working Group met on four occasions: 3 March 2011, 4 May 2011, 16-17 June 2011, and 26 July 2011. All meetings were held at the von Karman Institute for Fluid Dynamics, Rhode‐St‐Genèse, near Brussels, Belgium. The results of the working group were reported on 27 July 2011 at the 2nd QB50 workshop and then to the QB50 Steering Committee on the same day.

The task of the SSWG was to identify optimal sensor package options for the network part of the QB50 mission taking into consideration: scientific objectives; potential science instruments, their availability and their heritage; the very limited spacecraft resources of CubeSats; and the limited available financial resources.

The SSWG remained cognisant of the overall QB50 mission objectives and recognised that these were perceived differently by the variety of stakeholders including EC, the Science Community and CubeSat providers. The group also recognised the importance of the science context both in terms of what were the key scientific issues to be addressed, and also what missions have flown or are proposed for this area of endeavour.

Given the CubeSat nature of the mission, accommodation constraints for any selected sensor package were naturally important and given high priority in discussions. A wide range of sensor options were considered with additional options being introduced as they became known to the group. Expert advice was sought where appropriate and budgetary considerations were included in the SSWG deliberations.

Inputs were received from a wide variety of sources including other QB50 working groups, commercial organisations, science groups, and potential instrument providers. Meetings included teleconference sessions and presentations.

This report outlines the QB50 project, provides science context, describes earlier missions and future mission proposals in this area and then goes on to describe the science instrument options that were considered. A brief description of the payload architecture and payload interface requirement is then followed by the selected payload options and other recommendations.

This report was compiled from numerous inputs (attributed below) and edited by A. Smith.

2. The QB50 Project R. Reinhard, VKI

QB50 is a network 50 CubeSats in a ‘string‐of‐pearls’ configuration that will be launched together in the first half of 2015 by a single rocket, a Shtil‐2.1, from Murmansk, Northern Russia into a circular orbit at 320 km altitude, inclination 79°. The 50 CubeSats will comprise about 40 atmospheric double CubeSats and about 10 double or triple CubeSats for science and technology demonstration. All 40 atmospheric double CubeSats and most of the 10 double and triple CubeSats for In‐Orbit Demonstration (IOD CubeSats) will carry a set of standardized sensors for multi‐point, in‐situ, long‐duration measurements of key parameters and constituents in the largely unexplored lower thermosphere and ionosphere. These multi‐point measurements will allow the separation of spatial and temporal variations. Due to atmospheric drag, the CubeSat orbits will decay and progressively lower and lower layers of the thermosphere/ionosphere will be explored without the need for on‐ board propulsion. The mission lifetime of individual CubeSats is estimated to be about three months.

QB50 will also study the re‐entry process by measuring a number of key parameters during re‐entry, e.g. CubeSat on‐board temperature and deceleration. The re‐entry process will also be studied by comparing predicted (using a variety of atmospheric models and different trajectory simulation software tools and CubeSat ballistic coefficients) and actual CubeSat trajectories and orbital lifetimes, and by comparing predicted and actual times and latitudes/longitudes of atmospheric re‐entry.

The initial total network size in orbit is determined by the deployment sequence, deployment speed and direction, and can be selected anywhere between 500 and 5000 km. It is currently planned to deploy one CubeSat every orbit, i.e. every 86 minutes. All deployment will preferably take place on the dayside of the orbit. The deployment speed will be 2‐3 m/s and the deployment direction will be uncontrolled. The initial distance between individual CubeSats will be between a few tens and a few hundred kilometres. Orbital modelling has shown that due to density variations along the orbit and small differences in the CubeSat ballistic coefficients the separation distance will change, eventually leading to a non‐uniform distribution of CubeSats all the way around the Earth. In this way, the CubeSats will be able to explore temporal and spatial variations over a wide range of scale sizes, from a few tens of kilometres in the beginning to about 1000 km after a month.

A single CubeSat is simply too small to also carry sensors for significant scientific research. Hence, for the universities the main objective of developing, launching and operating a CubeSat is educational. However, when combining a large number of CubeSats with identical sensors launched at the same time into a network, in addition to the educational value, fundamental scientific questions can be addressed which are inaccessible otherwise. Networks of CubeSats have been under discussion in the CubeSat community for several years, but so far no university, institution or space agency has taken the initiative to set up and coordinate such a powerful network. CubeSat reliability is not a major concern because the network can still fully achieve its mission objectives even if a few CubeSats fail.

QB50 has been selected as the first large‐scale CubeSat network in orbit because a network of CubeSats in the lower thermosphere as compared to networks in higher orbits has the following advantages

- The lifetime of a CubeSat in the envisaged low‐Earth orbit will only be three months, i.e. much less than the 25 years stipulated by international requirements related to space debris (Code of Conduct), - A low‐Earth orbit allows high data rates because of the short communication distances involved, - In their low‐Earth orbits, the CubeSats will be below the Earth’s radiation belts, which is advantageous because CubeSats use low‐cost COTS components which are not radiation hardened, - The orbit of the International Space Station (ISS) is maintained between 335 km (perigee) and 400 km (apogee). If a network of many CubeSats is launched into an orbit that is above that of the ISS there is a danger of collision with the ISS when the orbits of the CubeSats decay due to atmospheric drag. If the initial orbit of the CubeSats is below 330 km there is no danger of collision.

For a network of CubeSats in the lower thermosphere/ionosphere the short mission lifetime is not a deterrent as for a university the primary purpose of a CubeSat is educational and the educational objectives can be fully achieved even if the orbital lifetime is short.

The lower thermosphere/ionosphere (90‐320 km) is the least explored layer of the atmosphere. Five Atmospheric Explorers were flown by NASA from 1963 until 1981 in highly elliptical orbits (typically: 200 km perigee, 3000 km apogee); they carried experiments for in‐ situ measurements but the time spent in the region of interest below 320 km was only a few tens of minutes. Nowadays, sounding rocket flights provide the only in‐situ measurements. While they do explore the whole lower thermosphere, the time spent in this region is rather short (a few minutes), there are only a few flights per year and they only provide measurements along a single column. Powerful remote‐sensing instruments on board Earth observation satellites in higher orbits (600–800 km) receive the backscattered signals from atmospheric constituents at various altitudes. While this is an excellent tool for exploring the lower layers of the atmosphere up to about 100 km, it is not ideally suited for exploring the lower thermosphere because there the atmosphere is so rarefied that the return signal is weak. The same holds for remote‐sensing observations from the ground with lidars and radars. The multi‐point, in‐situ measurements of QB50 will be complementary to the remote‐sensing observations by the instruments on Earth observation satellites and the remote‐sensing observations from the ground with lidars and radars.

Space agencies are not pursuing a multi‐spacecraft network for in‐situ measurements in the lower thermosphere because the cost of a network of 50 satellites built to industrial standards would be extremely high and not justifiable in view of the limited orbital lifetime of only a few months. No atmospheric network mission for in‐situ measurements has been carried out in the past or is planned for the future. A network of satellites for in‐situ measurements in the lower thermosphere can only be realised by using very low‐cost satellites, and CubeSats are the only realistic option.

Interest to participate in QB50 by signing a Letter of Intent (LoI) has been expressed by over 90 universities all over the world. This includes universities in 30 European countries and universities in Argentina, Australia, Brazil, Canada, Chile, China, Ethiopia, India, Israel, Peru, Puerto Rico, Russia, Singapore, South Korea, Taiwan, Turkey, Ukraine, USA and Vietnam. For many of these countries, the CubeSats that will participate in QB50 will be their first satellite in orbit and a matter of national prestige. QB50 will eventually involve well over 1000 people from all over the world: 50 CubeSat teams of 10‐20 students and engineers each, numerous atmospheric scientists, instrument, orbital dynamics and ground station experts, legal experts, as well as industry and space agency representatives.

For QB50, double‐unit CubeSats are foreseen, with one half (the ‘functional’ unit) providing the usual satellite functions and the other half (the ‘science’ unit) accommodating a set of standardised sensors for lower thermosphere/ionosphere and re‐entry research. University teams are free to use any space left in the ‘functional’ unit of the double CubeSat for a technology package or a sensor of their own choice. In addition to the 40 ‘atmospheric’ double CubeSats, up to 10 ‘special’ double or triple CubeSats for science and technology demonstration will be selected. Some of the special IOD CubeSats can also accommodate either the Set 1 or Set 2 standard sensor package (see Chapter 10). If less than 10 IOD CubeSats are selected, the number of atmospheric double CubeSats can be higher so that the total of 50 CubeSats is maintained. If less than 40 atmospheric CubeSats are selected, the number of special IOD triple or double CubeSats can be higher so that again the total of 50 CubeSats is maintained.

All 50 CubeSats will be launched together on a single launch vehicle, a Russian Shtil‐2.1, from Murmansk in Northern Russia into a circular orbit at 320 km altitude, inclination 79°. Due to atmospheric drag, the orbits of the CubeSats will decay and progressively lower and lower layers of the thermosphere will be explored without the need for on‐board propulsion, perhaps down to 90 km. The maximum achievable orbital altitude for a 230 kg payload is 320 km for a circular orbit. If fewer than 50 CubeSats are proposed and selected or if the deployment system is lighter than currently estimated (80‐100 kg) the orbital altitude can be increased slightly (but not above 335 km, the lowest perigee of the ISS), extending the lifetime.

The initial total network size in orbit is determined by the deployment sequence and the separation speed and direction. The optimal CubeSat deployment scenario is currently under study, bearing in mind

- Launch vehicle and deployment system constraints, - The need to identify each CubeSat individually as soon as possible after deployment, - The need to establish a telecommunications link between the various CubeSats and ground stations as soon as possible after deployment, - The need to mitigate the risk of collisions between the CubeSats and between the CubeSats and the third stage/deployment system assembly after each full orbit when the CubeSats return to the original point of release, - The need to achieve spreading of the CubeSats all the way around the Earth as soon as possible after deployment.

Regarding ground stations, the current situation is typically as follows: A university builds a CubeSat and launches it into low‐Earth orbit. It also builds or already has available a ground station to track the CubeSat and enable uplink/downlink telecommunications. The period in a low‐Earth orbit (about 600 km altitude) is typically about 90 minutes, but the duration of a satellite pass over the ground station is very short and varies from approximately 10 minutes in the best case to no coverage at all for most of the 16 daily orbits. When supporting only one satellite project the ground station is not in operation 97 % of the time. This is highly inefficient and often, despite on‐board data storage capability, a limiting factor in mission science return.

For very low‐Earth orbits (150‐300 km altitude) as envisaged for QB50 the situation is a lot worse. The following table illustrates the problem

Orbital altitude Duration of the Daily coverage by a longest pass single ground station 600 km 10 min 3 % 300 km 5 min 0.7 % 150 km 2.5 min 0.25 %

Moreover, there are sometimes mission critical operations requiring uninterrupted coverage for longer than a single pass and, worst of all, in the case of an on‐board emergency, there is no immediate access to the satellite. The situation would dramatically improve if the satellite could be tracked by numerous other ground stations along its track.

An international network, the Global Educational Network for Satellite Operations (GENSO), is now being set up. The QB50 launch is currently planned for April 2015. By that time, GENSO will be fully operational. It will eventually comprise more than 150 ground stations in different parts of the world, providing a vastly improved uplink and downlink capability for all CubeSats. The international QB50 network in orbit will be the first major user of the international GENSO network on the ground.

Parallel ‘QB50 Mission Control Centres’ will be set up at VKI, Stanford in the US and NPU in China with the following real‐time functions for all 50 CubeSats

- Comparing predicted with actual trajectories, using different trajectory simulation software tools, atmospheric models and CubeSat drag coefficients, - Monitoring the status and health of the 50 CubeSats and the deployment system, - Displaying which ground station is in contact with which CubeSat and displaying the link quality, - Predicting and continuously updating the approximate time and latitude/ longitude of atmospheric re‐entry for the 50 CubeSats.

The CubeSat teams will be in charge of the operations of their CubeSats and are expected to submit all science data, key housekeeping data and appropriate documentation to the QB50 Data Processing and Archiving Centre (DPAC) within 6 months after the end of the mission. The Centre will be located in the Belgian User Support and Operation Centre (BUSOC) in the Belgian Institute for Space Aeronomy (BIRA) in Brussels. The QB50 DPAC has the following tasks

- Serving as the single point of contact between the QB50 Project and the CubeSat teams regarding science and housekeeping data, e.g. format specification, - Verification (checking for completeness) and cataloguing of the data submitted to DPAC by the CubeSat teams, - Support of the CubeSat teams in producing the necessary documentation describing their data set, - Transfer of the data files into a uniform, user‐friendly format, - Handling of requests for data by the user community and providing clarifications to the user community if requested, - Archiving of all data and documentation

The correlated evaluation of all QB50 science data, together with simultaneously obtained sounding rocket and remote‐sensing data, will be made in 2015 by the Institute for Atmospheric Physics (IAP) in Kühlungsborn in Northern Germany. In the years before the launch of QB50, IAP will provide science support to the Project, for example, advice for the selection of the sensors for the science units of the double CubeSats and will interface with the DPAC on specifying the preferred, user‐ friendly formats for the science and housekeeping data. About 9 months after the end of the mission, IAP will receive all science data for scientific data analysis. Correlations of the data will be made and apparent discrepancies will be pointed out and analysed in cooperation with the CubeSat teams.

Furthermore, the QB50 science data will be correlated with remote‐sensing observations from the ground by lidars and radars up to 110 km altitude and with in‐situ measurements by experiments on sounding rockets up to 350 km altitude, if such data are available and are provided by the experimenter teams. IAP have their own lidar and radar sites but data from the most important lidar and radar sites located in other parts of the world will also be used. The QB50 science data will also be correlated with the relevant remote‐sensing observations of the MLT region by Earth observation satellites in higher orbits (600‐800 km). IAP has the intention to organise 'lower thermosphere data analysis and correlation workshops' with all interested parties.

3. Science Context A. Aruliah, UCL

The thermosphere region (90‐320 km altitude) to be explored by the QB50 network is highly important for atmospheric studies, yet has the fewest direct measurements due to its inaccessibility. Observations are limited to a few minutes of a rocket flight, or radar and optical observations from a handful of sites. The rest of our current understanding comes from theoretical models and semi‐ empirical models that use interpolation of measurements from above and below. The measurements come from satellites at much higher altitudes (400‐1000 km e.g. CHAMP and Dynamics Explorer), and highly elliptical orbits (described in next section); and measurements from lower altitudes using lidars and riometers. The beauty of the QB50 network is that for the first time we will have simultaneous multi‐point global coverage of the lower thermosphere for the 3 month period lifetime of the network. (Note that in order to maximise the ground‐based support it is important to consider the seasonal dependence of the operation of instruments, such as optical instruments that only operate during darkness.)

The most widely used semi‐empirical model of the thermosphere is the MSIS model, which gives neutral particle composition, density and temperature. This is extremely valuable, yet is predominantly based on observations made in the 1970s and 1980s from satellite and radar measurements. This should be compared with meteorological models that are continuously updated with observations from weather balloons, radars and satellites on an hourly basis. It is now predicted that the Sun is coming out of a grand solar maximum phase. The last solar minimum has been 2‐3 years longer than average, resulting in unusually quiet geomagnetic conditions. It is likely that a major revision of atmospheric models will be necessary to allow for what are the new “average conditions” and ranges. This will have a knock‐on effect on the modelling of Space Weather and the success of forecasting for the use of the satellite and technology industry.

The key features of the 90‐320 km altitude region are:

• Sharp increase in temperature with height: The mesopause is the coldest layer of the atmosphere. For a long time it was considered to be a single layer at 85 km but recent analysis of the TIMED satellite observations has shown a double layer (Xu et al., 2007). Above the mesopause the temperature rises rapidly until around 300‐400 km when it becomes a constant value (an isotherm). Below the turbopause (~100 km) there is a homogeneous mix of the chemical composition, but above this altitude molecular diffusion dominates and the density of each chemical drops off exponentially according to its mass. This causes a rapid decrease in density with altitude and a change in chemical composition. Regions of sharp changes in parameters require detailed measurements in order to model them correctly. • Planetary waves, tides and atmospheric gravity waves: Solar heating of the Earth’s atmosphere creates regions of high and low temperature, and consequently variations in chemical composition and density. Planetary waves and tides are a range of atmospheric motions labelled according to their spatial extent and period, and they have a global influence. Thus planetary waves have wavelengths of the order of the radius of the Earth. Tides are harmonics of the daily 24 hour heating cycle, with the 24 and 12 hour tides being dominant. At the other extreme, AGWs have wavelengths of the order of kilometres to

hundreds of kms, and these are generated by localised atmospheric perturbations, typically caused by thunderstorms, the irregular shape of mountain ranges and auroral activity. These atmospheric motions transport energy and momentum both horizontally and vertically, and across magnetic field boundaries. Tides are a dominant feature of the mesosphere and lower thermosphere. AGWs are critically important in lower atmosphere models, and are of increasing interest in upper atmosphere studies. A chain of cubesats would allow accurate in‐situ measurement of the frequency composition of atmospheric perturbations. • The E‐ and F‐region dynamos at altitudes ~120 km and ~300 km: There is a global Sq current system in the E‐region caused by the tidal movement of charged particles across the Earth’s magnetic field lines. Figure 3.1 from Matsushita (1968) shows the global Sq current system from 69 world‐wide magnetometer stations using measurements collected during the period of the first International Geophysical Year in 1957. Two other main features of the E‐region dynamo are the equatorial and auroral electrojets. The equatorial electrojet is a band along the equator on the dayside of fast moving charged particles in the E‐region. The auroral electrojets are equivalent bands circling the north and south poles (see next point). These are all strong features that appear as perturbations in ground‐based magnetometer measurements of the Earth’s magnetic field. However, the limitation of ground‐based observations is that they integrate the magnetic perturbation from currents at all heights. In‐ situ measurements of currents in the E‐region have not been done since the highly elliptical orbit satellites 1960s and 1970s.

