TIMECOP-AE Towards Innovative MEthods for COmbustion Prediction in Aero-Engines Author : Dr. Thomas LEDERLIN Combustion Engineer Turbomeca S.A. Methods Department BP17 64511 Bordes Cedex – FRANCE Email : [email protected] Tel : (33) 5 59 12 50 65 Introduction The aim of the FP6 TIMECOP-AE project (2006-2010) was to improve the necessary combustion prediction methods that enable the development of practical advanced combustion systems for future engines, with reduced emission levels and fuel consumption. Predictive tools are required to be able to reduce NOx emissions, to decrease the development time and costs of new combustion systems and to improve the operability of lean-burn combustion systems. Most promising approaches to satisfy future emission levels regulations are based on lean combustion technology. However, lean combustion compromises combustor operability, including ignition, altitude re-light, pull-away, weak extinction performance and thermo-acoustic instability behaviour. Therefore it is of prime importance to evaluate the behaviour of the flame during these transient phases in the design stage and modelling tools are required. Without these tools the development of advanced combustion systems relies on many costly and time consuming rig tests. The high-fidelity simulations proposed in TIMECOP-AE are therefore a way to increase our competitiveness. During the last years big advances have been made in the field of reactive Large Eddy Simulation (LES) with gaseous fuels. This approach gives promising results with respect to turbulence modelling and can be used to model unsteady processes. Within the framework of TIMECOP-AE, the LES tools have gained a new critical capability: modelling of the liquid fuel combustion process for conventional and low-emission combustors, over a wide range of operating conditions. The operating conditions include the above-mentioned transient phenomena, such as ignition or extinction. The developments achieved in the simulation tools are concerned with models for turbulence, chemistry, turbulence-chemistry interactions, and liquid spray models. The methods and models developed within TIMECOP have been evaluated against high quality validation data issued from several validation test-rigs, from academic burners designed to validate a specific model up to a generic combustor, representative of an aero-engine combustor (Figure.1 ). Figure 1: Generic combustor experimental model To reach the main objective of advancing LES methods into two-phase flows for gas turbine applications, TIMECOP was divided into 4 work packages and the technical activity distributed as follows: − Development of models and study of fundamental issues, handled by work package 1 − Production of experimental data for validation, handled by work package 2 − Implementation of 2-phase capability in numerical solvers and validation calculations, handled by work package 3 − Application of new simulation tools to industrial configurations, handled by work package 4 TIMECOP at a glance − 23 institutions from 8 European countries constituted the TIMECOP consortium: o Universities: University of Cambridge (UK), Technische Universität Darmstadt (GER), Karlsruher Institut für Technologie (GER), Technische Universiteit Eindhoven (NED), Imperial College (UK), Loughborough University (UK), Czestochowa University of Technology (POL), University of Rome (IT), Ecole Centrale Paris (FR) o Research centres: CERFACS (FR), ONERA (FR), DLR (GER), IMFT (FR), CNRS (FR), ICEHT (GR), IFPEN (FR) o Industries: Turbomeca (FR), Rolls-Royce Deutschland (GER), Rolls-Royce plc (UK), MTU (GER), SNECMA (FR), AVIO (IT) − The scientific production of the project is summarized here: o 7 test-rigs o 18 CFD codes or modules o 94 technical deliverables validated o 41 publications produced Scientific content WP1 - Fundamentals Within this work package, numerical models for two-phase flow, chemistry and ignition have been developed, improved evaluated and tested. Both Eulerian and Lagrangian two-phase models have been considered, and the performances of the two approaches have been compared. Chemistry models have been developed for application to LES. Approaches are based on the Flamelet Generated Manifold method, the Conditional Closure Model, the Field PDF method, and the Computational Singular Perturbation method. Furthermore, a specific spark ignition model has been developed. The models have been implemented in numerical solvers and exploited by industrial partners. WP2 - Validation experiments There are a number of factors that still prevent full utilisation of such sophisticated simulation methods, especially for spray flames, by industry. The need to improve the reactive LES capability is closely dependent on availability of