Numerical and experimental research for improved control of compressor surge – NumERICCS

The NumERICCS project aims at coordinating the efforts of the University/CNRS research groups of ENSAM-Lille, ECLille and UPMC, of the french aerospace research organization ONERA, and of the french aircraft-engine manufacturer SNECMA, towards the control of the destructive surge instability observed at the high-load limit of modern axial-flow compressors. This is a combined experimental and numerical work, whose expected outcome is the design, numerical simulation and experimental assessment of an original compressor-surge control device, using nonsynthetic (net massflow injection) jets ahead of the blades' leading-edge. The 48-month project will address multiple technological challenges to achieve this goal.

New experimental data, both with and without flow-control, will be acquired in 2 existing experimental set-ups which will be specifically modified during the project to insert the actuators designed to control the flow:
(1) a SNECMA-designed compressor-stage on the BERAC test-rig at ENSAM-Lille (LML) which will provide detailed aerodynamic data of rotating-stall and surge inception; several LML-patented (ECLille) actuators will be placed circumferentially, near the rotor's leading-edge, and optimized experimentally (massflow, frequency, phase, duty-cycle; LML) and computationally (distance from leading-edge, injection angle; ONERA); detailed measurements of the optimally-controlled flow will then be performed, and compared with full-stage RANS (with deterministic unsteadiness) CFD (ONERA).
(2) a generic nonrotating wind-tunnel set-up of an isolated cantilever blade with tip-clearance at ONERA-Meudon which allows for adjustment of several flow (tip-clearance height, flow-incidence, incoming boundary-layer thickness) and control (location, injection angle, massflow, frequency) parameters, to study their effects on the tip-clerance vortex; extensive numerical computations along with the construction of a surrogate model will be used in the initial phase of control-device optimization; although experimental data from this configuration cannot be directly transposed to compressor-control, they will be used instead to enhance our (limited under the present state-of-the-art) understanding of the influence of these parameters on the evolution of the tip-clearance vortex; furthermore, detailed measurements (LDV/PIV, hot-wire, multihole probes) will be acquired to construct a detailed data-base for LES multiequation subgrid turbulence model validation.

Extensive CFD computations, including advanced RANS models (2-equation, rij-eps and a new 12-equation rij-epsij model; ONERA, UPMC), zonal hybrid RANS/LES (ZDES; ONERA) and VR-LES (variable resolution multi-equation subgrid modelling; UPMC) will be systematically compared with measurements to assess the ability of various levels of modelling to predict the time-dependent controlled flow.

The large number of experimental and computational data will be specifically assessed to determine (1) the performance of different levels of CFD modelling for actively-controlled aeroengine compressors, (2) the possiblity of detecting and using precursors to trigger active control, (3) guidelines for optimal active-flow-control strategy, and (4) directives for future work.