Analysis On Several Key Technologies Of The Gear Contact Fatigue Test

The gear contact fatigue test mainly studies on gear pitting, fatigue life time and fatigue stress intensity. With practical project as its background, several aspects about gear contact fatigue testing were studied. The servo loading system was simulated since the accuracy of loading control affects the precision and reliability of testing. The method of getting pitting area ratio using digital image processing was studied because of it is the criterion of gear pitting failure. Aiming at the dispersed gear lifetime datum, the statistical processing of testing datum was studied. The main researches lie in the following aspects:(1) Different loading apparatuses were compared after introducing the importance of gear testing and classes of gear fatigue test rig. Observation methods of tooth face pitting were illuminated and compared. The purpose of testing datum statistical processing was illustrated.(2) According to the structure of test rig, the principle and demands of loading were analyzed. After establishing mathematical model of hydraulic loader and transfer function of loading control system, the static and dynamic performance of loading system were simulated. After combining PID controller, the system response of several input signal were analyzed.(3) The approaches of capturing image, image processing and analyzing were established. The processing steps were consisted of calibration, pretreatment, threshold segmentation and feature extracting. The gray transform approach based on image gray mean and variance was put forward. Two threshold segmentations were used for images with different characters. The character parameters of pitting image were set. A highly effective area label incorporation algorithm was put forward. Mathematical morphology operations of opening and closing were used to remove pit noises and small holes. In addition, large holes were filled with seed fill algorithm. Besides, the corresponding software was developed.(4) The character parameters of gear fatigue testing datum were discussed. After discussing common distributing functions of lifetime datum, parameter estimating, distributing fitting, estimating approaches of lifetime with reliability and degree of confidence were studied. Different distribution checkouts post that three parameter Weibull probability distribution is the best distribution to gear contact fatigue lifetime datum. The comparison between two stress-lifetime models indicates that Basquin model is more suitable for fitting S-N curve. Besides, the corresponding data processing software was developed.