In order to verify the correctness of the above crack propagation, the crack length and crack tip stress intensity factor data obtained from numerical simulation and test are plotted and fitted in origin software. It can be seen that when the stress ratio r = 0.6, the test is relatively stable, and more test data are collected by extensometer, Through the processing and analysis of the data, the relevant conclusions can be obtained more accurately. Therefore, when numerically simulating the stress intensity factor at the crack tip, the relevant data with stress ratio r = 0.6 is selected for analysis. The figure shows the relationship between crack length and stress intensity factor in numerical simulation and test when r = 0.6.
It can be seen from the figure that for the stress intensity factor at the crack tip under the same fatigue crack length, the development trend of the numerical simulation value is the same as that measured in the test, which increases with the increase of the crack length. However, the stress intensity factor at the crack tip obtained by numerical simulation is always less than the experimental value. When the fatigue crack length ranges from 9mm to 25mm, the two curves basically remain parallel, the gap is small, and roughly meet the linear growth relationship. With the increase of fatigue crack length, the stress intensity factor curve at the crack tip obtained by numerical simulation continues to maintain a close linear relationship, while the experimental value shows a nonlinear relationship, and the distance between the two curves gradually increases.
It can be seen that the compact tensile specimen in the test is sensitive to the change of crack length, and the crack length has a great influence on the change gradient of stress-strain field. Therefore, the actual measured data range value changes greatly, especially in the later stage of stable growth stage and rapid growth stage. This is because the processing defects such as inclusions and pores are not excluded in the actual rolling process of the steel plate. The numerical simulation is the ideal state of the test, and the set material properties are uniform, stable and isotropic. Therefore, the force received by the model in the simulation process is also uniform and stable, and is not affected by unstable factors such as processing defects of the material itself and test equipment. It is more reliable and accurate in simulating the stable propagation stage of fatigue crack propagation of compact tensile specimen of 42CrMo steel.