LAUSR

Bearing is a crucial transmission component of aero-engines under high-speed and heavy-load conditions. To design a roller bearing with high loading capacity and reliability, it is essential to focus on the relationship between the external load state (reaction force) and the internal load distribution (load distribution). Therefore, a measurement method for bearing reaction force using load distribution and radial basis function neural network is presented in this study. Unlike conventional static reaction force measurement methods, both the direction and magnitude of the reaction force are considered in the proposed method without modifications to bearing. First, an experimental system is designed to investigate the load distribution in a roller bearing under different reaction forces using strain variation measurements. Then, a finite element analysis is conducted, and simulation results of the strain variations at three interested points match well with the experimental measurements. Finally, a radial basis function neural network with strong nonlinear fitting ability is applied to construct the mapping relationship between strain variation and reaction force. The results demonstrate that the proposed method can predict the reaction force with high accuracy based on the strain variation at three measuring points.