Probabilistic fatigue life prediction for CSS-42L bearing in jet strengthen modification grinding using an improved WTP network

Abstract

Considering the high quality requirement of advanced aeroengine equipment, accurate prediction of probabilistic fatigue life (PFL) for CSS-42 L bearing in jet strengthen modification grinding, is an important issue for aeroengine manufacturing. Here an improved robust prediction algorithm of Wavelet Transform Package (WTP) Network is designed, therefore the decisive influence of jet strengthen modification grinding on the probabilistic fatigue life of aeroengine bearing, could be predicted and analyzed. A complete set of comparative experiments demonstrate that, when Pw is 23.5 MPa, Mc is 225 g, Mb is 342 g, Ma is 304 g, Fa is 0.26 kg/min, and Ts is 40.7 mm/min, and Bf is 69.442%, Cs is 70.336%, Nb is 69.002%, Sb is 54.334%, Os is 67.334%, Fp is 71.254%, At is 33.557%, W is 56.375%, Tr is 62.445%, an optimal data set of FLC is 60.5 × 105, PFLS is 0.337, FLGV is 83.74 × 102, CMFD is 8.42, FDP is 79.43, and NPFD is 4.445 (error tolerance = ±5%), could be obtained. This research is novel and merited in implementing novel PFL prediction for CSS-42 L bearing, proposing PFL indexes to evaluate the fatigue performance of bearing material in jet strengthen modification grinding, and designing an improved WTP network to achieve accurate PFL prediction as well. Eventually it could be summarized that this research realizes extraordinary predictive capability and facilitates the productive PFL management for aeroengine equipment.

Keywords

• Probabilistic fatigue life
• Prediction
• Aeroengine bearing
• Jet strengthen modification grinding
• WTP network