一种采用遗传算法优化Chaboche模型的转子轮盘疲劳寿命预测方法  

An Approach for Predicting Fatigue Life of Rotor Wheels with Genetic Algorithm Optimizing Chaboche Model

作　　者：史昱昆 王恭义 程凯 叶笃毅[1] SHI Yu-kun;WANG Gong-yi;CHENG Kai;YE Du-yi(Institute of Process Equipment,Zhejiang University;Shanghai Turbine Works Co.,Ltd.)

机构地区：[1]浙江大学化工机械研究所 [2]上海汽轮机厂有限公司

出　　处：《化工机械》2022年第2期183-189,288,共8页Chemical Engineering & Machinery

基　　金：国家自然科学基金项目(51675475)。

摘　　要：针对汽轮机一体化联合转子的抗疲劳设计,研究了一种采用遗传算法优化Chaboche模型参数的方法,并结合局部应变法进一步建立起考虑平均应力动态松弛的转子轮盘疲劳寿命预测方法。结合某型转子轮盘的疲劳寿命预测,开展了轮槽材料在应变比R=0下的低周疲劳试验与轮槽构件的疲劳试验。结果表明:经参数优化的Chaboche模型能够较好地模拟轮盘材料在各种应变幅下的滞回环形状和平均应力松弛行为,由此建立的寿命预测方法可显著提高轮槽构件的疲劳寿命预测精度,在研究条件下,预测寿命与试验值位于二倍分散带之内。Aiming at fatigue resistant design of turbine’s integrated rotor,making use of genetic algorithm to optimize parameters of Chaboche model was studied,including having local stress strain method combined to establish a fatigue life prediction method which considering mean stress relaxation for the integrated rotor’s wheels.Through combined with the fatigue life prediction of a rotor wheel,both low cycle fatigue test of wheel groove material at strain ratio R=0 and the fatigue test of wheel groove components was implemented to show that,the optimized Chaboche model can simulate hysteresis loop shape well and mean stress relaxation behavior of the rotor material under different strain amplitudes.The life prediction method established thereof can significantly improve the fatigue life prediction accuracy of wheel groove components and the predicted lifetimes and test values stay within the two-fold dispersion band under the present investigation conditions.

关 键 词：汽轮机 一体化转子轮盘 Chaboche模型 平均应力松弛 遗传算法 裂纹萌生寿命预测 

分 类 号：TQ051.2[化学工程]