accurate, comprehensive diagnostic measurement data to be used for validation. WP2 focused on development and application of advanced diagnostic techniques on geometries and flow problems ranging from very well defined, easy-to-characterise, academic test cases to industrial test cases. The former tests were used to support model development, the latter to validate models in presence of complex geometries and ambiguity in boundary conditions. A wide range of advanced visualization and diagnostic techniques (e.g. PIV, PTV, PDA, LDA, IMI, PLIF, OH* chemiluminescence, Mie scattering) have been used and often tested to the limit of their capabilities (Figure 2). Figure 2: Experimental visualization of an ignition kernel Attention has been paid to analyse a range of operating conditions, going from altitude relight up to cruise. Both reactive and inert experiments have been carried out. A good combination of single- and two-phase flow experiments has been conducted. The data collected has been then used to both define boundary conditions and validate LES predictions. Important aspects of evaporation, turbulence-chemistry interaction, droplet transport and droplet combustion have been investigated. Notwithstanding the obvious challenges posed by application of these advanced techniques, which often have gone through their own development process within TIMECOP-AE, the objectives of producing a comprehensive matrix of test cases for validation of LES methods has been achieved as per plan. In summary, the significant achievement of WP2 has been to stretch the capability of existing diagnostic techniques and so provide valuable LES validation data used elsewhere in TIMECOP-AE. In particular, demonstration has been provided that advanced laser diagnostic methods can be directly used on industrial geometries and can produce a wealth of information on the aerothermal behaviour of aero-engine combustors. WP3 - Numerical validation and implementation of fundamentals The aim of this work package was to integrate the fundamental models into the advanced CFD methods, in order to obtain the 2-phase reactive CFD capability and resolve the intrinsic unsteady behaviour of turbulent flows. To ensure the proper implementation of these new models, validations were first performed on academic experiments. Once validated, the advanced CFD methods were ready to be tested on complex 3D geometry experiments. All the tasks defined in the Description of Work of the TIMECOP project have produced state-of-the-art scientific results and contributions in multiple scientific publications in international conferences and journals. Such a prolific production is the result of the developments and validations of multiple LES strategies to handle two-phase reacting flows not only from a theoretical point of view but also for industry-like configurations. Despite minor adjustments all the tasks with strong links with WP4 (Exploitation) have provided new industrial tools with great potential as illustrated by the industrial partners. WP4 - Exploitation LES of reactive two-phase flow is the next evolution in CFD methodologies applied to the conception of aeronautical engines. It should complement and eventually replace existing RANS conception techniques. The justification of this evolution resides in the fact that engine performances and transient phases are not predictable with the only use of RANS. Some examples are flight relight, combustion instabilities, ignition and lean blow-off prediction, etc. Also RANS is very depending on turbulence modelling and there is no general turbulence model adapted to any geometry or engine operation circumstance. The basis principle of LES makes this technique naturally adapted to complex turbulent flows, such as those encountered in aeronautical combustors (Figure 3). Figure 3: LES velocity field in a model aero-engine combustor TIMECOP-AE has greatly helped introduce the LES tools into the industrial environment for aero-engine design. All industrial partners have computed one of the configurations that were experimentally tested in WP2. The results proposed have clearly shown that performing LES of two-phase reactive flows in an industrial context is feasible. Previous to TIMECOP-AE project, LES was mainly used by researchers without or with a weak link with the aeronautical industry. Several research projects (PRECCINSTA, ICLEAC, MOLECULES) have helped industrials evaluate and understand the interest to adopt such LES tools to improve the conception process. However, only gaseous-fuel combustion simulations had been applied to industrial configurations. TIMECOP-AE has brought 2-phase flow modelling, which is an essential feature in combustion simulation, into the industrial applications. LES technique still remains