| The rubber material of the rubber isolator is prone to fatigue failure under dynamic load cycles.Fatigue failure of rubber is one of the main factors affecting the vibration isolation effect of rubber components.In order to ensure that the rubber vibration isolator will not be fatigued in normal use,it is necessary to accurately evaluate the fatigue life of rubber.By establishing the fatigue life prediction model(S-N curve)or conducting fatigue bench test,the fatigue life of rubber can be effectively predicted.However,the rubber fatigue failure process has the characteristics of randomness and long periodicity.The traditional rubber S-N curve cannot solve the problem of randomness of rubber fatigue life.In addition,the fatigue bench test of rubber vibration isolators has shortcomings such as low efficiency,high cost,and low accuracy.In this paper,the probability analysis method is applied to predict the fatigue life of rubber components.Based on the fatigue test data of rubber components,two basic problems,namely,the probabilistic fatigue life prediction modeling of rubber(P-S-N curve)and accelerated fatigue bench test method,are studied.The main contents of this study are as follows.(1)The modeling method of P-S-N curve of rubber specimens under constant amplitude load is studied.Determining the distribution of fatigue life of rubber components is the key to establishing rubber P-S-N curve.Uniaxial fatigue tests are carried out on dumbbell-shaped cylindrical specimens.In order to solve the problem that it is difficult to determine the distribution of fatigue life of rubber components under small sample data,a method of extending the small sample data and quantifying its distribution law by using support vector machine is proposed.Then,the P-S-N curve of rubber is established.The probabilistic fatigue life of rubber components under constant amplitude load is predicted by using this model.The measured values and predicted values of probabilistic fatigue life of rubber specimens are compared.The results show that the dispersion of rubber fatigue life can solved by establish P-S-N curve.It was found that the predicted life was distributed within 1.5 times the dispersion line of the measured life.(2)The modeling method of residual probability fatigue life prediction of rubber components under variable amplitude load is studied.The establishment of rubber P-S-N curve and cumulative fatigue damage model is the key to predict the residual probabilisty fatigue life of rubber specimens.Firstly,aiming at the deficiency that the rubber P-S-N curve ignores the influence of strain ratio on the fatigue life of rubber specimens,a unified probabilisty fatigue life prediction model(U-P-S-N curve)of rubber considering the influence of strain ratio is established by constructing the equivalent effect amplitude calculation model.Secondly,aiming at the deficiency that the linear fatigue cumulative damage model cannot consider the influence of load sequence and interaction on the fatigue life of rubber specimens,a nonlinear fatigue cumulative damage model is proposed by constructing equivalent driving strain.Finally,based on the U-P-S-N curve of rubber and the nonlinear cumulative fatigue damage model,the residual probabilisty fatigue life prediction model of rubber specimens under variable amplitude load is established.The residual probability fatigue life of rubber specimens under variable amplitude load is predicted by using this model.The measured and predicted values of the residual probabilisty fatigue life of rubber specimens are compared.The results shown that the proposed model can improve the prediction accuracy of rubber residual fatigue life under variable amplitude load.It was found that the relative error between the predicted and measured values of the remaining probability fatigue life was less than 50%.(3)The P-S-N modeling method for rubber isolator under random load is studied.The compilation of load spectrum is the key to predicting the fatigue life of rubber isolator under random load.A modeling method for rubber isolator P-S-N based on load spectrum compilation is proposed.Based on the calculation model of equivalent amplitude,the load spectrum is simplified into equivalent load blocks.On the basis of equivalent load blocks,according to the P-S-N curve and the nonlinear fatigue cumulative damage criterion,the probabilistic fatigue life of mount under random load is predicted.In addition,the fatigue bench test of mount under load spectrum is carried out,and the predicted value and measured values of mount probabilisty fatigue life are compared.The results shown that the proposed method can accurately predict the probabilistic fatigue life of rubber mount under random load.(4)The accelerated fatigue bench test method of rubber isolator is studied.Load spectrum editing is an effective method to improve the efficiency of rubber isolator fatigue bench test.The existing load spectrum editing methods have the disadvantages of low identification accuracy and lack of adaptive window length selection when identifying small damage components.To solve this problem,taking rubber mounts as the research object,a load spectrum editing method of rubber isolators based on variational modal decomposition is proposed.The edited load spectrum is obtained by the proposed editing method.For comparison,four existing editing methods based on time domain pseudo-damage preservation,short-time Fourier transform,wavelet transform and S-transform are used to edit the load spectrum,respectively.On the premise of the same pseudo-damage retention,the editing effects of five different editing methods were compared from the aspects of time compression,statistical parameter error,power spectral density,and level count.In addition,the loading effects of the original load spectrum and the edited load spectrum are compared through the fatigue simulation of the mount.The results shown that the efficiency of fatigue bench test can be improved by the proposed load spectrum editing method and the percentage reduction in time was approximately 16.09%. |
PDF Full Text Request