Application of RSM and Optimization to Plume Signature Analysis

We propose using response surface methodology (RSM) as a tool to better understand the effects of engine design parameters and operating conditions on plume signature, and to improve signature predictions through optimized selection of analysis inputs. Often rocket engine operating conditions and design variables are not well known, either because the systems being simulated are foreign, or because hardware designs prevent the accurate measurement or calculation of key conditions. For example, true injector element flow may be difficult or impossible to obtain due to pressure mal-distribution in the injector manifold. This innovative application of RSM could provide a method of selecting unknown or poorly known inputs by using plume observations as an input selection driver. In fact, inputs could be optimized to determine the most likely engine hardware design and operating conditions based on plume observations. The plume signature predictions resulting from the RSM optimized inputs are bound to be more accurate than inputs developed using current selection methods since the inputs will be selected to reflect known signature data.