Methods for Diagnosing and Preventing Compressor Surge of an Aircraft Engine During the Start Phase

This paper presents work about spotting and stopping compressor surge as aircraft engines start. It puts forward a method using models which combines how gas flows, sound traits, and readings from electromechanical parts to guess when rotating stall and surge might occur. This includes a Start-Phase Surge Index (SPSI), a way to decide things using two levels (“hold” or “stop & motor”), and closely watched control of valves that let air out and blades that direct air. This method comes from studying conditions where things don't stay steady, the physics of why instability happens, tests of surge events, machine learning guided by models, and control using deep-RL to keep surge from happening. Basic steps and checks after repairs are added for engineers and people looking into incidents. The limitations of what can be spotted, what data is needed, and how airlines can use this are talked about. This work could keep incidents from happening, make start steps better, and teach ground support workers.