An Energy Dissipation Model And Life Prediction Of Aluminium Alloy For High Cycle Fatigue

Fatigue performance, as one of the most important mechanical properties ofengineering materials, is foundmental data for design and reliability assessment forengineering component. Fatigue test is required to obtain fatigue data, however,the test procedure is time-consuming and costly, which will prolong the design andproduction cycle as well as increasing cost. In the thesis, experimental andtheoretical methods for fatigue life prediction were studied based on the analysisof energy dissipation of high cycle fatigue.Aiming at the inhomogeneity in microstructure and mechanical performancefor welded joint, fatigue crack initiation characteristics of A7N01aluminium alloywelded specimen were investigated by notch fatigue test. The experimental resultsshow that the differences of fatigue life to failure among base metal, weld metaland heat affected zone (HAZ) are significant, but the difference of fatigue crackinitiation life is slight. There are distinguished differences on the ratio of fatiguecrack initiation life to fatigue life to failure for the three microzones. The stage offatigue crack initiation expends most of the whole fatigue life. Therefore，thefatigue crack initiation life cannot be ignored.Due to the high percentage of fatigue crack initiation life for fatigue failure, amodified high cycle fatigue model based on continuum damage mechanics wasproposed, which takes into account the influence of loading frequency on fatigueproperties. In the proposed model，the effect of strain rate on high-cycle fatiguedamage was considered. Fatigue test results indicate that the model can be appliedto both frequency-sensitive and frequency-insensitive materials.In order to monitor the temperature rise due to increasing of loadingfrequency, a real-time temperature detecting system based on accurate integratedtemperature sensor AD592was developed. The system consists of four AD592CNtemperature sensors, which can measure two actual temperatures and one relativetemperature simultaneously. The influence of variations of environmentaltemperature can be removed by the developed system. Meanwhile, the system,which has advantages of high precision and stability, is suitable to detect the sightchange of temperature during fatigue test in real time.According to the feature of temperature evolution curve, the characteristics ofenergy dissipation during high cycle fatigue were analyzed in macroscopic andmicroscopic scale. It is found that the temperature rise of the specimen is attributedto the movement of defects in materials. Under adiabatic conditions， the variationof internal storage energy in the first phase of the temperature evolution is slight and most of the expended mechanical energy transforms into heat which causestemperature of the specimen to increase rapidly. Thereafter, the increasing incycles causes a greater mobile dislocation density and an increasing in storageenergy. Correspondingly, the increasing rate of temperature falls in the secondphase of the temperature evolution. In the phase of crack instability propagation,the local energy at the crack tip releases quickly which causes a further rapidincrease immediately prior to failure. On the basis of the analysis of temperatureevolution, prediction method and model based on energy dissipation for high cyclefatigue life were proposed. The model has an explicit physics meaning. In addition,the only one parameter in the model is a material dependent constant called"limiting temperature rise for high cycle fatigue failure", which characterizes thecapacity for resisting high cycle fatigue failure. The physics meaning of theparameter can be expressed as the maximum temperature rise of material whenfatigue failure occurs for perfect crystal in adiabatic condition. Consequently, highcycle fatigue life can be determined by measuring the initial slope of thetemperature in several thousands of cycles using the proposed experimentalmethod.Temperature evolution of A7N01-T4aluminium alloy and the weldedspecimens under high-cycle fatigue load were obtained by the developedtemperature detecting system. Experiment results show that the temperature riserate increases with increasing stress amplitude and cyclic frequency. Thetemperature rise rate in the second phase of the temperature evolution is influencedby the heat transfer conditions. The proposed fatigue life prediction model wasverified by the measured temperature evolution and fatigue test data. The resultsshow that the parameters at frequency10Hz and128Hz are almost equal, which isin accord with analyses in theory. Fractographic observations of fatigue specimensshow that the influence of cyclic frequency on fatigue failure mode is insignificantfor both A7N01-T4aluminium alloy and the welded specimens.