Figure 3.1 From Matsuhita (1968)

• Auroral regions: The ionosphere within the auroral regions (poleward of 70˚ latitude) bears the clear signatures of a projection of the magnetosphere. Ionised particle energies indicate their origins within the magnetosphere (e.g. cusp, tail or lobe). Heating by auroral particle

precipitation; or resistive heating by currents (Joule heating) driven by the magnetospheric dynamo, are the main energy sources in the auroral region after solar heating. The magnetospheric dynamo is at least an order of magnitude larger than the Sq current system. One important consequence of auroral particle precipitation is the production of nitric oxide ions from high energy particles that penetrate deep into the atmosphere. These go on to destroy ozone which protects the Earth’s surface from UV radiation. Another consequence is radio blackout from solar flares, where precipitation increases ionospheric density so much that HF radio waves are absorbed. This is highly dangerous for communications on cross polar aircraft. Even on a day to day basis, the variability of the auroral ionosphere is considerable and persistent, which makes GPS unreliable even under geomagnetically quiet conditions. Standard GPS systems use fairly simple models of the ionosphere since other sources of error are considered to be greater. The global coverage of the ionosphere provided by the QB50 network would be extremely valuable in improving the modelling of small‐scale structures and temporal variation of the ionosphere. Multi‐satellite observations (e.g., the 4 Cluster spacecraft) have shown their value in distinguishing temporal and spatial variations.

4. Insitu measurements in the lower thermosphere – other projects

4.1 Past satellites in highly elliptical orbits R. Reinhard, VKI

NASA has launched 92 Explorer satellites, 10 of which made in‐situ measurements in the lower thermosphere, most importantly, NASA’s five Atmospheric Explorers (AE) and the Dynamics Explorer 2. As the lifetime of a satellite in the lower thermosphere in a circular orbit is relatively short due to atmospheric drag all satellites listed below were in highly elliptical orbits. These orbits, however, have the disadvantage that the time spent during each orbit in the lower thermosphere is only ~10 minutes out of an orbital period of ~2 hours.

In the 1960s, NASA also launched four spheres, each 3.65 m in diameter, to determine the air density at orbital altitude by tracking the orbits of the spheres from the ground. These were: Explorer 9, AD‐A (Explorer 19), AD‐B (Explorer 24) and AD‐C (Explorer 39). They were all launched into similar orbits with perigees around 600 km and apogees around 2500 km. The decay of their orbits due to atmospheric drag was tracked until they burnt up in the atmosphere at ~100 km altitude.

Missions with perigees in the thermosphere whose primary objective was to study the radiation belts (natural or artificial) or to determine the flux of micrometeorites are not included below.

Atmospheric Explorer A (Explorer 17) Orbit: 255 x 916 km inclination: 57.6° Launch date: 2 April 1963 re‐entry: 24 November 1966 Satellite mass: 183.7 kg shape: spherical, diameter 95 cm

Scientific payload • 2 neutral mass spectrometers • 2 Langmuir probes to measure ion concentration and electron temperature • 4 pressure gauges to measure the total neutral particle density

Atmospheric Explorer B (Explorer 32) Orbit: 276 x 2725 km inclination: 64.7° Launch date: 25 May 1966 re‐entry: 22 February 1985 Satellite mass: 224.5 kg shape: spherical, diameter 89 cm

Scientific payload • 2 neutral mass spectrometers • Ion mass spectrometer • 3 pressure gauges • 2 Langmuir probes • Optical and radio/radar tracking of the spherical satellite near perigee to determine air density

Atmospheric Explorer C (Explorer 51) Orbit: 149 x 4294 km inclination: 68°

Launch date: 16 December 1973 re‐entry: 12 December 1978 Satellite mass: 658 kg shape: multi‐sided polyhedron

Scientific payload • Open‐source neutral mass spectrometer • Closed‐source neutral mass spectrometer • Magnetic ion mass spectrometer • Bennett ion mass spectrometer • Cold cathode ion gauge • Low‐energy electrons • Atmospheric density accelerometer* • Neutral atmosphere temperature • Cylindrical electrostatic probes • Capacitance manometer • Photoelectron spectrometer • Retarding potential analyser/drift meter • EUV Spectrometer • 3 single‐axis accelerometers (miniature electrostatic analysers; the drag was determined from the electrostatic force required to re‐centre a proof mass)

Atmospheric Explorer D (Explorer 54) Orbit: 154 x 3816 km inclination: 90.1° Launch date: 6 October 1975 operations terminated: 29 January 1976 Satellite mass: 681 kg re‐entry: 12 March 1976 shape: multi‐sided polyhedron

Scientific payload • Cylindrical electrostatic probe • Atmospheric density accelerometer* • Photoelectron spectrometer • Retarding potential analyser/drift meter • Open‐source neutral mass spectrometer • Neutral atmosphere composition • Neutral atmosphere temperature • Magnetic ion mass spectrometer • Low‐energy electrons

Atmospheric Explorer E (Explorer 55) Orbit: 156 x 2983 km inclination: 19.7° Launch date: 19 November 1975 re‐entry: 10 June 1981 Satellite mass: 735 kg shape: multi‐sided polyhedron

• Bennett ion mass spectrometer • EUV Spectrometer

Energetic Particle Explorer D (EPED) (Explorer 26) Orbit: 171 x 8545 km inclination: 18.1° Launch date: 21 December 1964 telecommunications failed on 26 May 1967 Satellite mass: 45.8 kg

Scientific payload • 6 solid state electron and proton counters • 2 plastic scintillation counters, one omni‐directional (p: 40‐110 MeV, e: >4 MeV), the other unidirectional (p: >5.2 MeV, e: >0.5 MeV) • Boom mounted biaxial fluxgate magnetometer • Low‐energy proton/electron scintillation detector

SCubed A (Explorer 45) Orbit: 224 x 27031 km inclination: 3.5° (San Marco platform launch) Launch date: 15 November 1971 re‐entry: 10 January 1972 Satellite mass: 52 kg operations terminated: 30 September 1974

Scientific payload • 2 channel electron multipliers with electrostatic analysers • 2 solid state particle telescopes to measure the flux of electrons, protons and selected ions • Magnetic spectrometer followed by 4 solid state detectors to measure electrons in the range 35‐560 keV • Triaxial fluxgate magnetometer mounted on a 76 cm long boom • 2 perpendicular search coil magnetometers mounted on a 61 cm long boom • DC/AC electric field instruments, two 14 cm diameter metal spheres (mounted on the ends of 2 booms) with a centre‐to‐centre separation of 5 m

Magsat (Explorer 61) Orbit: 352 x 578 km inclination: 96.8° Launch date: 30 October 1979 re‐entry: 11 June 1980 Satellite mass: 158 kg

Scientific payload • 2 magnetometer sensors on a 6 m long boom behind the satellite • Vector fluxgate magnetometer • Scalar magnetometer (2 dual‐cell caesium vapour sensor heads)

Dynamics Explorer 2 (DE2) (Explorer 63) Orbit: 309 x 1012 km inclination: 90° Launch date: 3 August 1981 re‐entry:19 February 1983 Satellite mass: 420 kg shape: short polygon

Scientific payload • Triaxial fluxgate magnetometer • Vector electric field instrument, six 11 m long booms (9 m insulated, outer 2 m active), distance between the midpoints of the 2 m active elements 20 m • Quadrupole neutral/ion mass spectrometer

• Retarding potential quadrupole neutral/ion mass spectrometer • Ion drift meter to measure horizontal and vertical ion drift velocities with an accuracy of 50 m/s • Retarding potential analyser • Low‐altitude plasma instrument, an array of 15 parabolic electrostatic analysers to measure positive ions and electrons 5 eV‐32 keV • Langmuir probe to measure electron temperature and electron/ion concentration and satellite potential

FAST (Explorer 70) Orbit: 350 x 4175 km inclination: 83° Launch date: 21 August 1996 operations terminated: 1 May 2009 Satellite mass: 191 kg shape: cylindrical

Scientific payload • electric field instrument • Langmuir probe • Fluxgate magnetometer • Search‐coil magnetometer • Electrostatic analyser • Time‐of‐flight energy mass analyser

Space Technology 5 (ST5) 3 satellites in a ‘string‐of pearls’ configuration Orbit: 300 x 4500 km inclination: 105.6° Launch date: 22 March 2006 operations terminated: 30 June 2006 Satellite mass (each): 25 kg shape: 8‐sided cylinder Spin stabilised at 20 rpm height: 48 cm, diameter: 53 cm

New technologies tested

• Cold gas microthruster • Miniaturised X‐band transponder (100 kbps) • Variable emission coatings for thermal control • Highly‐sensitive magnetometer on a boom • Operations of all 3 spacecraft as a single constellation rather than operating them individually

The San Marco Programme was a collaborative programme between the Italian Centro Ricerche Aerospaziale (CRA) and NASA. In this programme, a total of five satellites were launched which all made in‐situ measurements in the lower thermosphere. The satellites were launched on Scout rockets. The first launch was from the Wallops Flight Facility in the US, the other four from the San Marco platform, a former oil platform. It was towed to an equatorial location off the coast of Kenya which, when combined with an easterly firing, provided the most energetically favourable launch.

The ‘atmospheric drag density accelerometer’ was flown on San Marco 2, 3, 4 and D/L. The drag was determined by measuring the displacement in three directions of the spacecraft outer shell subject to atmospheric drag with respect to the heavy inner structure (electronics, batteries, etc.). The outer shell was connected to the inner structure through a series of flexible arms. An elastic drag balance

16 system was able to detect the displacement of the outer shell, subject to the aerodynamic force, with respect to the inner body both in value and direction, resolving any relative translation along three mutually orthogonal axes. The translation of the elastic system was changed into voltages. From the drag at orbital altitude the ambient air density was derived.

San Marco 1 Orbit: 198 x 846 km inclination: 37.8° Launch date: 15 December 1964 re‐entry: 13 September 1965 Satellite mass: 115.2 kg

Scientific payload • Spherical ion probe to study positive ion composition and temperature • Electron‐content beacon, a high‐frequency radio transmitter to study ionospheric effects on long‐radio radio transmissions

San Marco 2 Orbit: 135 x 498 km inclination: 2.9° Launch date: 26 April 1967 re‐entry: 14 October 1967 Satellite mass: 129.3 kg shape: spherical, diameter: 66 cm

Scientific payload • Atmospheric drag density accelerometer • Electron‐content beacon, a high‐frequency radio transmitter to study ionospheric effects on long‐radio radio transmissions

San Marco 3 Orbit: 226 x 723 km inclination: 3.3° Launch date: 24 April 1971 re‐entry: 29 November 1971 Satellite mass: 163.3 kg shape: spherical, diameter: 75 cm

Scientific payload • Atmospheric drag density accelerometer

• Omegatron mass spectrometer to measure the density and temperature of N2 (US provided)

• neutral mass spectrometer to measure the density of N2, O2, O, Ar, He (US provided)

San Marco 4 Orbit: 232 x 905 km inclination: 2.9° Launch date: 18 February 1974 re‐entry: 4 May 1976 Satellite mass: 164.0 kg shape: spherical, diameter 75 cm

Scientific payload • Atmospheric drag density accelerometer

• Omegatron mass spectrometer to measure the density and temperature of N2 (US provided)

• neutral mass spectrometer to measure the density of N2, O2, O, Ar, He (US provided)

San Marco D/L Orbit: 263 x 615 km inclination: 3° Launch date: 25 March 1988 re‐entry: 6 December 1988

Satellite mass: 273.0 kg shape: spherical, diameter: 96.5 cm

Scientific payload • Drag balance and air density • Retarding potential analyser/drift meter to determine the ion velocity, plasma density and temperature (US provided) • Retarding Potential Quadrupole (RPQ) mass spectrometer to determine neutral winds, neutral particle temperatures and the concentration of selected gases (US provided) • 3‐axis electric field instrument (US provided) • Airglow Solar Spectrometer Instrument (20‐700 nm)

Azur (NASA/German BMWF collaborative) Orbit: 387 x 3150 km inclination: 103° Launch date: 7 November 1969 telemetry system malfunctioned early July 1970 Satellite mass: 70.7 kg

Scientific payload • Fluxgate magnetometer on a 80 cm long boom • 2 telescopes, stack of 7 silicon detectors (p: 1.5‐104 MeV, α: 6‐19 MeV) • Proton solid state detector (p: 0.25‐13.5 MeV, α: 2‐6.4 MeV) • Proton/electron detector (p: 20‐72 MeV, e: >1.5 MeV) • 4 Geiger tube electron counters (p: >0.7 MeV, e: >40 keV) • 2 Geiger Mueller counters (p: >12 MeV, e: >0.7 MeV)

AerosA (DLR/NASA collaborative) Orbit: 223 x 867 km inclination: 96.9° Launch date: 16 December 1972 re‐entry: 22 August 1973 Satellite mass: 125.7 kg shape: cylindrical

Scientific payload • Quadrupole neutral/ion mass spectrometer • Quadrupole mass spectrometer to determine the total gas density and the kinetic temperature and density of molecular nitrogen • Retarding potential analyser to determine plasma density • Impedance probe for measuring electron concentration (this experiment malfunctioned on Aeros‐A) • EUV Spectrometer

AerosB (DLR in Germany) Orbit: 217 x 879 km inclination: 97.4° Launch date: 16 July 1974 re‐entry: 25 September 1975 Satellite mass: 125.7 kg shape: cylindrical

Scientific payload • Quadrupole neutral/ion mass spectrometer • Quadrupole mass spectrometer to determine the total gas density and the kinetic temperature and density of molecular nitrogen • Retarding potential analyser to determine plasma density • Impedance probe for measuring electron concentration • Atmospheric drag analysis by tracking the satellite near perigee • EUV Spectrometer

BremSat 1 (ZARM/University of Bremen in Germany) Initial orbit: 344 x 363 km inclination: 56.9° Launch date: 3 February 1994 re‐entry*: 12 February 1995 Satellite mass: 63 kg shape: 12‐sided prism diameter: 48 cm, height: 52 cm * satellite transmitted data until 110 km orbital altitude

Scientific payload

• Measurement of density distribution and dynamics of dust particles • Measurement of atomic oxygen • Measurement of deceleration and temperature during satellite re‐entry • Magnetometer • gyroscope

Taiyo (“Sun”) (Institute for Space and Aeronautical Science (ISAS) in Tokyo Orbit: 249 x 3129 km inclination: 31.5° Launch date: 24 February 1975 re‐entry: 29 June 1980 Satellite mass: 86 kg shape: octagonal

Scientific payload • Bennett ion mass spectrometer to measure O, He, H • Electron density measurement • Electron temperature • Retarding potential analyser

4.2 The Drag and Atmospheric Neutral Density Explorer (DANDE) S. Palo, M. Pilinski, K. Kemble, University of Colorado

The need to accurately predict satellite positions is a leading aspect of space situational awareness and presents increased challenges in the specification of the spacecraft environment in low earth orbit. Atmospheric drag is the most important environmental perturbation for low orbiting spacecraft and the most difficult one to model and predict precisely. The goal of this mission is Table 4.2.1: Minimum success science requirements.

Figure 4.2.1: CU students preparing DANDE for testing. therefore a method for the characterization of satellite drag through the use of a dual‐instrument in‐ situ approach. The elements of this method include a novel acceleration measurement suite and a small Wind and Temperature Spectrometer (WTS) in order to measure both atmospheric density and wind. The spacecraft structure is also a part of this measurement system as it houses the instruments and interacts directly with the atmospheric gas. Accordingly, requirements on the level of cross‐sectional area variation were derived from the science analysis. In order to evaluate this approach in orbit, students at the University of Colorado at Boulder have developed a small spacecraft called the Drag and Atmospheric Neutral Density Explorer (DANDE). This small, spherical satellite addresses important needs of the defence and civilian community by measuring quantities which are crucial to the determination of atmospheric drag on spacecraft. DANDE is a collaborative project between the Colorado Space Grant Consortium and the Aerospace Engineering Science department. Top level science requirements, defined for the atmospheric measurements, are presented in Table 4.2.1.

The DANDE spacecraft addresses the aforementioned measurement challenges in its design (by having a determined coefficient of drag, cross‐sectional area, and by allowing for the removal of the effects of in‐track winds). In January of 2009, the satellite team won the national University NanoSat Program (UNP) (Franke et al. 2006) competition and was the first University NanoSat winner with a mission focused entirely on collecting scientific measurements. The spacecraft is under 50 kg in mass and 46 cm in diameter conforming to EELV Secondary Payload Adapter (DoD, 2001) requirements. The primary instruments, accelerometers and WTS, are aligned together to enable velocity vector scanning in the nominal attitude state. The instrument locations and relative sizes are shown in the left panel of Figure 4.2.2.

DANDE is spin stabilized around the orbit normal vector, meaning the instruments aligned with the ``equator'' of the spinning sphere will scan the velocity vector at a predicable rate. At the worst case solar maximum density conditions and a nominal 350 km circular, near‐polar orbit, the spacecraft is expected to spend approximately 3 months in orbit before re‐entry (Wilde 2008).

Figure 4.2.2: DANDE spacecraft assembly with one hemisphere removed (left).The DANDE spacecraft showing spin orientation (right).

In order to fully characterize spacecraft drag Table 4.2.2: SWAP envelope of the DANDE instruments. equation, the spacecraft velocity will be estimated Accelerometer Wind and using radar tracking and orbit determination while Subsystem Temperature the acceleration forces will be measured using the Spectrometer unique accelerometer system developed at the type modulated electrostatic drag signal deflection University of Colorado, Boulder (Pilinski 2009). The analyzer in‐situ horizontal wind vector will be measured mass 1.3 kg 1.6 kg using the DANDE Wind and Temperature volume 10.2 x 11.4 x 10.2 x 10.2 x 6.4 cm 7.6 cm Spectrometer (Pilinski 2009). In addition to power 4.9 W 2.2 W measuring drag with accelerometers, the data rate 0.07 Mbits/day 6.99 spacecraft will be tracked by radar and orbit‐ Mbits/day averaged drag acceleration will be generated from

orbital fitting for validation of the accelerometer measurements. Table 4.2.2 summarizes the SWAP envelopes of the two instruments carried by the DANDE spacecraft. DANDE is due for launch in mid 2012.

4.3 LowFlying Spacecraft (LFSC) Daedalus

Theodoros Sarris, Democritus University of Thrace

In the framework of the ESA‐Greece Task Force Programme, the LFSC‐Daedalus mission for the exploration of the upper atmosphere was studied by a consortium of institutes, led by the Space Programmes Unit of the Athena Research Centre in Athens. The mission goals of LFSC‐Daedalus are to perform multi‐point, in‐situ measurements in the largely unexplored lower thermosphere/ionosphere.

