One of the most severe failures in such aircrafts with reciprocating engines is related to combustion faults of single or multiple cylinders called misfire. It is clear that engine condition monitoring for robust and fast detection of such severe engine failures is very important for each aircraft to ensure safe and comfort flights. Although, most existing smaller aircrafts with reciprocating engines – especially older machines – do not have installed such monitoring and diagnosis systems. The idea behind the work presented here is to develop a monitoring and diagnosis solution which is cheap and easily installable to existing piston aircraft engines, but still with robust and fast failure detection performance.
In contrast to existing methods our solution is based on information acquired from a single acceleration sensor with very simple restrictions for attaching the sensor to the motor block. The sensor position is not critical; only the direction of the sensitive sensor axis needs to be aligned with respect to the primary engine case movement as discussed later. The acquired acceleration information is directly related to rigid engine block movements. We use a two stage hierarchical signal processing scheme based on knowledge from extensive analytical analysis of the power train system for such propeller-based aircrafts. Modal analysis evaluating signal components at relevant engine orders elucidate signal anomalies in case of combustion failures in the mechanical power train system. To ensure robust and accurate detection and identification of combustion failures, a high level engine health condition estimation has been developed to provide quantitative measure of the engine’s or each cylinder’s health condition.
Real world measurements carried out on an experimental aircraft engine test rig are in good agreement with the results predicted by our simulations. Our solution enables a very fast and robust condition monitoring solution for the detection and identification of misfire and other severe combustion failures.