Engine Load Prediction during Take-Off for the V2500 Engine

Abstract

The aviation industry faces an ever increasing pressure to reduce its cost in order to gain competitive advantages. Since aircraft maintenance contributes strongly with about 17% to the overall direct operating cost (DOC), maintenance providers are required to continuously reduce their cost share as well. As a result, a lot of effort is put into the exploitation of the potential of emerging digitalization technologies to predict upcoming system faults and, therefore, reduce the projected maintenance impact. The detection of early stage faults and prediction of remaining useful lifetimes (RUL) for various systems, including aircraft engines as high-value assets, has been a focal point for many research activities already. A key aspect – necessary for an accurate prediction of future behavior – is the correct mapping of ambient conditions that have led to the respective system condition. Therefore, it is necessary to combine data information throughout an aircraft’s life from different stakeholders to gain valuable insights. However, as the aviation industry is strongly segregated with many parties involved, trying to gain their own competitive advantage, the required information about the operating condition is often not available to independent maintenance providers. Thus, modeling engine degradation often needs to rely on estimated nominal conditions, limiting the ability to precisely predict engine faults. With this paper, we will develop a model that allows users to estimate the experienced engine load during take-off by only using publicly available information, i.e. airport weather information reports and public flight data. The calculated engine load factors are computed in terms of an engine pressure ratio (EPR) derate. The results are benchmarked with the actual engine derate, obtained for different operators and various ambient conditions, to enable an identification of challenges for the load prediction and areas of improvement. The developed model will help to adjust engine failure projections according to the experienced ambient conditions and, therefore, supports the development of better engine degradation models.