LFSC‐Daedalus comprises a main satellite, equipped with state‐of‐the‐art instruments for electrodynamics measurements, and four CubeSats accommodated in PPODs onboard the main satellite. The main satellite is launched into a highly elliptical polar orbit, with inclination of 80o perigee at 150 km; various apogee altitudes between 1000 and 10,000 km have been examined through a trade‐off analysis considering mission lifetime, radiation and launch costs. During the mission lifetime the main satellite performs four ‘excursions’ down to a perigee of 120 km, using its cold‐gas Fig 4.3.1 Main LFSC‐Daedalus satellite and propulsion system. At each of these ultra‐low released CubeSat perigee passes a CubeSat is released. Thereafter, the perigee of the main satellite is raised again to safe altitudes. The perigee altitudes of the CubeSats gradually decay due to atmospheric drag, providing second point measurements down to 95‐100 km altitude, where it is expected to stop operating.

In order to achieve two‐point measurements, which are critical in studies involving the gradients of field‐ aligned currents and the response of the thermosphere and ionosphere to energy inputs from above, conjunctions along the same flux tube are required between the main satellite and the sub‐satellites. As the different perigee altitudes and drag coefficients between LFSC and the sub‐ satellites lead to different orbital periods, extensive simulations were performed in order to identify the number and frequency of conjunctions; this involved tracing the field lines through each location Fig 4.3.2 Orbits of main satellite (green) and of the main satellite and recording instances when released CubeSat (red) the field lines passed within the vicinity of the CubeSat. These simulations have shown a considerable amount of conjunctions, establishing the impact that such a mission would have in resolving key questions in MLTI science. 4.4 Sounding Rocket experiments F.‐J. Lübken, IAP, Kühlungsborn, Germany

Stratospheric balloons can reach altitudes up to about 45 km. Above that altitude, sounding rockets are the only means for in‐situ exploration of the MLT region (Mesosphere/Lower Thermosphere). Sounding rockets are also used for UV or X‐ray astronomy which requires being above the bulk of the Earth atmosphere, or for carrying out experiments under microgravity conditions. These applications are presumably of less relevance for QB50. Sounding rockets are advantageous for some research due to their low cost and their ability to conduct research in areas inaccessible to either balloons or satellites. Furthermore, the lead time for experiments is very short and payloads can be developed in only about six months. Sounding rockets are also used as test beds for equipment that will be used later in more expensive satellite missions. The smaller size of a sounding rocket also makes launching from temporary sites possible allowing for field studies at remote locations, even in the middle of an ocean or at high latitude Antarctic sites.

The rockets are used to carry scientific instruments or microgravity payloads to altitudes ranging from a few tens to appr. 1500 km altitude. The first sounding rocket carrying a scientific payload (a cosmic ray detector) was the Aerobee A‐5 launched on 5 March 1948 reaching an altitude of 117 km. Since then, an estimated 1000 sounding rockets have been launched with apogees >120 km. Of these 1000 sounding rockets, approximately 500 were used for the in‐situ exploration of the lower thermosphere/ionosphere. Nowadays, each year about 30 sounding rockets are launched with apogees in the MLT region.

The rockets can be divided into two parts: the payload and the solid‐fuel rocket motor. After the launch, when the motor has used up its fuel, it separates from the payload and falls back to Earth. The payload continues into space and begins making measurements. The rockets are launched on parabolic trajectories; the total flight time is typically 10‐30 minutes. Most of this time is spent around apogee.

The payload mass on sounding rockets ranges from a few kg up to some hundred kg which includes one or several scientific experiments, an electrical power subsystem, on‐board data handling, housekeeping, a telecommunications subsystem, and sometimes sea or land recovery system for the payload.

Instruments on sounding rockets provide measurements in the lower thermosphere/ionosphere along a single column (nearly vertical profile) during their ascent and descent phases. It takes a sounding rocket and the separated payload only a few minutes to traverse the lower thermosphere/ionosphere (100‐300 km) during its ascent and another few minutes during its descent. The instruments on board the payload make measurements are only during a fraction of this traversal, i.e. in the region of highest interest as defined by the scientific objectives of the mission. Up to now, the total measurement time offered by all appr. 500 sounding rockets for in‐situ measurements in the lower thermosphere/ionosphere taken together over the last 60 years is only on the order of one day.

Microgravity payloads are equipped with a parachute which is opened at lower altitudes allowing a soft landing. The payload can then be recovered and investigated and data stored on board can be analysed.

Compared with QB50 CubeSats, payloads on sounding rockets for in‐situ exploration of the lower thermosphere/ionosphere have the following advantages

- large payload carrying capability, allowing to fly a much wider range of experiments, - no power limitations as the battery size is determined by the requirements of the payload, - a high data rate (20 Mbps or more) and, therefore, excellent temporal and spatial resolution, - experiments can be modified (improved or repaired) until shortly before the launch, - quick turnaround time which allows to modify and adapt the payload for the next flight according to new results. It takes only a few years from project start until scientific data are available which makes sounding rocket experiments an ideal tool for students. and the following disadvantages

- short measurement time in the region of interest, - measurements only along a single column (vertical profile), i.e. no measurements as a function of latitude or longitude, - no multi‐point measurements.

Instruments for exploring the lower thermosphere/ionosphere on sounding rockets

There are a large range of instruments which have been employed on sounding rockets. The following list provides only a tentative and incomplete overview. Some of these instruments are still under development and have not been flown yet. However, they are of interest for QB50. • Neutral mass and ion mass spectrometer • Magnetometer for magnetic field measurements • Langmuir probe for electron temperature and density measurements • Particle detectors for charged particles and, since recently, also for neutral particles, including spectral mass information (e. g. ECOMA, MASS) • Measurement of total density and its fluctuations (e. g. CONE) • Resonance lamps for the detection of reactive trace gases (atomic oxygen, water vapour, etc.) • PHLUX (Pyrometric heat flux experiment) • FIPEX (Flux‐Φ‐Probe EXperiment) for measuring atomic oxygen and other gases • Wind measurements by ground tracking of a luminescent cloud created in the MLT region. • Tracking foil clouds or falling spheres to measure winds and neutral densities • Measurements of plasma components (electrons, protons) at various energies • Electric field measurements, AC and DC components • Absolute electron density measurements by Faraday rotation • Plasma diagnostic probe for electron perturbation measurements • Multi‐spectral imaging of noctilucent clouds • MASS (Mesospheric Aerosol Sampling Spectrometer)

Tables 4.3.1 and 4.3.2 provide a list of some examples of sounding rocket motors and launch sites used today. The list is not exhaustive and does not claim to be complete. Number Country Sounding rocket of apogee payload stages Brasil VS‐30/Orion 2 434 km VSB‐30 2 270 km 400 kg VS‐40 2 650 km 500 kg Canada Black Brant V 1 250 km 270 kg Black Brant IX 450 km 170 kg Black Brant X 3 1300 km 68 kg Black Brant XI 3 590 km 250 kg Black Brant XII 4 1500 km 135 kg India RH‐300 Mk II 1 RH‐560 Mk II 2 334 km 100 kg Japan S‐310 1 210 km 50 kg S‐520 1 300 km 100 kg SS‐520 2 1000 km 140 kg Russia MR‐20 250 km 130 kg USA Nike‐Improved Orion 2 190 km 68 kg Castor 4B (used for MAXUS) 1 720 km 850 kg Terrier Malemute 2 700 km 90 kg Terrier‐Improved Orion 2 200 km 45 kg Terrier Lynx 2 380 km 50 kg Terrier Oriole 2 340 km 360 kg Table 4.3.1: Examples of sounding rocket motors in use today reaching a ceiling > 120 km.

Country Launch site Location

Norway Andøya Rocket Range 69.29°N, 16.02°E Sweden Esrange 67.89°N, 21.10°E USA Poker Flat Research Range 65.13°N, 147.48°W Russia Plesetsk Cosmodrome 62.71°N, 40.29°E USA Kodiak Launch Complex 57.44°N, 152.34°W Russia Yasny Cosmodrome 51.21°N, 59.85°E Russia Kapustin Yar Cosmodrome 48.58°N, 46.25°E Kazakhstan Baikonur Cosmodrome 45.96°N, 63.35°E China Jiuquan Satellite Launch Center 41.12°N, 100.46°E China Taiyuan Satellite Launch Center 39.14°N, 111.97°E USA Wallops Flight Facility 37.85°N, 75.48°W USA Vandenberg Air Force Base 34.77°N, 120.60°W Pakistan Tilla Satellite Launch Center 33.40°N, 73.30°E Israel Palmachim Air Force Base 31.88°N, 34.68°E Japan Uchinoura Space Center 31.25°N, 131.08°E Japan Tanegashima Space Center 30.39°N, 130.97°E USA Cape Canaveral Air Force Station 28.47°N, 80.56°W China Xichang Satellite Launch Center 28.25°N, 102.03°E Pakistan Sonmiani Satellite Launch Center 25.19°N, 66.75°E India Satish Dhawan Space Centre 13.74°N, 80.24°E India Thumba Equatorial Rocket 8.53°N, 76.87°E Launching Station Brazil Alcântara Launch Center 2.32°S, 44.37°W Table 4.3.2: Active launch sites for sounding rockets reaching a ceiling >120 km, ordered by latitude, most commonly used launch sites in bold

4.5 Armada A. Ridley, University of Michigan

Armada is a proposed NanoSat constellation mission to study the thermosphere reaction to energy input across all scales. It is a collaboration between UM, NRL, GSFC, SNC, Wallops, UC, UAF, NOAA and APL

Armada’s science objectives are:

• To determine the interplay between local, regional and global scales in defining the thermospheric response to magnetospheric inputs.

– How does the plasma‐neutral coupling drive thermospheric change over various scales?

– What part of the thermospheric response occurs in the form of propagating waves and how do the waves and mean circulation mutually interact?

• To quantify how winds and composition conspire to produce the observed density response over regional and global scales.

– What is the relative importance of thermal expansion, upwelling and advection in defining total mass density changes?

– How do neutral winds, temperature and composition serve as drivers for global ionospheric variability?

• To understand how ions and neutrals interact to produce meso‐scale structures and dynamics

– What are the in‐situ processes that lead to neutral and plasma structures characteristic of the low‐latitude ionosphere?

– How do upward propagating tides and planetary waves contribute to the variability of the low latitude IT system?

Armada will comprise a constellation of 48 nano‐satellites each of mass <6 kg deployed into 6 orbital planes with equal numbers of satellites in each. Each satellite will carry an identical and highly effective instrument suite which will thereby provide global coverage and allow investigation of the evolution of the thermosphere.

Figure 4.5.1. shows an Armada satellite with deployed solar panels. Aerodynamic forces will provide functional redundancy with attitude control and give additional power. Note the front end point into the RAM direction.

Attitude Determination and Control will be provided with momentum wheels, magnetic torquers, sun sensors, a star tracker and a GPS. Fig 4.5.1 Armada satellite

Communications will be S‐Band for downlink and UHF for uplink.

The Armada science payload will be the GSFC/NRL Wind Ion Neutral Composition Suite (WINCS) which will measure:

• Neutral density, temperature and wind • Ion density, temperature and drift • Ion and neutral composition

A single deployment vehicle will insert the NanoSats into their orbits and planes, each with an inclination of between 81 and 83.5°. The nominal mission lifetime is 2 years and the initial orbit height will be ~ 500 km, (slightly elliptical).

Deployment will involve the release 8 satellites at the same time with 3 m/s delta‐V, spaced out in angle, giving slightly different orbital periods. 5 maneuvers will make slight adjustments to the inclination providing angles of: 81.0°, 81.8°, 82.2°, 82.3°, 82.7°, 83.5°. This allows complete longitudinal deployment in ~1 year and yet keeps 2 planes within ~10° over the entire mission. 82° is chosen to maximize the amount of time on the auroral oval. By putting the planes at slightly different inclinations they precess at different rates, slowly spreading them out.

One of the most challenging aspects of Armada is how to get all of the data down to the ground. Each satellite generates roughly 300 MB/day, or almost 15GB/day across the constellation. On‐board data compression will be essential. The situation is also mitigated with 32 GB of onboard storage, so data can be cached for ~100 days. It will be possible to download 6s averages for entire mission and higher cadence data for events. NRL has developed autonomous, low cost ground stations to handle constellations like this. These can be distributed strategically to allow near real‐time downloading of data.

5. QB50 Candidate Sensors

5.1 Introduction Information regarding sensor technologies was gathered by the SSWG and is presented below in terms of potential QB50 payload instruments. The technologies covered were:

• Accelerometer • Energetic Particle Sensors • FIPEX (atomic oxygen sensor) • GPS • Ion Mass Spectrometer • Langmuir Probe • Laser Retroreflector • Magnetometer • Neutral Mass Spectrometer • Spherical EUV and Plasma Spectrometer (SEPS) • Thermal Sensors (Incl. Bolometric Oscillation Sensor) • Wind Ion Neutral Composition Suite (WINCS – on proposed Armada mission)

The information covers the following aspects:

• Scientific justification • General description and principle of operation • Performance • Mass • Power • Data rate • Operations and Commanding • Special requirements • Heritage (e.g. Technology Readiness Level) • Cost • Development schedule

Information on mass, power, data rate, TRL and cost are summarised in Table 5.14.1

5.2 Ion and Neutral mass spectrometer D. Kataria, MSSL

Science case Atmospheric variables are often derived using assumptions about the composition and the mean molecular mass. However, localized composition changes can change production and loss processes and be the cause of anomalous measurements that are derived from mean molecular masses. For example, above 300 km atomic oxygen is the dominant neutral constituent. Heating that causes upwelling of molecular N2 and O2 increases recombination and changes the mean molecular mass resulting in an overestimation of the ion temperature, amongst other variables measured by ISR radars. In‐situ measurements carried out by QB50 of the density and composition of the ions and neutral particles and their spatial variation will provide key measurements for significantly improved determination of atmospheric variables and provide key inputs for atmospheric models.

Description The Ion and Neutral Mass Spectrometer (INMS) is a miniaturised analyser designed for sampling of low mass ionised and neutral particles in the spacecraft ram direction with the instrument resolutions optimised for resolving the major constituents in the lower thermosphere, i.e., O, O2, N2.

The key sensor components consist of a collimator/ion filter, an ioniser and a charged particle spectrometer. Particles enter the aperture into the ion filter region where charged particles can be rejected. This is followed by a series of baffles for collimation and further charged particle suppression. Collimated neutral particles are subsequently ionised in the ionizer by a 50 eV electron beam Figure 5.2.1: Schematic of the followed by mass selection in the analyser. With an energy principle of working of the resolution of 3%, the analyser will provide clean separation of the INMS major constituents. The spectrometer can be operated in different modes, optimised for ions or neutral particle analysis.

Performance Of the two critical components, the ionizer and the analyser, proof‐ of‐concept/prototype devices have been breadboarded and are undergoing detailed testing. Figure 5.2.2 shows the prototype of the miniaturised analyser (in gloved hand) along with MSSL’s conventional, albeit state‐of‐the‐art, plasma analyser. Furthermore, ChaPS (Charged Particle Spectrometer), a complete charged particle instrument, has been developed for in‐flight demonstration on TechDemoSat, a UK technology demonstration Figure 5.2.2: Miniaturised charged satellite to be launched in the first quarter of 2013. Performance particle analyser along with the Improved Plasma Analyser characteristics of ChaPS are given in Table 5.2.1.

Calibration Performance “standardisation” (performance characterisation and calibration, including thorough characterisation of performance variability) of the sensors across the network is a key requirement for QB50. MSSL has considerable expertise with calibration and qualification of large numbers of sensors and the instrumentation activities are supported by state‐of‐the‐art characterisation and

29 calibration facilities for this as well as qualification facilities including vibration and thermal vacuum. For example, MSSL supplied 10 analysers each for Cluster 1 and 2 and recently celebrated 10 years of successful operation of the analysers in space with all 10 instruments continuing to provide excellent performance. The INMS sensors will be calibrated before delivery and will include cross‐ calibration of the sensors across the network.

Characteristic Key view direction Spacecraft Ram Energy range 0.1 to 28 eV Energy resolution <3 % Elevation resolution 5° Azimuth resolution 5° Geometric factor ~1x10‐6 cm2 sr eV/eV Time for full distribution 1 s Table 5.2.1: Performance characteristics of ChaPS

Operations and Commanding The control and overall operation of the instrument will be carried out through simple command sequences executed by the CubeSat on‐board computer/microcontroller. It is envisaged that in‐ flight, the INMS will have two modes, standby and science, with the data handing and control carried out over the I2C interface, nominally operating at 100 kHz clock speed. All the parameters required for control and operation of the instrument will be contained within the command. During in‐flight operations, it is expected that several such sequences would be stored in memory and run at different times in an orbit over a chosen time scale and the data acquired stored on‐board for downlink.

Special requirements Key requirements for the INMS is accommodation of the sensor such that its aperture views the spacecraft ram direction and for the CubeSat to provide attitude control to within ±10° with pointing knowledge of ±5°. This will require a clear face of the CubeSat with no solar panels. INMS will use the 3.3V and 5V available from the CubeSat bus for powering the instrument and the I2C interface for command and control.

Heritage MSSL has long been recognized as a centre of excellence in the provision of instrumentation for space missions and exploiting the data returned by them with involvement in all major missions conducted since the 1970's by ESA and NASA. One of the key areas of expertise is in the provision of particle instruments. The laboratory is currently developing highly miniaturised low‐energy particle analysers, sensors that are ideally suited for CubeSat missions and as described above, has recently built the ChaPS instrument shown in Figure 5.2.3 for in‐flight demonstration on Figure 5.2.3: Photo of the the UK’s TechDemoSat mission. The ion sensor on ChaPS meets all the ChaPS instrument. requirements of the QB50 mission for studies of the ion environment. ChaPS has been built and calibrated and has successfully undergone vibration and thermal vacuum testing, placing it currently at TRL 7/8. The instrument is scheduled for delivery in February 2012 with launch in Q1 2013.

