Exploring the effects of removing process-intrinsic constraints on gas turbine design

Abstract

Design by decomposition makes complex design problems tractable. However, decomposition also requires fixing key interdisciplinary design variables, decisions that often lock in much of the final product performance, at the very beginning of the design process. We suggest that a design process based on decomposition thus introduces "processintrinsic " constraints, in addition to the "hard" constraints driven by physical or customer requirements, which can have important and unexpected implications on the achievable performance in later design stages. Using an integrated design methodology which preserves the inherent multidisciplinary coupling between components and disciplines and does not require assumption of the relevant interactions, the advantages of removing, or "softening," process-intrinsic constraints can be explored. In this paper we investigate how removing these constraints may improve the adiabatic efficiencies of a high pressure compressor and turbine spool and thus reduce specific fuel consumption. The results indicate that integrating the aerothermal and mechanical analyses can reduce specific fuel consumption by 0.20% but if the process-intrinsic constraints on shaft work, component weight, shaft speed, and bleed air cooling flows are removed a reduction of 0.42% is possible. The cost performance benefit of removing constraints exhibited bimodal behavior. Conservative performance improvements were achieved with and without process-intrinsic constraints but design time decreased as more constraints were removed. More aggressive improvements were achievable only by removing constraints but only with significant computational expense.