The mitigation of aviation emissions, in terms of their environmental impact, is a priority for both air quality (local impact) and the greenhouse effect (global impact). The main margin for progress is in the field of combustor technology. Lean combustion technology is the breakthrough that should enable high-level reductions in NOx emissions, both at airports and in cruise. In addition, lean combustion also assists the reduction of particulates. Injection systems are the most critical issue in achieving a satisfactory level of lean combustion, and are the focus of this project.
The project aimed to achieve sufficient maturity in lean combustion for the single annular combustor application. The objectives were a 80% reduction in NOx emissions in relation to the CAEP2 regulation limit during the LTO (Landing and Take-Off) cycle, and low NOx emission indices at cruise speed (EINOx=5g/kg as a target).
From the technological point of view, two essential features of TLC were that:
- the five manufacturers involved would be all exploring lean combustion applied to single annular combustor architecture;
- an ambitious reduction in NOx levels during LTO cycle emphasising on NOx levels at cruise speeds, particulates reduction during LTO and at cruise speeds.
In relation to the general scope, the detailed objectives were as follows:
- To intensively explore various fuel staged LP or LPP injection systems covering various engine compression rates (OPR ∈ [15, 35]) and all operating regimes. Emission performance characteristics were measured. Risk of auto-ignition, flash-back and stability was also assessed.
- To develop appropriate non-intrusive measurement techniques to enhance knowledge and understanding of all the phenomena occurring in the combustor and the injection system. To acquire high diagnosis capability for:
- NOx measurements: gas analysis;
- Particulates: SMPS on samples and LII;
- Combustion process (OH, kerosene, T, velocities, turbulence): LIF, CARS, PIV, LDA, PDPA. These techniques are based on laser use and are non-intrusive measurements.
- To calibrate CFD tools (RANS or LES codes, combustion and emission formation models) for the purposes of combustion and emission prediction. To acquire or assess prediction capability. The choice was made not to carry out numerical developments. Development activities are largely covered in recently completed or still on-going projects.
- To develop systematic optimisation procedures: definition of a lean injection system concept and associated design parameters, definition of design criteria and constraints, definition of design objectives, application of advanced optimisation algorithm.
- To extrapolate real engine performance for the future. To assess what is technologically feasible in the long term.
The project consisted of four Work Packages:
- WP1 treated the validation and the adaptation of non-intrusive measurement techniques to the conditions foreseen for the tests, in particular for HP tests (up to 30 bar).
- Highly instrumented experimental campaigns were carried out under realistic conditions (pressure up to 22 bar and high temperature) within WP 2. Advanced experimental diagnosis developed in WP1 were applied and complemented by more conventional measurements like gas analysis. The whole set of measurements permit to have a complete evaluation and understanding of the system performances: Spray characterisation, pre-mixing, pre-vaporisation, combustion process and efficiency, NOx emissions, particulates, radiation, lean extinction limit.
- An important part of the tasks within WP 3 was to demonstrate the applicability and effectiveness of optimisation methods applied for lean injection systems. Based on the results of the injector combustor optimisation design criteria was derived for lean injection systems using either combined experimental/ numerical investigation or empirical/ systematic approaches. Five different fuel injector/ combustor applications have been investigated by means of CFD and partially by the use of optimisation techniques.
- The scope of the WP 4 was the calibration of CFD tools (RANS or LES codes, combustion and emission formation models) for the purposes of combustion and emission prediction and to acquire or assess prediction capabilities.
From the results and discussions, the following general conclusions can be made:
- The calculation mesh must be chosen very carefully because different meshes can give drastically different results. Unstructured meshes composed by tetrahedral cells give problems at wall and difficulties in the convergence. Hybrid meshes, composed by both tetrahedral and hexahedral (near the walls) work better.
- Non-reacting gaseous turbulent flows are correctly predicted by most models, either RANS (for the average flow) or LES. RANS results are sensitive to the turbulence model and extensive parametric studies have been made. LES results compare very well with experimental data on both mean and fluctuating flows.
- Comparisons for the liquid phase show that RANS results give correct trends and flow structure. However only LES results agree well with the measurements on velocity profiles (mean and fluctuating).
- The Eulerian and Lagrangian approaches for the liquid phase give the same results and none of the two methods can be said superior.
- Reacting cases (RANS only) give access to the flame structure with again a correct trend. However quantitative comparison with experimental data show that the simulations fail to predict emissions (CO and NOx).
Numerical simulation appears to be more and more performant, in particular the LES approach which is truly predictive. Efforts on the development of such tools and their application to industrial configurations must be continued, in particular: the description of the flame chemistry needs to be improved to allow a correct prediction of emissions, important missing features are the description of the atomisation of the liquid, which has a great impact on the flame structure heat transfer must be better described, including conduction in the solid walls and radiation.
Taking advantage of all the knowledge gained in previous European or national projects, the TLC project allowed the development and design of lean injection systems with fuel staging for single annular combustors to mature. Evaluation of various concepts, use of design optimisation procedures, large number of tests, development and application of advanced experimental diagnosis for realistic combustor conditions, strong support of numerical diagnosis with the latest European CFD tools, have contributed to this goal.