Development schedule

The schedule for the development of the INMS is provided in Table 5.2.2

No. Development activity Time [months]

1 Development of Ion sensor Completed

2 Ionisation source proof‐of‐concept testing Completed

2 Ionisation source breadboard development 2

3 Development of INMS prototype 6

4 Development of flight sensor 4

5 Production and calibration of flight sensors 6‐9

Table 5.2.2: Schedule for the development of the INMS

5.3 FIPEX: Insitu Atomic Oxygen Measurement in LowEarth Orbit T. Schmiel, S. Fasoulas, A. Weber, TU Dresden

Science case FIPEX of TU Dresden (Flux‐Φ‐Probe‐Experiment) is able to distinguish and measure the time‐resolved behaviour of atomic and molecular oxygen as a key parameter of the lower thermosphere. Atomic oxygen is the dominant species in these regions and therefore its measurement is crucial in the validation of atmosphere models. Moreover, erosion of spacecraft surfaces due to interaction with atomic oxygen is a serious concern and merits in‐situ study in its own right.

Figure 5.3.1 Sketch of FIPEX sensor principle with a third reference electrode (Schmiel, 2009; Hammer et al 2009)

Description The measurement is based on solid oxide electrolyte micro‐sensors. The working principle of the developed oxygen sensors as shown in Figure 5.3.1 is based on the ion conductivity of ceramic materials. For oxygen conducting solid state electrolytes, e.g. yttrium‐doped zirconia, the conductivity starts at high temperatures and so the sensor operates at an elevated temperature, heated by an electrical resistance. Oxygen is “pumped” from one electrode to the other by an applied direct voltage and in accordance with Faradays’ law; the measured current is proportional to the mass flux by electrolysis. To distinguish between atomic oxygen (O) and molecular oxygen sensor elements with different cathode materials are used.

The design of the FIPEX sensor element as developed for the CubeSat SOMP is simple and compact with a relatively low power consumption (operating only one sensor at a time) and high sensitivity in high vacuum conditions. The sensor geometry is 20x3.5x0.5mm with a heater power of less than 1.6 Watts at 660°C. Figure 5.3.2 shows the sensor for SOMP and its integration into the sensor unit periphery leading to a complete unit mass of 15 g (without electronics and harness). See Table 5.3.1 for the key design parameters used for the design of the CubeSat SOMP.

Sensor unit Dimension 36 x 30 x 12 mm³ No. of sensors 2 Type of sensors Atomic Oxygen Mass 15 g (excluding harness) Field of View ~180 deg (free flow)

Heating Power < 1.6 W Electronic / PCB Sensor 1 + 1 spare, no parallel operation Dimension#1 80 x 100 x 10 mm³ Power (includes sensor 12 V: 2700 mW#2 heating power) 5 V: 100 mW 3.3 V: 200 mW Mass 70 g (excluding harness) Table 5.3.1 Characteristic of the SOMP FIPEX Experiment (based on CDR report, Deckert et al 2011)

#1 Different dimension like PC/104 is possible;

#2 This special power line is available in the SOMP CubeSat bus. Alternatively this power line can be included in payload electronics for about 15 g extra mass. In this case a 8 V or the 5 V power line is used instead.

Figure 5.3.2 Oxygen sensor for CubeSat SOMP (left) (Hammer et al 2009) and the sensor unit (right)

Performance FIPEX is able to distinguish and measure atomic and molecular oxygen at very low ambient pressures. A typical characteristic of different oxygen sensors in ultra high vacuum can be seen in Figure 5.3.3a. It is possible to measure oxygen partial pressures down to 10‐10 mbar (e.g. particle density 106 atoms/cm³) with a resolution of better than 2x10‐10 mbar (Schmiel 2009). A typical result of the difference in densitites between atomic and molecular oxygen, taken during the flight of FIPEX on the ISS mission is depicted in Figure 5.3.3b.

For development and calibration, the Atomic Oxygen Exposure Facility of the TU institute is used, that can produce high vacuum conditions down to 10‐9 mbar total pressure. The residual species are determined by a SEV quadrupole mass spectrometer. For the simulation of the atomic oxygen environment, microwave induced oxygen plasma dissociated to atomic oxygen is expanded through a small aperture into the test chamber.

Figure 5.3.3a Typical characteristics of different oxygen sensors in ultra high vacuum (Schmiel 2009)

Figure 5.3.3b Difference in density between atomic (above) and molecular (below) oxygen taken during the flight of FIPEX on the ISS mission, data taken on 23‐24th April 2008 (Schmiel 2009)

Operations and Commanding The commanding and data handling requirements for FIPEX are not demanding. FIPEX is controlled by an MSP430 micro controller that is linked to the central data handling and control of the CubeSat. There are no special mode requirements for FIPEX and so no special or complex commanding requirements. There will be a measurement mode and a standby/safe mode. If not in use the sensor can be deactivated. It is only necessary to be able to set a schedule for measurement and verify the correct functionality through appropriate housekeeping. Direct commanding is not required during normal operation. The assumed data rate for one sensor will be about 16 kByte per hour. It is not required that FIPEX measures the whole time.

Special requirements The FIPEX design for SOMP assumes that a 12 V power line is available, which is not a common power line for CubeSats. Alternatively, this can be included in the sensor electronics. Additionally, it is assumed that a battery is provided by the CubeSat bus which is able to buffer peak power consumption during heating of the sensor elements. If not available for QB50, a battery can be included in the sensor electronics. The sensor needs to be in free flow and to determine the actual flux, the attitude of the satellite with respect to its direction of motion needs to be known. For SOMP the required accuracy of attitude determination is 5°.

Heritage FIPEXonISS was an experiment launched on STS‐122 (1E) Shuttle flight on 7 February 2008 and deployed on the Columbus External Payload Facility on the platform EuTEF (European Technology Exposure Facility). It provided measurements of the time‐resolved behaviour of atomic and molecular oxygen. The next natural step would be a temporally and spatially resolved measurement which is possible with QB50. FIPEX on QB50 consists of three parts: sensors, sensor unit and electronics. Due to the heritage with FIPEXonISS, the experiment has a TRL of 9. Based on the experience of the ISS experiment, the design was changed to the requirements of the SOMP mission. The SOMP design has currently a TRL 6. TRL 8 will be reached in the autumn of 2011. The launch is planned for the beginning of 2012. The TRL are summarized in Table 5.3.2.

Part Heritage TRL Sensor Sensor principle: FIPEXonISS [2] TRL 9 Sensor typ: TRL 8 Sensor unit Possible: FIPEXonISS [2], TRL 9 CubeSat SOMP [4] TRL 6 Electronics Possible: FIPEXonISS [2], TRL 9 CubeSat SOMP [4] TRL 6 Table 5.3.2 Technology Readiness Level (TRL) and Heritage

Figure 5.3.4 Test PCB for FIPEX electronics (130 x 90 x 10mm³)

Development schedule QB50 is likely to have different requirements for FIPEX than SOMP and the development work packages are shown in Table 5.3.3 below.

No. Work Package Name Time Remarks [month] 1 Development of an Engineering Model for FIPEXonQB50, 2‐9 depending on adaptation including: effort and requirements − Clarification of Requirements − Adaptation for mass production − Adaptation of the sensors to the orbital height − Adaptation of sensor unit according to QB50 requirements − Adaptation of measurement electronics Assembly of Engineering Model 2 Qualification of Housing and Electronics 3 if required due to design (Vibration, Thermal, …) changes or launcher specification 3 Manufacturing of Flight Hardware 4‐5 includes FM electronics (Reproduction of EM) and housing but EM sensor 4 Production of the sensors, calibration 6‐8 for assumed 70 sensors, (delivery of flight sensors shortly before final integration) Table 5.3.3 FIPEX Development Schedule

5.4 MultiNeedle Langmuir Probe T. A. Bekkeng and J. Moen, University of Oslo

Science case The University of Oslo multi‐Needle Langmuir Probe system can provide absolute electron density measurements at ~1 m spatial resolution for a 320 km orbit, and thereby map regions of F‐region plasma turbulence of particular relevance for communication and GNSS signals in polar and equatorial regions. A Langmuir probe will provide an absolute measurement of plasma density and so provide a means of cross‐calibration for an ion‐mass spectrometer.

Description The m‐NLP system consists of one DAQ PCB (80.00 x 82.60 mm, but can easily be adapted to PC104 standard size), and four thin booms to support the probes, as seen in Figure 5.4.1.

Figure 5.4.1: Three of the four probes incl. booms shown, mounted in the CubeSTAR CubeSat.

The signal flow in the m‐NLP system is shown in Figure 5.4.2.

Figure 5.4.2: Block diagram of the instrument signal flow

All four probes are biased at different bias potentials, with respect to the platform potential.

Performance Current measurement range 3 decades (i.e. 1 nA to 1 µA), but adjustable by in-flight automatic gain control Electron density range 108 m-3 to 1012 m-3 (adjustable to match mission requirements)

Accuracy 16 bit raw data, but downsampled to 8 / 10 / 12 bit data product Sampling rates Up to 7 kHz, adjustable by uploadable selection commands Table 5.4.1 Langmuir Probe performance summary

Calibration strategy and inorbit approach Pre‐flight calibration is done by EGSE equipment supplied by the University of Oslo. In‐orbit calibration is done by running sweeps on all probes i.e. once a week, to calibrate for effects from probe contamination etc.

Data rates and options and operational modes Mode: Complete scientific coverage ~1.25 MB per orbit On‐board processed: 100% Duty cycle: Mode: Partial scientific coverage ~312.5 kB per orbit On‐board processed: 25% Duty cycle: Mode: Irregularity survey mode 8.6 kB per orbit 100% Duty cycle Table 5.4.2 Langmuir Probe data rates and modes

Required commanding The following commands must be sent over the I2C bus:

• Selection of which operational mode to be active • DNEL signal to ensure proper storing of measurement data before system power down

Any special requirements – thermal stability, booms and mechanisms, … Four spring‐loaded booms to support the probes have to be placed, optimally, at the end of the CubeSat facing the direction of flight. To avoid wake effects, the CubeSat should have an attitude control within ±25° along the axis of motion.

Heritage (TLR with evidence) The m‐NLP system has been flown successfully on four sounding rockets between Dec. 2008 and Dec. 2010. For all four flight the m‐NLP system provided high quality data of electron density, and a one second sample of data is shown in Figure 5.4.3, showing small‐scale density structures.

Figure 5.4.3. One second of electron density data from the ICI‐2 campaign

The TRLs for the different parts of the m‐NLP system are given in Table 5.4.3.

Electronics Probes Booms TRL 6 ‐ System/subsystem model TRL 6 ‐ System/subsystem model TRL 3 ‐ Analytical & experimental or prototype demonstration in a or prototype demonstration in a critical function and/or relevant environment (ground or relevant environment (ground or characteristic proof‐of‐concept space) space)

Table 5.4.3 Langmuir Probe sub‐system TRLs

Development schedule None of the parts used in the system design have significant lead times

Development stages, sounding rocket version of the instrument, serving as the prototype for the satellite version Version two of the mNLP system for sounding rocket (8‐NLP), August 2011 ready for final integration Launch 1 of the 8‐NLP system ICI‐3 sounding rocket, Dec. 2011, Svalbard Launch 2 of the 8‐NLP system NASA 36.273 MICA Sounding rocket, Poker Flat, Alaska Table 5.4.4 Langmuir Probe general development milestones

Development stages, CubeSat version Breadboard model of preamplifier stage November 2010 Testing of probe prototype in ESTEC plasma chamber February 2011 PCB design and production Finalized August 2011 FPGA programming Finalized October 2011 Functional tests January 2012 Environmental / vibration testing Finalized by the end of 1st quarter of 2012 Launch of mNLP system on the CubeSTAR CubeSat 2013 Table 5.4.5 Langmuir Probe QB50 specific development milestones

All subsystems have been partly of fully tested on prototype level, except of the on‐board memory for the experiment.

5.5 Magnetoresistive magnetometer P. Brown, Imperial College London

The main science goals of the magnetometer are the detection of Equatorial Spread F events which are typically accompanied by field enhancement across regions of low electron density. It is envisaged that the magnetometer would also provide context information for the particle measurements and could be used as an input for spacecraft attitude determination.

Description The sensors are implemented as plastic or ceramic packages and co‐located with drive electronics in a potted sensor head ideally at the end of a rigid boom. The sensor head has a mass of around 20 g and 10 cm3 volume. This is within the typical payload resource envelope of a CubeSat mission and may be compared with a 200 g fluxgate sensor encompassing a volume well over 100 cm3. The sensor delivers the three components of the magnetic field (nominally in the range 0‐10 Hz although up to 100 Hz is possible) plus a temperature measurement. Ideally, two three‐axis sensors would be included for differential (‘gradiometer’) measurements of the platform‐generated field. The outboard (OB) sensor is mounted at the end of the boom and the inboard (IB) sensor fitted at the base of the boom or directly on the magnetometer electronics card. The sensors connect to the electronics card via a lightweight harness.

Each single axis sensor is implemented as a Wheatstone bridge. The bridge output voltage varies in response to a change in current flow modulated by the field dependent resistance. The control electronics is implemented as a closed loop system featuring dedicated bipolar driving. The closed loop control improves the overall linearity and limits sensitivity drift. The MR driver ensures the sensor maintains low noise operation by holding the sensor at its optimum position on the sensor transfer function.

Figure 5.5.1 shows the exposed miniaturised sensor head and an example of a complete magnetometer (PC104 PCB, sensor harness and enclosed sensor head, protected by an Aluminium housing) as implemented for a 3U CubeSat mission. Other configurations are possible depending on specific mechanical requirements. The sensor head houses the AMR sensors and associated drive circuitry implemented as a thick film hybrid on a LTCC substrate. The drivers are included in the head to limit power loss in the sensor harness. The sensor is typically potted in epoxy to protect the sensor components and provide thermal and mechanical stability.

Figure 5.5.1 Left: Miniaturised magnetometer sensor head; Right: Complete magnetometer

The magnetometer requires a single electronics card compatible with a standardised CubeSat platform of dimension 9 cm by 9.6 cm and 1 cm height clearance either side. Options to reduce the electronics volume envelope are possible if common control, power and data processing electronics are envisaged.

The instrument summary is shown in Table 5.5.1. The instrument cadence and duty cycle are dependent on the requirements of the specific configurations but the sensor measurement bandwidth is between 0‐200 Hz and the instrument hardware configuration is compatible with both always‐on and duty cycled operations.

Characteristic Requirement

Mass 15 g (sensor – includes un‐potted Aluminium mount) 25 g (sensor –potted) 80 g (electronics) 12 g (1 m harness) Power 240 mW (attitude mode), 500 mW (science mode) Volume Sensor head 10 cm3, Electronics 173 cm3 (9 cm x 9.6 cm x 2 cm) Sensitivity 10 nT (attitude mode), <2 nT (science mode) Cadence 1 vector/s (attitude mode), 10 vectors/s (science mode)* Telemetry <100 bits per second Mechanical A rigid boom of dimension at least 30cm is desirable in order to limit spacecraft contamination of the magnetometer measurement Pointing No active requirement but attitude knowledge required to recover magnetic vector direction. Noise Density 150 pT Hz‐½ at 1 Hz Temperature ‐50°C to +120°C *Other cadences possible Table 5.5.1 Magnetometer summary

Calibration Imperial’s 3‐axis, high quality Helmholtz coil calibration facility allows us to calibrate the sensor head against a high end fluxgate sensor on the ground. We will perform this calibration before delivery, including a characterisation of temperature effects, in order to maximise the precision of the on‐ orbit data.

Performance The instrument has a sensitivity of better than 2 nT and a Noise Spectral Density (NSD) of less than 150 pT/√Hz above 1 Hz. Offset drift is of order 1 nT/oC. This is expected to improve following further technical developments on the sensor design.

Special requirements A boom of length >20 cm is required. Solar panels should be wired so as to minimise the magnetic field at the magnetometer.

The unit requires one digital (3.3 or 5 V) and one analogue (12 V or 15 V) line. The data interface to the bus can be either analogue or digital depending on specific requirements. The implementation shown above has an SPI digital interface incorporating a 24‐bit ADC sampling each component at 16 Hz. The instrument range is ±50,000 nT. The ADC includes several IO lines and these will be utilized

41 to implement the command and control lines from the spacecraft on‐board computer. Imperial College can verify CubeSat interfaces through test with a development board (Pumpkin Inc www.cubesatkit.com).

Heritage The Imperial College London Space and Atmospheric Physics group has a long heritage in the provision of magnetometer hardware for space science missions. The development of a space‐grade solid state magnetometer has been undertaken as part of a UK/ESA funded research program resulting in a design of a tri‐axial magnetometer using sensors based on Anisotropic Magneto‐ resistance (AMR). The first magnetometer flight model was recently delivered for integration on the NSF CINEMA CubeSat which is due for launch in summer 2012 and is currently undergoing an environmental test campaign. The magnetometer is currently at TRL 6.

5.6 Accelerometer T. Beuselinck, Redshift

Accelerometer data together with an understanding of the drag coefficient (known from geometry, orientation and temperature etc.) and ballistic coefficient enables the calculation of atmospheric density.

Description The candidate accelerometer for QB50 is inevitably constrained by available technologies. Most high‐accuracy accelerometers for space use have been based on the Honeywell QA‐3000 series (or lower‐cost QA‐2000 series) acceleration sensor. Although this sensor is very accurate and stable it cannot be considered to be a quasi‐steady accelerometer in itself. Although the noise of this sensor is higher than the atmospheric drag signal that is to be measured in a typical space mission the very symmetrical nature of this noise allows the recovery of significant measurements using analog or digital filters. However, also the bias drift of this sensor is higher than the signal to be measured. Some strategies have been used to overcome this problem (e.g. by periodically rotating each sensor over 180°or by continuously rotating the sensor to be able to remove a slowly drifting bias drift) with various success. A similar strategy is to be used by the DANDE mission, where the full spacecraft is rotated. RedShift has used a similar strategy to successfully extract very low‐frequency acceleration information from the Foton accelerometer measurements based on measurements of the (non‐ controlled) attitude of this spacecraft.

MEMS sensors offer an appealing alternative to the Honeywell sensor (much smaller, lower mass, lower power). However, up to recently these sensors could not match the performance of the Honeywell sensor. Recent developments claim to match or exceed some of the Honeywell sensor specifications, but these sensors have not yet been used in space. Therefore, a dedicated test campaign would be required to evaluate such sensors for the QB50 application (e.g. using droptower or parabolic flight testing). In addition, these sensors exhibit an important temperature dependency. Therefore, besides thermal compensation (by post‐processing), active thermal control is expected to be necessary.

