为了检查表征疲劳和蠕变疲劳裂纹特征的J积分计算方法精度进行了一系列试验。结论为:(1)在计算疲劳J积分中,用循环应力应变方程代替Kumar的J积分方程中的稳态应力应变方程,可提高简化方法的精度。(2)在计算蠕变疲劳J积分中,用循环载荷保持过程中近似表征蠕变行为的蠕变方程代替Norton形式的蠕变方程,可提高简化方法的精度。 A series of experiments were conducted to examine the accuracy of the J-integration method that characterizes fatigue and creep fatigue cracks. The conclusions are: (1) In the calculation of fatigue J-integral, the steady-state stress-strain equation in Kumar’s J-integral equation can be replaced by the cyclic stress-strain equation, which can improve the accuracy of the simplified method. (2) In the calculation of creep fatigue J-integral, the creep equation approximating creep behavior during cyclic load-holding instead of the Norton-type creep equation can be used to improve the accuracy of the simplified method.