Tooth Root Stress Prediction of Helical Gear Pair With Misalignment Errors Based on PSO-BP Neural Network

Misalignment errors (MEs) cause uneven load distribution across the faces of the gear teeth that increase contact and tooth root stress (TRS), moving the peak TRS to the edge of the face width, seriously affecting the gear system transmission performance. This article employs the finite element method (FEM) to calculate the TRS of helical gear pairs under different MEs. The effect of MEs on the distribution of TRS is analyzed, and the characteristics of TRS distribution under different types of MEs are determined. Subsequently, a TRS testing platform is established, and the accuracy of the finite element model is validated through strain tests by introducing MEs through changes in the position of journal bearings. Finally, based on the particle swarm optimization-improved backpropagation (PSO-BP) neural network, a correlation model between MEs and the distribution characteristics of TRS along the tooth width is established. Utilizing this model, it is possible to calculate the maximum TRS at different positions of the helical gear tooth root under any MEs.