Performance Based on the available orbit information RedShift estimates that the drag experienced by a typical QB50 CubeSat (without any drag‐enhancing devices, e.g. without deployable solar panels) will fluctuate (within each orbit) between 0.5 µg and 2 µg (in ram direction at 320 km) to >1000 µg at 120 km. The expected fluctuations are due to normal atmospheric density variations in the orbit. In order to acquire scientifically useful data much smaller fluctuations need to be measured (target resolution 50 ng, target accuracy ±100 ng). These fluctuations cannot be reliably measured with any currently available off‐the‐shelf acceleration sensor. In order to measure significant signals a strategy similar to the DANDE mission could be used (increase the CubeSat drag by using deployable items, rotate the sensor or the whole CubeSat around an axis perpendicular to the orbit plane) at the cost of very stringent requirements on the spacecraft design (e.g. attitude measurement and control systems performance). This seems out of scope for the QB50 project. Even if this was possible the additional constraints (e.g. low data downlink volume) would lead to even more stringent requirements (e.g. a fixed pre‐defined angular rate ‐or alternatively high angular stability‐in order to allow for accurate in‐flight data processing and acceleration signal recovery). Moreover,

43 based on RedShift experience the measured signal could be contaminated by other sources than atmospheric drag (e.g. the Honeywell sensor based devices that RedShift built are sensitive to magnetic fields, corresponding to a signal of 1 µg amplitude for the Earth magnetic field variations in orbit). This kind of error in the measurement signal can be interpreted as a scientific result and could be impossible to correct for if only a limited amount of downlink data (the result of extensive on‐ board processing) is available. Although on‐board data processing is certainly possible RedShift knows of several examples where this failed (although the device had been extensively tested on ground the output of a German accelerometer saturated in flight).

Note that a sensor located out of the spacecraft COG will also measure accelerations caused by attitude changes (rotations) and by attitude control system actions (typically up to 100µg). If mechanisms are active in the spacecraft the sensor will measure vibrations (typically mg‐level amplitude at higher frequencies). In ground tests also it is almost impossible to prevent higher‐level vibrations. Therefore any accelerometer requires a significant dynamic range.

Based on this discussion RedShift proposes to trade‐off between two strategies: 1. Using a 1‐axis Honeywell‐type servo‐accelerometer in a rotating CubeSat 2. Using a 3‐axis MEMS sensor set‐up without CubeSat attitude constraints

Scientific relevance to QB50 Measuring the atmospheric drag experienced by a CubeSat by using a objectives micro-accelerometer allows the study of the Earth atmosphere density distribution. Description – including Typical 1-axis micro-accelerometer design based on the Honeywell QA- principle of operation 2000 servo-accelerometer sensor (or similar non-ITAR item). Performance – sensitivity, • Typical target measurement range: >±100 µg (internal device range >± • range, accuracy etc. Typical target resolution: 50ng • Typical target accuracy: <±100ng Calibration – strategy and On-ground calibration on test jig in-orbit approach Mass 0.2 kg Volume 100mm x 100mm x 30mm Power – equivalent primary 1W total required power rails: +15V: <500mW -15V: <500mW +5V: power together with <100mW +3.3V: <500mW secondary power per LV power rail Data rates and options 500kByte per day Operational modes On / Off Required commanding On / Off, angular rate Any special requirements – Adapted mission strategy: • increase the drag using deployable items • thermal stability, booms and rotate CubeSat around axis perpendicular to orbit plane at constant mechanisms, … angular rate Heritage (TRL with Accelerometer: TRL 7 (several Foton missions and one ISS payload evidence) subsystem) To be developed: On-board data processing and mission strategy constraints. Availability Honeywell sensor: ITAR (alternative available) Cost Development: 250k€ Recurrent (per flight-qualified accelerometer): 15k€ Remark: non hi-rel, standard commercial electronics. Mission support: 150k€ (including support in post-mission data analysis) Development schedule Development (up to qualification): 15 months Includes flight model manufacturing, starting after 6months. Development risk – what Low development risk: Accelerometer (based on flight-proven design) are the risks and how are Medium risk: Development of on-board data processing algorithm. High- they to be mitigated. risk: Adapted mission strategy implementation (by CubeSat developers) Table 5.6.1 Option 1: Servo‐accelerometer based solution. All values are estimates

Scientific relevance to QB50 Measuring the atmospheric drag experienced by a CubeSat by using a objectives MEMS accelerometer allows the study of the Earth atmosphere density distribution. Description – including 3-axis MEMS accelerometer design based on European sensor principle of operation technology. Performance – sensitivity, • Typical target measurement range: >±100 µg • Typical target resolution: range, accuracy etc. 50ng • Typical target accuracy: <±100ng Calibration – strategy and On-ground calibration on test jig, complemented with free-fall development in-orbit approach testing and post-flight correction of low-frequency drift (using mathematical models of orbit and attitude dynamics). Mass 0.05 kg Volume 50mm x 100mm x 30mm Power – equivalent primary 0.35W total power Data rates and options 100kByte per day Operational modes On / Off Required commanding On / Off Any special requirements – Adapted mission strategy: • increase the drag using deployable items thermal stability, booms and (preferred) • active temperature control using thermo-electric cooling mechanisms, … modules Heritage (TRL with No flight heritage evidence) Availability European supplier Cost Development: 150k€ Recurrent (per flight-qualified accelerometer): 10k€ Remark: non hi-rel, standard commercial electronics Mission support: 150k€ (including support in post-mission data analysis) Development schedule Development (up to qualification): 9 months Includes flight model manufacturing, starting after 6months. Development risk – what High-risk: MEMS accelerometer without flight heritage (mitigated after are the risks and how are dedicated tests) they to be mitigated. Table 5.6.2 Option 2: MEMS Accelerometer based solution. All values are estimates

Tradeoff conclusions Option 1 has the advantage of an existing flight heritage, but consumes a large portion of the available resources (mass, volume, power) and imposes a dedicated attitude (mission strategy).

Option 2 lacks flight heritage but shows great potential relating to reduction of mission resources. Considering the educational nature of the QB50 project, the challenge of using these innovative sensors can be considered acceptable.

Special requirements Note the potential interference from other CubeSat elements including Attitude Control System jitter and EMC.

Heritage RedShift designed, built and operated several high‐accuracy accelerometers for the Russian Foton spacecraft and was also involved in the post‐flight data processing. Based on this experience the need for a low‐cost quasi‐steady micro‐accelerometer was identified and a first prototype of such a device has been demonstrated on Foton‐M3. The last few years this technology was not developed any further due to lack of funding. Although some recent developments could lead to a resumption of this development, this would not lead to a product in time for the QB50 project.

5.7 GPS receiver

5.7.1 Sérgio Cunha, University of Porto

GPS receivers operate by measuring relative phases of signals transmitted by the GPS satellites situated in circular MEOs of approximately 26,000 km radius. The signals that can be measured are: • Pseudo‐ranges, obtained from the phase of a 1,023,000 chips per second pseudo‐random binary sequence (PRBS) modulated over the carrier frequencies, which are L1 = 1575.42 MHz and 1227.60 MHz. The chip length in space is about 300 m and receivers can measure this signal phase to an accuracy of few meters. The PBRS total length is roughly 300 km in space and any possible ambiguity regarding which multiple of the sequence length is being received is solved by information contained in a 50 bps message stream also modulated over the L1 carrier. • Carrier phases, obtained as the phase of the L1 and L2 carriers relative to a reference signal generated by the receiver and common to all satellite tracking channels. These measurements have only about 1 cm error. Although lock is maintained from one measurement to the next, there is an ambiguity in the initial integer number of cycles for each consecutive set of epochs without loss of lock. It is not possible to solve this integer ambiguity from other measurements (only bound this integer value) using information just from one satellite and one epoch, given the short period (one wavelength, of size 19 cm for L1 and 24 cm for L2). Carrier phases are used to smooth the pseudo‐range measurements from epoch to epoch, reducing the error of the latter to about one meter. Maintenance of lock between epochs is very important for carrier phase measurements, which implies operating the receiver continuously during a period of measurements. • The Doppler measurement is the instantaneous frequency offset from the L1 and L2 carriers relative to the frequencies generated from the clock of the receiver. Solving simultaneously for the receiver clock drift, it allows computing the instantaneous vector velocity of the receiving antenna. The measurements are not exactly instantaneous, as the tracking loop of the receiver contains filters that integrate over some period of time. However, this period is in the order of magnitude of one to ten milliseconds, which for the purpose of what is addressed here is negligible and can be considered instantaneous.

Positioning with GPS The most common way of using a GPS receiver is in stand‐alone mode, computing position and velocity estimates together with the respective timestamps (PVT estimates) from the pseudo‐range and Doppler measurements. The accuracy of the position estimates is in the order of a few meters (say, 10 m); the accuracy of velocity estimates is in the order of centimetres per second. Such error has several components: atmospheric delay (which is the most significant near the surface of the Earth, having less impact in orbit), GPS satellite orbit and clock uncertainties and pseudo‐ range/Doppler measurement noise. It is relevant to mention that, except for the measurement noise, these errors have significant low frequency components. Therefore, low‐pass filtering will have almost no effect in containing them. The next step to improve accuracy with GPS is differential positioning with pseudo‐ranges. Through this method, two GPS receivers located at a distance much closer than the distance between the receivers and the GPS satellites exchange measurements (in this case, pseudo‐ranges). The common part of the errors affecting both receivers is cancelled out, enabling the computation of the vector arm between them with an accuracy of about half a meter. The error source dominating now is the pseudo‐range measurement noise, which has a broad spectrum, and the differences between the GPS orbit and atmospheric errors as seen by the two receivers. The later exhibit low frequency components and are more significant with the distance between receivers.

Further accuracy can be obtained by processing the carrier phase measurements in differential mode. One simple procedure is to compute the evolution of the vector arm between receivers from one epoch to the next. The error of such estimate is in the order of one centimetre for epochs that are close in time. They contain a low frequency drift, due to GPS satellite ephemeris errors and atmospheric perturbation changes. The absolute relative arm between receivers can also be computed from the carrier phase measurements. This implies solving an integer ambiguity problem in search for the unique differential position and, simultaneously, the number of full cycles added to the measurements of each GPS satellite that will result in the minimal residuals, in the vicinity of the pseudo‐range differential solution. When the two receivers are at close range, this produces absolute vector arm measurements with an accuracy of one centimetre. At longer distances, atmospheric perturbations make this procedure harder. However, the use of multiple epochs enables the solution of these uncertainties, producing not only accurate measurements, but also a characterization of the atmosphere in terms of propagation delay. Although short distance between receivers means tens of kilometres for land based applications, the impact of the atmosphere is much less significant in orbit, allowing to extend this limit to thousands of kilometres. Differential GPS processing implies co‐registration of data from two (or more) receivers. Carrier phase positioning, besides measuring and transmitting the raw measurement including carrier phases, requires also that the centre of phase of the antennas are known. If this centre is not independent of the angle of sight, which is the case for most antennas, then the receiver antenna attitude has also to be known (a few degrees of attitude error is acceptable).

Velocity and acceleration derived from GPS Roughly speaking, in order to study the density of the atmosphere in low Earth orbit one has to estimate the acceleration suffered by the body in orbit. If it is a measurement of the instantaneous velocity that is sought, then it is better to compute it from Doppler measurements. However, when the average velocity between two instants is sought, it is better to compute it from position differences. In this case, more accurate relative position estimates provide better average velocity estimates. This calls for carrier‐phase measurements. Even in a stand‐alone setup (non‐ differential), it is standard procedure to use carrier‐phase measurements to smooth the pseudo‐ range measurements. For average velocity estimations, tracking the carrier‐phase between epochs will lower the measurement noise (from the meter level to the decimetre or centimetre level). However, these measurements will still suffer from the variations of the atmospheric perturbations and from ephemeris errors. In differential mode, these errors are reduced in the proportion of the distance to the reference station (which could be another satellite), relative to the distance to the GPS satellites. This accuracy applies to the differential measurement between the two stations. In any case, there is no need for carrier‐phase ambiguity resolution (absolute position errors in the 10 meter range are acceptable for accurate velocity determination). Therefore, the value added by using double frequency receivers for velocity determination is ionospheric perturbation variations estimation and the availability of more measurements per epoch. Similarly, the most accurate average acceleration estimates are obtained from double differences of positions and not single differences of Doppler derived velocities. Considering that time intervals Δt of 10 or more seconds separate the sample times of these double differences, it can be assumed that each position measurement error (excluding atmospheric perturbations) is sufficiently independent. Therefore, the accelerating error is roughly estimated as being given by Δa = 2σ/Δt2, where σ = 1 meter (or more) for pseudo‐range based position estimates and σ = 1 centimetre for carrier phase based measurements. Therefore, using carrier phases with Δt = 10 seconds will provide an acceleration estimate in the order of 20 mgal and Δt = 5 minutes will provide an acceleration estimate in the order of 0.02 mgal.

This error estimate only considers the receiver noise. The acceleration accuracy is further degraded by the double differences of the position errors induced by ionosphere and ephemeris errors, the first one being dominant. In LEO orbits, few minutes are enough for the ionospheric errors to start to de‐correlate, affecting the acceleration estimation error. This error is reduced when the data is processed in differential mode, against another receiver. In this case, it is the accuracy of the differential acceleration that is improved. It is also reduced if double frequency receivers are used, as the ionospheric delay can be estimated from the comparison of the L1 and L2 measurements.

Accuracy needs for atmospheric studies Atmosphere density studies based on satellite acceleration measurements in the Thermosphere requires acceleration estimates of 0.1 mgal over periods of 10 to 20 minutes. This is only compatible with carrier phase based acceleration measurement, in differential mode. This requires operating the receiver continuously (to avoid losses of lock), registering raw data every 15 seconds, during periods of 30 minutes or more (e.g. half an orbit or one complete orbit). In this case, there is no need for precise absolute positioning, avoiding the burden of ambiguity resolution (which would require in practise double frequency receivers). Ionospheric delays could be better estimated by double frequency receivers, but differential processing between receivers and the use of ionospheric delay estimates from GNSS augmentation systems can also mitigate this problem. Ionosphere tomography requires dual frequency receivers, as it compares the delay between the two frequencies to estimate the ionosphere delay. It involves registering carrier phases without loss of lock (continuous operation) for periods compatible with those stated above for atmosphere density studies.

Receiver types There are several types of GPS receivers. Disregarding those employed for space high‐end operations of those for military application, attention is drawn to two significant types: geodetic type receivers and mass market receivers. Geodetic receivers are mostly concerned about precise positioning. They are usually double frequency receivers, providing trustworthy carrier phase measurements. They also use the full bandwidth of GPS spectrum (10 MHz), requiring the use of double frequency GPS antennas with compatible bandwidth. These receivers tend to favour precision, which make them more subject to losses of lock due to signal blockage. Therefore the antennas quality is a concern (including maintenance of good RHCP in both frequencies to mitigate multipath from the surface of the Earth). Mass market receivers are mostly concerned to bounding the position error to 10 meters. They are single frequency receivers, tracking the coarse acquisition signal and thus requiring a 1 MHz bandwidth only. Carrier phase measurements, if present, are many times noisy (there are exceptions). They provide fast reacquisition times, especially after short time losses of lock. Space operation imposes further constraint on the GPS receivers. First, COCOM restrictions must not be applied, short‐listing the set of available devices. Furthermore, LEO satellites have linear speeds that largely exceed those of the GPS receivers, requiring the tracking loops in the receiver to withstand and search within a higher Doppler shift range. GPS receivers for small satellites are mostly adaptations of mass market receivers, although specific designs are now available. The number of options in the market is low.

Data rate The data volume of a GPS receiver performing sporadic time, position and velocity estimates just for satellite orbit estimation is insignificant. In fact, used just for this purpose, the GPS receiver can be used with a very low duty cycle, producing an order of magnitude of 100 bytes every many minutes. When performing measurements for other more demanding applications, the case is different. Assuming full raw measurement acquisition (pseudo‐range, carrier phase and Doppler) every 15 seconds (for a period of several consecutive minutes), a receiver can produce the equivalent to 200

48 bytes per second (data converted from specific receiver protocol to a more compact format, yet without compression). This volume can be further reduced by compression, although going significantly below 50 bytes per second is not to be expected.

Power consumption A representative value for power consumption of mass market GPS receivers today is 0.2 W (value for continuous operation). Older receivers tend to consume a bit more, going up to 0.5 W. Space ready receivers tend to comply with this order of magnitude. Geodetic grade receivers usually require a bit more power, a figure in the order of 1.5 W to be considered a general value. In a case, it is not foreseen that a GPS receiver can be operated with a high duty cycle in a CubeSat. If used for tracking, synchronization and orbit determination purposes only, it can be switched on for a few seconds (a couple of minutes in case of a cold start) and be switched off for several minutes. If used for more demanding applications, it might be required to operate continuously during a certain period of time. Accordingly, the interval between periods of continuous operation will have to be large.

5.7.2 Jens Frederik Dalsgaard Nielsen

Science case GPS and GPS solutions to be used on‐board a number of the QB50 cubesats may have an operational as well as scientific purpose. RTK (Real Time Kinematic) GPS is an advanced GPS receiver which can deliver extra ordinary high accuracy – less or equal to 10 cm. In general, there are three intervals of GPS accuracy: low accuracy (>= 10m), med accuracy (>=1‐2m) and high accuracy (>=5cm). For ionosphere observations (density, drag) high accuracy may be preferred. In many cases medium accuracy is believed to fulfil most demands. Low accuracy does not have relevance for primary observations and results but may be feasible for localization and time stamping on scientific observations. Less than RTK accuracy is questionable as reference for some experiments where altitude is of highest importance. Other experiments as magnetic field observations will benefit very much even from non RTK GPS systems.

Description In 2011, there were approx 30 GPS satellites at an altitude of 20000 km in operation. Every GPS satellite carries out two orbits/24 hours providing 100% GPS coverage. Each satellite transmits GPS signals on a number of channels, each with their quality and usage. Not all channels can be decoded by everybody. Different channels may have different applications. GPS is a triangulation‐based system based on regular reception of position and time from a number of GPS satellites based on timing of received GPS “messages” from different satellites. Depending on different approaches for real time analysis of the GPS signal, very different accuracies can be obtained. Space qualified non‐ RTK GPS devices may have accuracy up to 1‐5 m range whereas RTK GPS receivers may reach cm accuracy.

