| Rubber isolators are usually subjected to substantial static and dynamic loads, and oftenfail due to nucleation and growth of defects or cracks. Prevention of such mechanical fatiguefailures necessitates thorough understanding of the deformation mechanisms of the rubbermaterials during cyclic loading, so as to predict the fatigue life of rubber components moreaccurately. The investigation on the mechanical fatigue of rubbers has become one of thecurrent international hot issues. In this thesis，a series of fatigue behaviors for filled natrualrubbers under uniaxial tension, uniaxial tension/compression, and multiaxial loads, areinvestigated in the frame of mechanics of fatigue design. The major studies can besummarized as follows:(i) Two factors that might influence the tension fatigue model of filled natural rubbersused in rubber isolators are investigated. One is the damage parameter and the other is thespecimen geometry used in the fatigue experiment. The uniaxial tension fatigue experimentsare carried out for three typical types of filled natural rubber specimens: a dumbbell simpletension specimen (STS), a dumbbell cylindrical specimen (DCS), and a hollow cylindricalspecimen (HCS). The commonly used damage parameters (Green-Lagrange strain,Almansi-Euler strain, engineering strain, logarithmic strain, stretch ratio, etc.) for fatigue lifeprediction are described and discussed. The fatigue life prediction models using the measuredtension fatigue life of the STS and different damage parameters are developed, and acorrelation coefficient is used to characterize the error between the measured fatigue life andthe estimated one using the developed fatigue life prediction model. It is concluded that alldamage parameters discussed in this paper can be used to estimate tension fatigue life withcorrelation coefficient greater than0.9. The fatigue life model with the STS is appropriate topredict the fatigue life of the DCS and the HCS, which shows that the relationship betweenthe tension fatigue life and the damage parameters is independent of the geometry of thespecimens. One can thus carry out tension fatigue test using only a STS to model the tensionfatigue of rubbers.(ii) The effect of strain ratio R on the fatigue life of filled natural rubbers used in rubberisolators is investigated experimentally and numerically. A uniaxial tension/compressionfatigue experiment is conducted on DCS rubber specimens subjected to loads representingdifferent R-ratios. The experimental fatigue data are used to formulate two preliminary fatiguemodels based on peak strain and strain amplitude as the damage parameters. The deficienciesof these two models in predicting fatigue life over a wide range of R ratios are discussed, and an alternative life prediction model is proposed. The proposed model incorporates the effectof R-ratio using an equivalent strain amplitude. It is shown that the proposed model couldeffectively predict fatigue life over a wide range of R-ratios with an improved accuracy,praticually for loads of negative R ratio.(iii) The fatigue crack growth (FCG) experiment and modeling method for filled naturalrubbers used in rubber isolators under variable amplitude loads are carried out using anedge-crack pure shear specimen. Variable amplitude loads are imposed on the edge-crack pureshear specimen, and such load provides a more effective way of obtaining the measured FCGdata under different load levels than the conventional constant amplitude load. The commonlyused data processing techniques for getting the crack growth rate (crank length versus numberof cycles) in metal materials, the secant method or the local incremental polynomial method,are not applicable for computing the crack growth rate of rubber material, since the FCG dataunder the variable amplitude loads embodies large fluctuating locally. Based on the knownFCG law and the measured FCG data under variable amplitude loads, a power function isproposed to fit the measured crack growth length and number of cycles using the least-squarestechnique. The crack growth rate is thus calculated from the determined power function, and aFCG prediction model for filled natural rubbers is established from the crack growth data andthe associated tearing energy. To validate the developed FCG model, a dumbbell specimenmade of the same rubber compound as the pure shear specimen is manufactured and is used tocarry out the tensional fatigue experiment. The comparsions between the measured tensionalfatigue life of the dumbbell specimen and that evaluated from the established FCG modelvalidates the proposed data processing method for FCG data of filled natural rubbers underthe variable loads.(iv) A new method for calculation of cracking energy density (CED), which is related tothe tearing energy of rubbers under multiaxial loads, is proposed in the frame of continuummechanics. Using the measured fatigue crack growth characteristic of rubber materials and thecalculated CED of rubbers under external loads, one can predict fatigue life of rubbercomponents. To calculate CED using the output strain from the finite element software(ABAQUS) as inputs, the formula of CED under the principal coordinate system is derivedand the required integral technology is given. Six hyperelastic constitutive models (Ogden,Mooney-Rivlin, Neo-Hookean, Yeoh, Arruda-Boyce and Van der Waals) for rubber materialsare used in the method for calculating CED. It is shown that the CED calculated from theproposed method is valid and effective to unify the uniaxial and equiaxial fatigue life ofrubber materials. (v) The multiaxial fatigue life of rubber isolators is predicted by combing the calcualtedCED and the FCG model of the studied rubber material. Using the measured FCGcharacteristic of rubber material and the calculated CED of rubbers under external loads, aformula for calculating the fatigue life of rubber components under multiaxial loads isestablished, and is applied for predicting a typical rubber isolator. It is shown that thepredicted fatigue life of the rubber isolator agree well with the measured fatigue life within afactor of two, and the predicted crack location and orientation are comparable with themeasured results. The fatigue life prediction method for rubber isolators under multiaxialloads can thus be used as an effective and low cost tool for up-front knowledge of rubbercomponents in the design stage. |
PDF Full Text Request