Measuring Method And Experiment Of Bending Strength Of Gear Based On Embedded Mode

With the rapid development of modern production, science and technology, geartransmission is largely applied in complex working conditions, such as extreme environment,overload, etc. Therefore the strength of gear and its measurement, especially the bendingstrength, also put forward higher and newer requirements. The measurement of stress andstrain and assessment plays a vital role in the quality of entire product. Along with continuousinnovation of measuring method, measurement method of engineering stress and strain alsokeeps pace with fast development and it offers strong technical support and theoretical basisfor measurement of gear strength.This paper adopts grooving or drilling hole on the gear to embed micro sensor module toachieve stress measurement, however, the changes of gear structure will affect the stress anddeformation of gear and also affect load bearing capacity of gear. In order to control itsinfluence to a certain range, so grooved or drilled gear is analyzed by FEA. Firstly, necessaryparameters are determined to establish3D entity parameterized model of different embeddedmode gear. Based on UG software, parametric modeling, assembly and motion simulation ofgear is carried out to verify accuracy of built gear. Gear model of different mode is analyzedby ANSYS software based on finite element analysis (FEA) to obtain the gear stressdistribution and maximum stress, and influence of different groove or hole on gear strengthhave been found to determine the optimal result. Meanwhile, it also provides a basis forstructure design of gear.Secondly, in this paper, the bending stress of gear is predicted by neural network basedon MATLAB. The results of finite element analysis of gear bending stress are regarded assamples. Different hole dimensions or groove dimensions are defined as input and bendingstress is considered as output. Nonlinear model of different neural network is established andtrained to attain the relationship between input and output. The bending stress of gear israpidly calculated with high accuracy while trained parameters of neural network are withinsample scope. The foundation for strength design of gear is laid.Finally, through embedded measuring experiment of gear, the measuring results of stressare acquired. According to the result of finite element analysis, the feasibility of theembedded measuring experiment is verified in the application of measurement of bendingstress to provide certain reference significance and guiding significance to actual assessmentof gear strength.