Performance GPS receivers normally deliver position and time information with rates of 1 to 5 Hz. The highest precision of GPS can be achieved by Carrier‐Phase Enhancement (CPGPS) enhancement of GPS – RTK GPS – which analyses the phase of the received radio signal from the GPS satellites. The wavelength is about 0.19 m. It is claimed that accuracies about 1 % is reachable, meaning a position accuracy about 0.0019 m. One problem with CPGPS (aka RTK GPS) is the lack of knowledge of which wave you are within so position may be of multiple of 0.19 m wrong. CPGPS receivers are known to obtain

49 accuracies up to mm in extreme implementations – but more normal are in cm range. Non CPGPS space qualified receivers may operate in m range accuracy. Time obtained from a GPS will be within 10/100 ns to 1 µs accuracy – the former accuracies from CPGPS systems. In real operation 100 ns is considered as the limit. DLR Phoenix‐XNS has accuracy within 1 m. The mass, power nad volume required for the DLR Phoenix non RTK are 20 g, 0.85 W and 50 cm3.

Calibration

The only calibration to be carried out is the initialization. A cold initialization (meaning no seed of position and time) may take up to 15 minutes. A warm start with a priori position and time may take 30 seconds (from DLR Phoenix module and AlCatel Topstar 3000). During calibration and in operations in general the patch GPS antenna shall point “upwards”. A tolerance for this pointing is normally within +/‐ 30°.

Data rates

The DLR Phoenix operates with a baudrate of 9k6 to 57k4 using the NMEA plus protocol.

Operational modes

After initialization, Phoenix can deliver NMEA encoded data with up to 5 Hz. Communication is by an RS232 interface and there are a number of options depending on the specific model. Operations will consist of: 1. Switch GPS on

2. Set configuration parameters by RS232 interface

3. Set predicted position and time for a fast warm start‐up

4. Wait: warm startup <= 30 seconds, cold startup up to 15 minutes

5. Switch off

Special requirements For proper operation the GPS antennas must “point up in space” within some tolerances. For that, an effective active attitude control system (ADCS) is necessary. The commercial suppliers mentioned above do not clearly give pointing accuracies but it is expected that upward pointing must be within +/‐ 30° and this attitude must be maintained during the initial as well operational phase. Antennas will be patch antennas with a coverage area that is compatible with a 2 unit cubesat although solar cell panels may conflict with side panel location. This will depend whether all surface of the QB50 CubeSats will be available.

Heritage The Phoenix GPS has been flown on the Prisma formation flying mission and is still in operation. It has also flown on Proba‐2.

Cost Indicative price for Phoenix‐XNS including LNA and patch antenna is €12000 and €10000 without antenna system. The price is for non commercial research use by universities. The price is open for negotiation for QB50 depending on quantity and does not cover thermal and vacuum testing as well

50 as inspection by a space inspection certified company. If DLR are requested it will be carried out by an external company. The DLR supplied antenna may be to big 8cm x 8 cm x 2.5cm but DLR states that smaller antennas can be used.

Development schedule and risk The units are ready to purchase although the Interface SW has to be developed. There is no major risk on the HW part. Major risk may be in SW development but is not critical.

5.8 Laser Retroreflector Zoran Sodnik, ESTEC

In Satellite Laser Ranging (SLR) a global network of stations measures the instantaneous round trip time of flight of ultrashort pulses of light to satellites equipped with special reflectors. This provides instantaneous range measurements at millimetre level precision (standard is 25 mm) which can be accumulated to provide accurate orbits and a host of important science products. From a sequence of distance measurements, using orbital dynamics calculations, the orbital parameters, including altitude and velocity can be derived. SLR is the most accurate technique currently available to determine the geocentric position of a satellite in Earth orbit. The International Laser Ranging Service (ILRS) has been formed by the global SLR community to enhance geophysical and geodetic research activities (http://ilrs.gsfc.nasa.gov). The laser ranging services are provided by the ILRS free of charge.

NASA’s next generation SLR station, currently under development, is fully automated and unmanned. It will operate 24 hours a day and will provide subcentimetre position precision to the full suite of satellites in Earth orbit. The SRL2000 design is expected to greatly reduce the costs for station construction, operations and engineering maintenance.

Corner Cube Retroreflectors (CCR) are passive devices that are used to reflect back a laser beam originating from a ground laser ranging station for the purpose of measuring the distance from the ground station to a satellite, in the case of QB50 to a CubeSat. A set of three mutually perpendicular reflective surfaces, placed to form the corner of a cube, work as a corner cube retroreflector. CCRs are completely passive and require no power. They can be purchased from many suppliers. Edmund Optics, for example, use CCRs fabricated from Schott N‐BK7 glass, a borosilicate crown optical glass which has high homogeneity, low bubble and inclusion content. They come in different sizes. For a 320 km orbit, a CCR with a diameter of 12.7 mm is probably sufficient. Its housing has an outer diameter of 31.75 mm and a height of 18.08 mm. The mass of a CCR including its housing is less than 50 g. The CCR can be flush mounted on one side of the CubeSat. The cost is about 200 € per CCR.

With a single CCR the angle between the retroreflector face and the ground station cannot be larger than ~45°. Therefore, some CubeSats may not be visible or not be visible during the whole pass of the tracking station. With ± 25° control in roll angle and more accurate control alignment to the ram direction, two laser retroreflectors should be adequate.

From ESA’s Optical Ground Station (OGS) on Tenerife, Spain, the QB50 CubeSats can also be detected via its liquid nitrogen cooled CCD camera by tracking the CubeSats with the telescope.

5.9 Thermistors/thermocouples Z. Várhegyi, VKI

Science case The surface of the satellite reflects the molecules of the surrounding atmospheric environment. These molecules form a reflected beam, which modify the properties of the surrounding atmosphere in the vicinity of the spacecraft.

The reflected beam properties are largely influenced by the surface temperature, therefore the accurate knowledge of surface (sensor) temperature is essential. Depending on the position and orientation of sensors, the reflected beam alters the measured velocity, density, kinetic temperature and concentration of species compared to the undisturbed environment. In general, modeling of the flow field around the satellite is necessary to reconstruct the undisturbed (free stream) values from the values measured with the sensors. Flow field modeling is also a key step in accurate prediction of momentary drag coefficient of the satellites in order to retrieve atmospheric density from accurate trajectory measurements such as radar or GPS.

Some of the instruments may require temperature measurements for temperature compensation of their reading due to the temperature‐dependent working principle. Thermal/power management of the satellite also requires temperature measurements of the surface and onboard equipment.

Description The following sensors are candidates: thermistors, resistance temperature detectors, and integrated‐circuit temperature sensors. All of these sensors are small, analog, and commercial off‐ the‐shelf electronic components. The range of voltage, current and power makes them suitable for use on CubeSats. For the interpretation of drag measurements a ±1° accuracy in surface temperature measurement is needed.

Thermistors: thermistors are thermally sensitive resistors made of semiconductor materials. Resistivity of the sensor is a known function of temperature. The sensor, as a usual resistor, has two legs. Measuring the resistance of the sensor, connected to a Wheatstone bridge, and then digitizing the output voltage by A/D conversion makes it possible to calculate the temperature of the sensor.

Resistance temperature detectors (RTD): these sensors have the same working principle as thermistors but use metallic materials. Thin film platinum resistors are one example.

Integrated‐circuit temperature sensors: these sensors are transistors with three legs, designed to have a voltage output linearly dependent of temperature if the input is fixed. The output can be directly digitized by an A/D converter.

Performance Sensor Range [°C] Resolution [°C] Sensitivity Range Primary Power Thermistors ‐55 ‐ +150 ±0.05 – ±0.2 nonlinear 4300 steps 2‐40 mW RTDs ‐50 ‐ +500 ±0.15 ‐ 3850 ppm/C 3667 steps 3‐10 mW IC temp. Sensors ‐50 ‐ +150 ±0.5 ‐ ±2 5‐10 mV/C 400 steps max. 0.5 mW Table 5.9.1 Temperature sensor performance and power summary

Calibration Calibration of temperature sensors must include temperature characteristics, time response, thermal resistance of mounting, noise/power supply sensitivity and ageing. Calibration of un‐ mounted sensors is possible with high precision in an oil bath. Calibration of mounted sensors can be done the same way or in tempered air in contact with solid masses of known temperature and large thermal inertia. In‐flight calibration is only qualitatively possible with cross‐calibration and reference to a thermal model.

Mass and Volume Part mass/sensor [g] volume/sensor [mm3] Sensing element 1 5x3x3 incl. mounting Cables, soldering 5 dia. 1 x 100 x 2 or 3 cable Electronics 5 10x10x3 Mounting balance included above Total 15 g Table 5.9.2 Temperature sensor mass and volume

Data rates and options For surface temperature measurements, at least one temperature sensor per side is recommended. Altogether 12 sensors provide cross‐check on each side and good coverage. The minimal time interval of significant change in side temperature is determined by the angular velocity of the satellite attitude. As a worst case scenario, 5 Hz is taken. Although none of the above temperature sensors are expected to have a time constant smaller than 1 s, this angular motion may be recoverable by transfer functions, if needed. Worst case situation can be encountered after deployment, during/after antenna or boom opening, after collision, in case of attitude control system malfunction, and approaching/during re‐entry. For the 5 Hz worst case angular speed, a sampling frequency of 10 Hz is necessary (reaching the Shannon limit). In nominal conditions, the sampling time interval can change between 1 and 100 s, depending on the actual angular rate of the attitude motion. The resolution of the sensors can be captured by a 12 bit A/D converter, producing 1.5 byte data per measurement/channel. In order to optimize the collected amount of data, adaptation of the sampling interval to the angular rate is recommended. This can be done automatically or by ground command. Parallel with the reduction of sampling, digital averaging/filtering can be easily realized.

Table 5.9.3 summarizes the data rates of 12 sensors, with a 12 bit A/D conversion for different sampling times.

Sampling Resolvable angular rate [°/s] Data rate Data collected in 100 interval [s] [byte/s] minutes [kB] 0.1 1800 180 1080 0.2 900 90 540 0.4 450 45 270 0.8 225 22.5 135 1.6 113 11.3 67.5 3.2 56 5.63 33.8 6.4 28 2.81 16.9 12.8 14 1.41 8.4 25.6 7 0.70 4.2 51.2 4 0.35 2.1 102.4 2 0.18 1.1

Table 5.9.3 Temperature sensor data rates

Operational modes Data collection rate together with averaging or filtering should be selected per channel or on a grouped basis.

Required commanding Power voltage supply for the sensors and A/D converters should be switched on/off. Data collection rate together with averaging or filtering should be selected per channel or on a grouped basis.

Special requirements Nearby power dissipative elements should be avoided or their influence on temperature measurement should be carefully modeled. Quantitative characterization of the surrounding thermal structure in terms of thermal coefficients and surface radiative properties is necessary for post‐flight heat flux history reconstruction.

Heritage Sensor Flight heritage Thermistors yes [1],[2] RTDs yes [3] IC temp. sensors yes [4] Table 5.9.4 Temperature sensor heritage [1]: AAUSat, http://www.cubesat.auc.dk/dokumenter/ADC‐report.pdf [2]: Measurement specialties, http://www.meas‐spec.com/ [3]: SEEDS satellite, http://cubesat.aero.cst.nihon‐u.ac.jp/english/seedsdetail_e.html [4] Compass‐1, http://www.raumfahrt.fh‐aachen.de/compass‐1/publications.htm

Development schedule Sensors and associated electronics are available on the market as COTS products. Their cost is small and the application in CubeSats is widespread, adaptation is straightforward with little risk.

5.10 QBOS (Bolometric Oscillation Sensor) M. van Ruymbeke, Royal Observatory of Belgium

Science case The bolometer will measure the radiative and thermal fluxes as well as the ambient temperature with a very large dynamical range. The science goals can be summarized as:

• Provide in‐situ measurements of radiative fluxes at selected wavelengths (Visible, IR). Since the IR flux from the Earth will vary due to the changes in albedo, cloud coverage, aerosols etc, detection of radiative flux variations with a very high dynamical range will yield valuable information on the Earth’s energy balance. • Study gas‐surface interactions. One of the strongest sources of uncertainty in the determination satellite drag (i.e. atmospheric density) in rarefied environment of drag is due to the interaction of gas particles with the surface as the molecules exchange angular momentum and energy with the surface. • Flow field modeling is also a key step in accurate prediction of drag coefficient of the satellites in order to retrieve atmospheric density from accelerometers or accurate trajectory determination using radar, GPS and/or laser reflectors. The surface of the satellite reflects the molecules of the surrounding atmospheric environment. These molecules form a reflected beam, which modify the properties of the surrounding atmosphere in the vicinity of the spacecraft. The reflected beam properties are largely influenced by the surface temperature, therefore, the accurate knowledge of surface (where the sensor is) temperature is essential. Depending on the position and orientation of sensors, the reflected beam alters the measured velocity, density, kinetic temperature and concentration of species compared to the undisturbed environment. In general, modeling of the flow field around the satellite is necessary to reconstruct the undisturbed (free stream) values from the values measured with the sensors.

Some of the instruments may require temperature measurements for temperature compensation of their reading due to the temperature‐dependent working principle. Moreover, the monitoring of the thermal state will help to identify and correct unpredicted signals from other instruments. Thermal/power management of the satellite also requires temperature measurements of the surface and on‐board equipment.

Description Thermistors are used in a Wheatstone bridge arrangement to provide extremely precise temperature measurements. The thermistors are small in dimension and the required voltage, current and power are suitable for use on CubeSats. Temperature differences as small as few µK have been measured on the ground and can be expected to be measured in space.

Figure 5.10.1: Example of Thermistors bridge recording temperature with a µK precision in an underground laboratory.

Performance By configuring the thermistors as bolometers, radiative flux measurements can be made. The proposed bolometer adapted to the CubeSat could be a 8 channels system which can register simultaneously at a given sampling rate (typically 10 s) the thermometric and/or optometric signatures from the four faces of the CubeSat. It consists of a mass placed on the exterior of the satellite (at ambient temperature) and attached to the body of the satellite by a thermal shunt. The re‐emitted radiative energy of the small mass is measured along with the gradient of temperature in the shunt, which provides the conductive heat flux. The sum of both corresponds to the incident radiative flux. With 8 sensors, 6 fluxes in the 3 directions can be obtained and, using the satellite orientation in space (with respect to the Earth and the Sun), the total amplitude and direction of the incident flux coming from the Earth and from the Sun can be obtained. A complexity is added as the incidence angle modulates the incidence flux. The final design of the proof mass and the mission scenario/satellite orientation will be modulated when this is known.

We assume the electronics, the thermistors, the attachment hardware and cables to be ~20 g for 8 channels. The proof mass would be typically at the level of few grammes per channel, leading to a total mass of typically 40 g for the whole instrument. The total mass will depend on the geometry of the proof mass and the exact number and positions of the different small sensors. A simpler arrangement would be a device with a single sensor based on a two‐channel measurement (temperature and/or irradiance variations). Nearby power dissipative elements should be avoided or their influence on temperature measurement should be carefully modeled.

Part Description Performance Mass Volume Power number Range from ‐55 to +155°C thermistor sensor 0.1g 3mm³ 0 8 relative: 10µ°C absolute:0.2°C 10g cable & assembling connection 100x1mm3 0 17 +10g Range from ‐40 to data +125°C 17mW electronics acquisition 22 bits 15g 40x40x7mm3 1 (3.4V/5mA) system TBC 8 channels 0.1s to 60s from 9.6k to communication Rs422 TBC TBC 1 115kbits/s

Table 5.10.1 Technical characteristics of the Bolometer

Operations and Commanding For a nominal configuration, monitoring the temperature of the 4 faces of the CubeSat, the data from the output channels will be recorded simultaneously at a period of 10 s leading to a rate of 24 Bytes/10 s. Lower sampling rates could be considered. Pre‐flight calibration will be necessary to convert the measured thermistor values into absolute temperatures. Calibration of temperature sensors must include temperature characteristics, time response, thermal resistance of mounting, noise/power supply sensitivity and ageing. Calibration is possible with high precision by contact with solid masses of known temperature and large thermal inertia. Ageing patterns of thermistors are sufficient for ground‐based calibration application prior to observation in orbit.

Special requirements Quantitative characterization of the surrounding satellite thermal structure in terms of thermal coefficients and surface radiative properties is necessary for post‐flight heat flux history reconstruction.

Heritage The design of the bolometer is very simple and does not require any deployment or mechanical parts. The hardware parts exist and are space qualified. It must be noted that space qualification will change the masses given above as it will require specific non‐miniaturized and adapted hardware. The simplicity of the measurement method (already proven on a space mission), allows a large flexibility to modify/adopt the configuration of the sensors, on a best‐efforts basis, linked to the available resources and scientific objectives. Table 5.10.1 summarizes the mass, power, volume, data volume/rates, commands, and characteristics. The proposed approach for the bolometer is a direct heritage of the BOS on board the PICARD satellite (Figure 5.10.2). The BOS signal integrates solar, Earth and satellite irradiative energy induction, in addition to the conductive flux through the thermal shunt, with absorption and emission of radiation of the black and white front faces.

Figure 5.10.2 BOS onboard the PICARD satellite

5.11 Silicon Detector M. Messerotti and A. Gregorio, University of Trieste

Scientific relevance to QB50 objectives The Silicon Drift Chamber (SDC) described here is a silicon detector to monitor the radiation environment in order to build a map of the flux of radiation impinging on the instrument. This device is the one that has been chosen to be used on AtmoCube, the CubeSat produced by the University of Trieste. AtmoCube is one of the CubeSats that is supposed to fly on board the Vega Maiden flights in early 2012.

The Earth's atmosphere and magnetosphere allow adequate protection of the humans on the ground, but astronauts in space are subject to potentially lethal doses of radiation. The interaction of high‐energy particles with living cells, measured as radiation dose leads to chromosome damage and, potentially, cancer. Large doses can be fatal immediately; high energy solar protons are particularly hazardous. Radiation risk estimations at altitudes around 350 to 600 km appear to be one of the basic scientific problems in the planning and design of manned and un‐manned missions.

The SDC will be able to sample both particles and photons in selected energy sub‐bands over a wide range of altitudes, ranging from about 350 to 1200 km. In principle, this should allow both flare effects detection and Earth inner radiation environment mapping. Note that this will be the first time a Silicon Drift Detector will be used in the space: it could be considered a very important possibility to test its behaviour on board a space mission.

