A Design Optimization Framework for Enhanced Compressor Stability Using Dynamic System Modeling


This thesis constitutes the second year effort of a joint engineering project initiated in 2001 by Snecma Moteurs, ENSAE, ECL and MIT. The long term objective of this joint project is to conceive, design, implement and operate an advanced core compressor for an unmanned air vehicle. This thesis addresses the issue of compressor design in the light of enhanced stability, introducing a novel approach with respect to industry-practice: compressor dynamic stability is to be considered as a prime design variable. This present study focuses on the development and implementation of a compressor design optimization framework for enhanced stability based on an existing reduced order dynamic system modeling approach [15]. Stall margin, a common metric for compressor stability, is chosen as an optimization objective in the light of performance constraints. Modifications of the shape of the bladerow loss buckets are the optimization design variables since they directly impact compressor performance and dynamic stability. An optimization framework is defined, with the goal to redesign a baseline compressor through geometric modifications, optimized for enhanced stability. The framework is comprised of a mean line calculation including end-wall effects, and the computation of the unsteady flow field perturbations using the existing dynamic compression system model. A solution to the resulting inverse-blade-design problem introduced by the optimization problem is devised. Changes in bladerow loss buckets are linked analytically to a modification of the blade channel angle. The inverse-design optimization framework is implemented on a 3-repeating-stage compressor leading to a 6.4% improvement of the stall margin while changes in performance are kept to a minimum of 2% loss in pressure ratio and 0.5% loss in efficiency. The definition of a new compressor geometry is obtained from the baseline compressor through the introduction of modifications to the blade channel angles of the blade-rows which range between -0.3' and 2.7. Thesis Supervisor: Professor Zoltin Spakovszky Title: C.R. Soderberg Assistant Professor of Aeronautics and Astronautics

Extracted Key Phrases

64 Figures and Tables

Cite this paper

@inproceedings{Perrot2013ADO, title={A Design Optimization Framework for Enhanced Compressor Stability Using Dynamic System Modeling}, author={Vincent Perrot}, year={2013} }