Description The SDC is a detector realized on high‐resistivity n‐doped silicon substrate, having a surface of ~2 cm2 and a thickness of 300 µm. SDC is particularly suited for low‐noise spectroscopy, given its working principle based on the “transverse depletion” concept, i.e. the possibility to deplete a large silicon wafer (e.g. n type) via a small n+ electrode (the anode) that is positively biased with respect to a set of p+ diodes (field electrodes or cathodes) placed on both surfaces of the wafer. The small dimensions of the collecting electrode ensure low capacitance and hence low noise and good energy resolution. The measurement energy range extends up to 70 keV. The expected rate of events due to charged particles is of the order of 4 Hz at the Earth’s poles. It should be noted that this is a small and simple SDC. Being the very first silicon drift chamber in space, we have decided to use the easiest design but this implies worse position and energy resolutions. A picture of the SDC and its schematic working principle is given in Figure 5.11.1.

Figure 5.11.1 A large area (56 cm2) SDC developed for the ALICE at LHC (left) and a schematic illustration of the charge transport in an SDC (right)

The SDC sensor and the read‐out electronics are both mounted onto an ad hoc designed printed circuit board, fitting the available space inside the satellite. The estimated power consumption of the complete read‐out electronics is of the order of 0.5 W. Tests are ongoing on the detector and the read‐out electronics which are under development at INFN (Istituto Nazionale di Fisica Nucleare) laboratories in Trieste.

A linearly varying potential is applied to the field electrodes. When an ionizing event creates electron‐hole pairs inside the silicon substrate, holes are rapidly swept across the substrate towards the p+ electrodes, while electrons are collected in the potential minimum created by the symmetric bias in the middle of the wafer and drift under the action of the constant drift field toward the low capacitance anode. The design of the SDC includes an integrated voltage divider (realized via lightly doped p+ implants), so that no external connections (apart from the “High Voltage” and the “Reference” connections for the first and the last drift electrode) are needed to provide the drift field.

The SDC is read‐out via a low‐noise front‐end chain based on a Charge Sensitive Preamplifier, a semi‐ Gaussian shaper with adjustable peaking time and a peak detector. The output of the peak detector is used also to feed two low offset comparators: the first one detects higher energy events (over 70 keV), whereas the second one generates a trigger signal for starting the energy measurement for particles with energies lower than 70 keV. The analog signals are converted with an 10‐bit ADC and data are stored in the FPGA which hosts the "brain" of the spectro‐dosimeter. The SDC sensor and the read‐out electronics are both mounted onto an ad hoc designed printed circuit board, fitting the available space inside the satellite.

Performance Being a single anode device, no measurement will be performed on the position of the impinging particle. The measurement energy range extends up to 70 keV, with a resolution of the order of 2.5 keV ΔEFWHM. More sophisticated (multi‐anode) SDCs and read‐out electronics systems can provide better energy and position resolution.

Calibration Silicon drift chambers can provide an internal calibration measurement but the device proposed here is not implementing this possibility. The calibration of the device depends mainly on its temperature. Once the device is characterized on ground, a thermal sensor should be used to verify and estimate the resolution.

Mass and Volume The mass of the full device is of the order of 80 g. The SDC developed for the AtmoCube can be located on two electronics boards that fit the Cubesat standards (about 10 cm×10 cm) and may be compatible with the pc104/pc104 plus standards. The height of these boards is of the order of 2 cm. A dedicated window on the satellite structure (for example a thinning of the aluminum in correspondence of the detector), should be foreseen in order to allow the particle flux measurement.

Power The potential to be applied to the voltage divider is 600 V (corresponding to a drift field of ~600 V/cm) and the current flowing in the divider is of the order of 20 µA, so that the power consumption

60 associated with the sensor bias is of the order of 12 mW. The estimated power consumption of the complete read‐out electronics is 0.5 W.

Data rates and options The expected rate of events due to charged particles at low altitude is of the order of 4 Hz at the poles. The SDC is producing 512 bytes per measurement. By considering an acquisition sampling rate of the order of 5 Hz, this corresponds to a data rate of the order of 20 kbps.

Operational modes The SDC considered here has few operational modes: on/off, measurement start and stop, data read. More sophisticated silicon drift chambers can provide a calibration measurement but this is not implemented here.

Special requirements Silicon devices show better performance at low temperature even if the SDC can still work at room temperatures. Keeping the temperature below 0° C makes the device more stable while above this temperature the accuracy of the measurement becomes dependent on the temperature itself (see Figure 5.11.2 below).

Figure 5.11.2 Energy resolution of the SDC as a function of the temperature

Heritage The INFN team in Trieste that developed this device has a well‐known and recognized experience in silicon detector development and, in particular, on Silicon Drift Chambers that are implemented also on the Alice experiment at LHC (see references at the end of this report).

Availability The SDC is not a commercial device and was developed for the AtmoCube mission. This does not exclude the possibility of using it on additional missions, indeed we consider AtmoCube as a test bench for this detector on future space missions. Full documentation will be ready in a short while as part of the package to be delivered to ESA for AtmoCube. Better performance in terms of spatial and energy resolution can be achieved by improving the type of detector and the read‐out electronics. This would require an additional implementation period that depends upon the new requirements. If the device should be produced on a larger scale (tens of nanosatellites), the situation needs to be discussed and agreed directly with INFN that could eventually enter in a collaboration.

Cost The full cost for supply of the SDCs for QB50 could range from 100 to 160 k€ depending upon the number produced.

Development schedule The first detector should be ready to be used on AtmoCube in a few months. If a similar detector is chosen, the development phase will be completed and a larger number of SDCs only need to be produced. An improvement in the detector performance can require an additional development phase that needs to be estimated. It should be noted that the INFN team has a wide experience in this field and slightly different chambers with better energy and position resolution already exist. With further development, the power consumption could be reduced to <300 mW.

Development risk The development of the AtmoCube detector is at the very end and in a few months it will be ready to be fully tested and verified. Up to now, we had no major problems in its development and tests but this will be only clear once the detector will be working in space. The only risk could be related to the availability of a large number of sensors that requires an ad‐hoc production of the silicon devices and the implementation of the read‐out systems. INFN is not an industry but a research institute that is used to working also on a large numbers of devices but the necessary manpower should be planned ahead and dedicated to this purpose.

5.12 Spherical EUV and Plasma Spectrometer (SEPS) R. Brunner, W. Konz, G. Schmidtke, IPM

General description and objectives In the thermospheric/ionospheric (T/I) altitude regime of the QB50 mission the solar extreme ultraviolet (EUV) irradiance is the primary energy input driving most of the versatile physical processes ending in the heating of the electrons, ions and neutral particles. For this reason, the objectives of SEPS are:

• temporal variability and the absorption of the EUV photons • energy and altitude distribution of the thereby generated photoelectrons

SEPS operation shall also contribute to the QB50 science objectives by measuring electron, ion and neutral particle temperatures and densities.

In the flow diagram of Figure 5.12.1 the SEPS objectives (filled rectangles) and related SEPS modes of operation are visualized schematically. These measurements will be well supported and inter‐ compared by other QB50 experiments such as Neutral Mass Spectrometer (NMS), Ion Mass Spectrometer (IMS)), FIPEX instrument, Multi‐Needle Langmuir Probe (m‐NLP) and GPS Receiver.

Figure 5.12.1 Flow graph of the SEPS objectives contributing to T/I aeronomy <320 km altitude and correlated operational modes of SEPS

Functional description of the sensor SEPS The SEPS sensor consists of three isolated spheres/electrodes, each connected to sensitive floating electrometers and voltage sources, each (Figure 5.12.2). By setting different potentials to the electrodes (metallic sphere, inner grid and outer grids), measurements of the ambient plasma parameters and of the extreme ultraviolet radiation can be performed. This way, several measurement modes are possible: Langmuir‐ (LM), Plasma Shielded Langmuir‐ (PSLM), Retarding Potential Analyzer Electron‐ (RPAEM), Retarding Potential Analyzer Ion‐ (RPAIM), Extreme Ultraviolet

Radiation‐ (EUVRM) and EUV Occultation‐Mode (EUVOM). The latter one will strongly depend on the measurement velocity and performance of the electronics for the EUVRM and on the orientation of the corresponding CubeSat. All these modes are potential candidates for the CB50 mission.

Mode Electrode 1 Electrode 2 Electrode 3 (sphere) (inner grid) (outer grid) Langmuir (LM) +8…-8 +8…-8 +8…-8 * Plasma shielded Langmuir (PSLM) +20…-70 Vpl Vpl * RPA plasma electron (RPAEM) +20 +10…-70 Vpl * RPA plasma ion (RPAIM) -20 +70 …-10 Vpl EUV (EUVRM) +70…-70* -50 +50 EUV Occultation (EUVOM) +70…-70* -50 +50 Table 5.12.1. Different operational modes of SEPS sensor. * 30 V for example for QB50

In the Langmuir Mode (LM) for example, the shape of the measured current signals are very similar to the typical shape of standard Langmuir probes. In addition, the sensor can be kept in a neutral behaviour with respect to the plasma environment by setting the potential of the electrodes 2 and 3 (inner and outer grid) to the value of the plasma potential (Plasma Shielded Langmuir Mode, PSLM). PSLM is providing the advantage over the LM that a larger energy range can be monitored without increasing the detection currents significantly and, hence without disturbing the spacecraft potential.

Figure 5.12.2‐a: The SEPS sensor (schematics) b: SEPS engineering model EUV measurements at several photon energies between 20‐160 nm and plasma measurements, as have been carried out at the BESSY synchrotron, at the IPM test chamber and at the ESTEC plasma chamber also in comparison with Langmuir probes, show characteristic features as expected.

Typical deduced plasma parameters are listed in Table 5.12.2. The EUV indices are parameters useful as inputs for models or services for space weather analysis.

Deduced parameter Derived from measurement modes η , η e i Electron density, ion density T e Plasma temperature Е , Е e i Energy distribution of electrons and ions Vsc, V pl Spacecraft potential, plasma potential

EUV EUV spectra, important range ~ 6 – 70 eV, rough resolution EUV , EUV , TEC therm Several deduced indices like EUV … solar activity EUV total electron content, EUV solar activity etc.

Table 5.12.2: Parameters which can be deduced from SEPS measurements

Adaption of SEPS to QB50 requirements and current status of development The design and manufacturing of the Development Model (DM) have been finished at TRL 5. In view of the science objectives in Section 1.1, it is proposed to install two or three SEPS sensors onboard QB50 satellites by operating SEPS in selected modes, each to derive specific aeronomic parameters. The following modes of operation are relevant: 1. RPA Electron‐Mode: Temperature, density and photoelectron distribution 2. RPA Ion‐Mode: Density and temperature 3. EUV‐Mode: Total EUV irradiance variability 4. Langmuir‐Mode: Electron temperature 5. Plasma Shielded Langmuir‐Mode: Electron temperature 6. SEPS Testing‐Mode: Testing SEPS capabilities. This set of SEPS sensors onboard the QB50 satellite network would allow to investigate the variability of T/I parameters in view of the transfer of solar energy by photoelectrons with altitude over three solar rotations. In addition, it would offer the chance to study the influence and propagation of space weather events in the T/I system at the given altitude regime from 320 km down to about 100 km, a region that is barely accessible.

In order to realise the proposed installation of two or three SEPS sensors the EM (see Table 5.12.3) has to be adapted to the QB50 interfaces. However, since it is expected that the full range of ±70 V will not be available, the corresponding limitations of the measurements will be discussed in more detail. For example, if the voltage range available will be ±30 V most of the integral photoelectron flux will be measured and the most important energy range from 0 eV to 30 eV will be resolved. The sensor could be integrated into a QB50 CubeSat as shown in Fig. 5.12.7.

Figure 5.12.7. Proposed location/configuration for sensor SEPS

For the sensor electronics, the estimated power consumption is below 2.4 W for continuous operation (without considering the efficiency of DC/DC converter). All three electrometers will need their own power supply for floating potential measurements. The medium power consumption over one orbit can be optimized by defining optimal operation conditions with a duty‐cycle concerning power and scientific minimal requirements (< 1.0 W).

Further options reducing power consumption:

- by reducing the voltage of scan generator electronics (50 V to 30 V) - stop gap solution: cancel of electrometer unit 3 (elimination of power C and Electrometer 3, reduction of data volume of about 30%)

Special thermal stability is not necessary, an operation within 0 ‐ 40 °C for example is no problem. The sensor does not need special pointing stability during measurement; on the other hand, the knowledge about the orientation in space should be available for correction calculation of the analysis.

Figure 5.12.8: SEPS electronic interface. All three electrometers will need their own power supply for floating potential measurements.

Values for the QB50 network Sensor mass < 110 g Sensor dimension 50 x 50 x 75 mm Mass of PCB electronics ~ 200 g (TBC) (without DC/DC converter and electronic box) Data rate ~ 10 kbit/s (depending on level of pre‐analysis and mode operation) Sensor orientation preferable in ram direction (angular independent) Electrometer 10 pA – 10 µA, 16 bit A/D Electrical potential range ± 30 V, ~ 5 mV resolution Power consumption ~ 2.4 W max. (continuous operation), < 1.0 W (in normal operation with duty‐cycle) Table 5.12.4: SEPS specifications

For the QB50 application, the sensor has to be adapted in its size and its electronics has to be redesigned with commercial electronic parts. The estimated development time frame for these activities will be 12‐15 months.

5.13 WINCS A. Nicholas, NRL

The Winds‐Ion‐Neutral Composition Suite (WINCS) instrument was designed and developed jointly by the Naval Research Laboratory (NRL) and NASA/Goddard Space Flight Center (GSFC) for ionosphere‐thermosphere investigations in orbit between 120 and 550 km altitude. The three WINCS instruments are: the Wind and Temperature Spectrometer WTS, the Ion‐Drift and ion temperature Spectrometer IDTS, and the Neutral and Ion Mass Spectrometer NMS/IMS. The WINCS design provides the following measurements in a single package with a low size, weight, and power: 7.6 x 7.6 x 7.1 cm outer dimensions, 0.8 kg total mass and < 2 W total power: neutral winds, neutral temperature, neutral density, neutral composition, ion drifts, ion temperature, ion density and ion composition.

Science Goals 1. Characterize neutral atmosphere composition, winds and temperatures at spacecraft altitude for analysis of global mass coupling, energy deposition, and momentum transfer. 2. Characterize plasma density, temperature, and drifts for studies of ion‐neutral coupling, electric field morphology, and regional heating phenomena. 3. Address optimal ingestion of thermospheric space weather data into climatological and general circulation models.

(a) (b) (c)

(f) (e)

Figure 5.13.1. Panel (a) layout of the three spectrometers on WINCS on 3 square openings shown with overall cube dimensions 7.6 x 7.6 x 7.1 cm. WTS and IDTS are paired off in two spectrometer modules WTS1/IDTS1 and WTS2/IDTS2 with mutually perpendicular fields of view as shown by the two pairs of long slits. The GEMS NMS/IMS pair aperture is at the lower left. An exploded view of the instrument is shown in panel (b). Panels (c) through (e) show the focal plane array, WINCS electronics stack and the stack mounted in the electronics enclosure base

Sensor Concept The WINCS suite of spectrometers is based on the small deflection energy analyzer (SDEA), which allows for a series of measurements to be made over an angular space for a set of discrete energy ranges. This combination of energy and angular measurements allows one to

measure the density, winds (drifts) and temperature of local neutrals (and ions). Figure 5.13.2 depicts the sensor viewing geometry. Table 5.13.1 lists the WINCS observables, while the bus and data requirements are listed in Tables 5.13.2 and 5.13.3 respectively.

Figure 5.13.2. In a satellite moving to the right with velocity VS, ambient atoms and molecules pass the entrance aperture to form a pattern at the detector plane, as shown by the flux contour rings. The center of the flux pattern corresponds to the peak of the angular distribution of the particle flux, determined by the wind velocity W. Measuring the energy distribution at a known angle on detector gives the magnitude of VT and knowledge of angle yields W.

Table 5.13.1. WINCS Measurement Capabilities Instrument Parameter Range Resolution WTS Density 103‐1010 cm‐3 <3% WTS Temperature 100‐5000 K<1% WTS Wind +/‐ 2000 m/s 16 m/s

Neutrals Table 5.13.2. Bus Requirements 3 10 ‐3 GEMS NMS Composition 10 ‐10 cm <3% Mass 800 g IDTS Density 103‐107 cm‐3 <3% Power 2W (6‐8 V) IDTS Temperature 100‐5000 K<1% Volume 7.6 x 7.6 x 7.1 cm Telemetry (Raw) 23 kbps Ions IDTS Drift +/‐ 2000 m/s 16 m/s Pointing 2° of RAM 3 7 ‐3 GEMS IMS Composition 10 ‐10 cm <3% Knowledge <.0.1°

Table 5.13.3. WINCS Suite Raw Data # Bits/ Steps/ Bits per Sensor Anodes Channel Channel Sample WTS Horizontal 16 16 20 5120 WTS Vertical 16 16 20 5120 IDTS Horizontal 16 16 20 5120 IDTS Vertical 16 16 20 5120 Mass Spec. 1 10 200 2000 Housekeeping N/A N/A N/A 240 Total 22720

5.14 Sensor Resource Summary: Table 5.14.1

Sensor Science area Mass Power Volume (cm3) Accommodation Bytes/s Unit Dev. TRL Comments (g) (mW) Cost cost ( k€) (k€) Neutral Mass Low mass neutrals 450 500 645 16 60 4 Spectrometer Ion Mass Low mass ions 450 500 645 16 230 4 Spectrometer Characterise ions WINCS 800 2000 410 7.6 x 7.6 x 7.1 2875 500 6 and neutrals 3.6 x 3.0 x 1.2 Atomic and Req. 12V power or FIPEX 70 3000 + 4.4 4.0 120 6‐9 2 sensor pack molecular oxygen 15g additional mass 8 x 10 x 1 Absolute electron 400- 4 booms of length 10- Raw: 231 Langmuir Probe 120 66.1 3.0 3.25 6 4 probes density 1000 17 cm Proc: 1.6 5 x 5 x 7.5 EUV photons SEPS 310 1000 + See fig 5.12.7 1250 4.0 3 photoelectrons 10 x 10 x 5 12 with 10 Magnetometer Magnetic field 132 240-500 183 30cm Boom 11.1 50 5 vectors/s Accelerometer Drag 50 350 5 x 10 x 3 1.2 10 150 3? MEMS based Drag; electron Attitude control 1o?; Probably not sensitive GPS 20 850 50 22.5 (DRL) 9 column density antenna enough Thermal disturbance: 180 for 0.1 Thermal sensors 180 4.1 6 - 480 2 per surface 0.5 9 12 sensors Calibration of other sec sampling sensors Map radiation Aluminium window Silicon Detector environment up to 80 500 400 2500 4 Pref. below 0°C 70 keV Radiative flux Sensors on the Q-BOS Gas-surface 40 17 4 x 4 x 0.7 2.4 low 9 8 sensors exterior interaction Laser retroreflector Position 30 0 17.3 Two facing Earth 0 0.2 9 2 retroreflectors

6. QB50 CubeSat physical constraints and payload architecture The basic 1U CubeSat is considered to be:

• 10 x 10 x 10 cm • 1 kg • 1 W orbit average power

Given the working hypothesis that the payload should occupy no more than a single unit, then these requirements are taken as general constraints to payload selection. However, the orbit power average can accommodate higher power levels for shorter times. Instruments with high power demands will not operate throughout the orbit. It is anticipated that in order to perform the coordinated scientific measurement necessary, operation during night‐time will be required. The high level of atmospheric drag and the need to align some sensors with the velocity vector means that very rapid control of the CubeSat attitudes will be necessary once released. The alignment shall be that the CubeSat long axis points in the direction of motion with the sensor package leading.

Coordinated measurements will require time‐tagged commands executed onboard and absolute knowledge of time to ~1‐10 ms (tbc). It is expected that the use of a low‐performance GPS will be the chosen approach to deal with onboard time.

7. Selection For the 2U atmospheric CubeSats, a conceptual division into a 1U platform and 1U sensor payload was envisaged as the maximum that could be expected from the CubeSat providers. For the atmospheric CubeSats it was therefore necessary that the 1U constraints be met for the sensor package. Note, however, that power constraints could be met through operational constraints.

Other selection criteria included:

• Alignment with mission science theme • Appropriate sensitivity • Technical maturity • Production feasibility • Relative simplicity of operation • Relative simplicity of accommodation • Payload complementarity • Cost

Selection Categories

Sensors or Sensor Packages were categorised as follows:

• Not selected o Not appropriate o Not feasible within spacecraft resources available • Proposed as a Special o Too expensive for multiple deployment o Not requiring a network deployment o Large element of technology demonstration • Selected • Still under consideration o Not essential but with significant value scientifically, cost not justified within QB50 budget however some accommodation might be found if alternative funding is available

Special sensors/sensor packages The resource requirements of the following sensors/sensor packages did not allow accommodation in a standard 2U atmospheric CubeSat in combination with other sensors, or the technology was not ready and the required development would not be funded in time, or the sensor was considered of lower scientific priority. However, even if they could only be flown on one or two CubeSats they would nevertheless contribute significantly to the exploration of the lower thermosphere/ ionosphere and also provide technology demonstration opportunities. The PIs of these sensors/sensor packages are encouraged to submit proposals for the category of 2U/3U In‐Orbit‐ Demonstration (IOD) CubeSats.

• Magnetometer • SEPS • WINCS

• high‐precision GPS receivers • Silicon Drift Chamber

Recommended for flight on QB50

The following sensors were selected for the standard 2U atmospheric CubeSats.

• Neutral Mass Spectrometer • Ion Mass Spectrometer • FIPEX sensors for atomic and molecular oxygen • Multi‐Needle Langmuir Probes • Laser Retroreflectors • Thermal sensors

Unfortunately, various constraints prevent the accommodation of all these sensors on all 2U atmospheric CubeSats (for details see Section 9). Where possible, the selected sensors should also be included in appropriate 2U/3U IOD CubeSats.

8. Recommendations 1. Overall it is not feasible to fly 50 2U CubeSats and 6 Specials due to mission constraints. It is therefore recommended that: 1.1 The total number of satellites shall be 50 1.2 The number of 2U/3U IOD CubeSats should not be more than 10. Therefore, the minimum number of 2U atmospheric CubeSats is 40. 1.3 Where possible, the 2U/3U IOD CubeSats should include a sensor package 2. Simplicity should be the theme of the mission and should cover: 2.1 Payload operations 2.2 Sensor package interfaces 2.3 Lines of communication and management structures 3. All QB50 CubeSats shall be subject to critical review and selected on the basis of an evidenced proposal. 4. The expectations of the 2U atmospheric CubeSat providers should be explicit and unambiguous. Selection should require compliance to such expectations.

These expectations include:

4.1 Provision of mass budget, power, specific electrical interface, specific data and command interface, minimum 9.6 kbits/s UHF download. 4.2 Attitude control (TBC velocity vector 5 deg, 1 deg knowledge, 25 deg roll), GPS (position knowledge, onboard time), sensor accommodation with unrestricted field of view as appropriate. 4.3 Integration of sensor elements into 2U CubeSat 4.4 Conformance with test, materials, handling, … 4.5 Very early in‐orbit commissioning 4.6 Delivery of sensor package data products in a defined format 4.7 Progress reporting and delivery on time 5. Support to CubeSat providers should be given through workshops and general communications. 6. Avoid ITAR restricted technologies where possible. 7. Separation of Functional Units and Sensor Units, while desirable from a logistics sense, is probably impractical and disadvantages CubeSat providers. 8. While an S‐band transmitter is desirable it was not thought practical to demand this of all CubeSat providers and so a UHF/VHF communications package is foreseen with a maximum downlink of 9.6 kbits/s during contact.

9. Baseline sensor package configurations Two complementary configurations, equally divided across the 2U atmospheric CubeSats, are recommended:

Set 1 : 20 x [NMS, FIPEX, Laser Reflectors, Thermal]

• Mass: 760 gms • Power: 2.7 W (reduced to 1 W through operational constraints) • Volume: 770 cm3

Set 2 : 20 x [IMS, Langmuir Probe, Laser, Reflectors Thermal]

• Mass: 780 gms • Power: 1.5 W (reduced to 1 W through operational constraints) • Volume: 717 cm3

Additional numbers of both sets will be required for 2U/3U IOD CubeSats and/or where the number of 2U atmospheric CubeSats exceeds 40. It is anticipated that this will require an additional 3 of each set. The above numbers contain no margins which might be expected to be in the order of 30 ‐ 40%. The actual requirement to be placed upon CubeSat providers will depend upon an accommodation study.

Please note that the sensor sets proposed above are based on the recommendations of the sensor selection working group. This is, however, preceded by the corresponding details provided in the Call for Proposals which represents the current position.

10. References Deckert, D., Hahn, R., Klier, A., Weber, A. Technical Summary CubeSat SOMP, interner Bericht, ILR, Technische Universität Dresden, ILR‐RSN P 10‐12, 2011.

Dobbin, A. L.; Aylward, A. D.; Harris, M. J., Three‐dimensional GCM modeling of nitric oxide in the lower thermosphere, Journal of Geophysical Research, Volume 111, Issue A7, CiteID A07314, 2006 Matsushita, S., Sq and L Current Systems in the Ionosphere, Geophys. J. R. Astr. Soc, 15, 109‐125, 1968

DoD, Secondary Payload Planners Guide For Use On The EELV Secondary Payload Adapter, 2001

Fasoulas, S.: Experimentelle und theoretische Charakterisierung einer hochenthalpen Stickstoffströmung zur Wiedereintrittssimulation, Dissertation, IRS, Univ. Stuttgart, Feb. 1995

Franke, S., Pilinski, M., Diaz‐Aguado, M., Forbes, S., and Hunyadi, G., “The University Nanosat Program from Concept to Flight: A Dual Student Program Perspective on What Works and What Does Not," Proceedings of the 20th Annual AIAA/USU Conference on Small Satellites, Logan, UT, August 2006

Hammer, F.,, Schmiel, T., Fasoulas, S.,, Messerschmid:, E. From Space to Earth ‐ a Novel Solid Electrolyte Oxygen Sensor System for In‐Situ Measurement and Process Control, Advances in applied plasma science, Vol.7, 2009

Pilinski, M. D., “An Innovative Method for Measuring Drag on Small Satellites", Proceedings of the 23rd Annual AIAA/USU Conference on Small Satellites, Logan, UT, August 2009

Schmiel, T. Entwicklung, Weltraumqualifikation und erste Ergebnisse eines Sensorinstruments zur Messung von atomaren Sauerstoff im niedrigen Erdorbit, Dissertation, TU Dresden, Sierke‐Verlag, 2009

Wilde, M., Mission Operations and Simulation for the DANDE spacecraft, Master's thesis, Institute of Astronautics, Technische Universitat at Munchen, 2008.

Xu, J., H.‐L. Liu, W. Yuan, A. K. Smith, R. G. Roble, C. J. Mertens, J. M. Russell III, and M. G. Mlynczak (2007), Mesopause structure from Thermosphere, Ionosphere, Mesosphere, Energetics, and Dynamics (TIMED)/Sounding of the Atmosphere Using Broadband Emission Radiometry (SABER) observations, J. Geophys. Res., 112, D09102, doi:10.1029/2006JD007711.

11. Acronyms

AAUSAT AAlborg University SATellite (a series of 1U CubeSats)

ADC Analog to Digital Converter AGW Atmospheric Gravity Wave AMR Anisotropic MagnetoResistance APL Applied Physics Laboratory BESSY Berliner Elektronenspeicherring‐Gesellschaft für Synchrotonstrahlung BIRA Belgian Institute for Space Aeronomy (in Brussels) BOS Bolometric Oscillation Sensor BUSOC Belgian User Support and Operations Centre CCD Charge‐Coupled Device CCR Corner Cube Retroreflector CHAMP CHAllenging Minisatellite Payload CHaPS Charged Particle Spectrometer COCOM COordinating COMmittee for multilateral export controls COG Centre of Gravity CONE COmbined measurements of Neutrals and Electrons COTS Commercial‐Off‐The‐Shelf (components) CPU Central Processing Unit DANDE Drag and Atmospheric Neutral Density Explorer DLR Deutsches Zentrum für Luft‐ und Raumfahrt (German Space Agency) DM Development Model DNEL Disconnect Non Essential Loads

DPAC Data Processing and Archiving Centre (in Brussels) ECOMA Existence and Charge state of meteOric dust grains in the Middle Atmosphere EELV Evolved Expendable Launch Vehicle EGSE Experiment Ground Support Equipment EM Engineering Model EMC Electro‐Mechanical Cleanliness EUV Extreme UltraViolet EUVOM Extreme UltraViolet Occultation Mode

EUVRM Extreme UltraViolet Radiation Mode EuTEF European Technology Exposure Facility FIPEX Flux‐Ф‐Probe EXperiment FPGA Field Programmable Gated Array GEMS Gated Electrostatic Mass Spectrometer GENSO Global Educational Network for Satellite Operations GNSS Global Navigation Satellite System GPS Global Positioning System GSFC Goddard Space Flight Center IAP Institute for Atmospheric Physics (in Northern Germany) IC Integrated Circuit IDTS Ion Drift and ion Temperature Sensor ILRS International Laser Ranging Service IMS Ion Mass Spectrometer INFN Istituto Nazionale di Fisica Nucleare INMS Ion and Neutral Mass Spectrometer I/O Input Output IOD In‐Orbit Demonstration (CubeSats) IPM Institut für Physikalische Messtechnik IR InfraRed ISS International Space Station ITAR International Traffic in Arms Regulations LEO Low‐Earth Orbit LFSC Low‐Flying SpaceCraft LHC Large Hadron Collider LM Langmuir Probe MASS Mesospheric Aerosol Sampling Spectrometer MCC Mission Control Centre MEMS MicroElectroMechanical Systems MEO Medium‐Earth Orbit MLT Mesosphere and Lower Thermosphere MLTI Mesosphere, Lower Thermosphere and Ionosphere

78 m‐NLP multi‐Needle Langmuir Probe MSIS Mass Spectrometer Incoherent Scatter (model of the atmosphere) MSSL Mullard Space Science Laboratory (in the UK) NMS Neutral Mass Spectrometer NOAA National NPU Northwestern Polytechnic University NRL Naval Research Laboratory NSF National Science Foundation (in the US) OGS Optical Ground Station (in Tenerife) PHLUX Pyrometric Heat fLUx eXperiment PVT Position, Velocity and Time PSLM Plasma Shielded Langmuir Mode RPA Retarding Potential Analyser RPAEM Retarding Potential Analyser Electron Mode RPAIM Retarding Potential Analyser Ion Mode RTD Resistance Temperature Detector SDC Silicon Drift Chamber SDEA Small Deflection Energy Analyser SEEDS Space Engineering EDucation Satellite (a 1U CubeSat built by Nihon University, launched on 28 April 2008) SEPS Spherical EUV and Plasma Spectrometer SLR Satellite Laser Ranging SNC Sierra Nevada Corporation SOMP Student Oxygen Measurement Project SSC Surrey Space Centre SSWG Sensor Selection Working Group SWAP Sun Watcher using Active Pixel system detector and image processing TBC To Be Confirmed T/I Thermosphere/Ionosphere TIMED Thermosphere Ionosphere Mesosphere Energetics and Dynamics TRL Technology Readiness Level UAF University of Alaska Fairbanks UC University of Colorado (in Boulder)

UHF Ultra High Frequency UM University of Michigan UNP University Nanosat Program UV UltraViolet VHF Very High Frequency VKI von Karman Institute for Fluid Dynamics WADIS WAve propagation and DISsipation in the middle atmosphere WINCS Wind Ion Neutral Composition Suite WTS Wind and Temperature Spectrometer ZARM Zentrum für Angewandte Raumfahrttechnologie und Mikrogravitation (University of Bremen)

12. Contributors

Dr. Anasuya Aruliah, Atmospheric Physics Laboratory, Department of Physics & Astronomy, University College London, London WC1E 6BT, UK, [email protected].

Tom Beuselinck, Redshift Design & Engineering, Europark Noord 21A, 9100 Sint Niklaas, BELGIUM; [email protected]

Dr. Raimund Brunner, Fraunhofer Institute for Physical Measurement Techniques, 791110 Freiburg, Germany; raimund [email protected]

Prof. Tore André Bekkeng, Dept. of Physics, University of Oslo, P.O.Box 1048, Blindern, 0316 Oslo, Norway; [email protected]

Dr. Patrick Brown, Imperial College London, London, UK; [email protected]

Prof. Sérgio Reis Cunha – University of Porto, Faculty of Engineering, Rua Dr. Roberto Frias 4200‐465 PORTO, Portugal; [email protected]

Prof. Jens Frederik Dalsgaard Nielsen, Automation & Control Department of Electronic Systems, Aalborg University, Fredrik Bajersvej 7C, 9220 Aalborg East, Denmark; [email protected]

Prof. Stefanos Fasoulas, FIPEXonISS Principle Investigator, Professor for Space Transportation Technology, Universitat Stuttgart, Institute of Space Systems, 70569 Stuttgart, Germany; [email protected]‐stuttgart.de

Prof. Anna Gregorio, University of Trieste ‐ Physics Department; INFN ‐ Istituto Nazionale di Fisica Nucleare, sez. Trieste; INAF ‐ Istituto Nazionale di AstroFisica, OAT – Trieste, via A. Valerio, 2 ‐ Trieste, I – 34127, Italy; [email protected]

Dhiren Kataria, Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey, RH5 6NT, UK; [email protected]

Kyle Kemble, Aerospace Engineering Sciences, University of Colorado at Boulder, Boulder, CO 80309‐ 0429, USA

Dr. Werner Konz, Fraunhofer Institute for Physical Measurement Techniques, 791110 Freiburg, Germany; [email protected]

Prof. Vaios Lappas, Surrey Space Centre, University of Surrey, Guildford, Surrey GU2 7XH UK; [email protected]

Prof. Franz‐Josef Lübken, Institute for Atmospheric Physics, Kühlungsborn, Germany; luebken@iap‐ kborn.de

Prof. Mauro Messerotti, Astronomical Observatory, Via G. B. Tiepolo, 11 34131 Trieste, Italy; [email protected]

Prof. Jøran Moen, Dept. of Physics, University of Oslo, P.O.Box 1048, Blindern, 0316 Oslo, Norway; [email protected]

Dr. Andrew Nicholas, Naval Research Laboratory, USA, +01‐202‐767‐2441; [email protected]

Prof. Scott Palo, Aerospace Engineering Sciences, University of Colorado at Boulder, Boulder, CO 80309‐0429, USA; [email protected]

Marcin Pilinski, Aerospace Engineering Sciences, University of Colorado at Boulder, Boulder, CO 80309‐0429, USA

Dr. Ruedeger Reinhard, Von Karman Institute for Fluid Dynamics (VKI), Chaussée de Waterloo 72, 1640 Rhode‐Saint‐Genèse (Brussels), Belgium ; [email protected]

Prof. Aaron J. Ridley, Dept. of Atmospheric, Oceanic and Space Science, 1416 Space Research Building, University of Michigan, Ann Arbor, MI 48109 – 2143, USA; [email protected]

Dr. Michel van Ruybeke, Royal University of Belgium ; [email protected]

Prof. Theodoros Sarris, Democritus University of Thrace, Greece; [email protected]

Dr. Gerhard Schmidtke, Fraunhofer Institute for Physical Measurement Techniques, 791110 Freiburg, Germany; [email protected]

Dr. Tino Schmiel, Faculty of Mechanical Engineering, Institute for Aerospace Engineering, Space Systems and Utilization, Technische Universitat Dresden, Germany; tino.schmiel@tu‐dresden.de

Prof. Alan Smith, Director, Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey, RH5 6NT, UK; [email protected]

Zoran Sodnik, ESA Optics Section (TEC‐MMO), ESTEC, P.O.Box 299, Noordwijk, The Netherlands; [email protected]

Zsolt Várhegyi, Aeronautics & Aerospace Dept, von Karman Institute, Brussels, Belgium; [email protected]

Dr. Andreas Weber, Faculty of Mechanical Engineering, Institute for Aerospace Engineering, Space Systems and Utilization, Technische Universitat Dresden, Germany; andreas.weber@tu‐dresden.de