CAJ | 학술논문

传统的焊点疲劳仿真分析方法主要有基于力的LBF方法和基于应力的LMS方法，针对这两种方法所存在的预测精度不高、建模和计算效率偏低等问题，提出一种新的焊点疲劳寿命预测方法，该方法从断裂力学的角度考虑焊点的疲劳失效，采用一种模块化焊点模型，通过有限元提取焊点周围各节点的节点力和力矩，进而推算出焊点裂纹在板厚扩展路径上的平均应力强度因子作为疲劳寿命的评价参量。通过对HSLA340GI和DP600GI两种高强钢材料、4种不同厚度的电阻点焊接头进行剪切疲劳试验，将该平均应力强度因子与试验所得的焊点疲劳寿命数据进行双对数回归分析，得到一条平均应力强度因子与疲劳寿命的拟合曲线作为焊点的疲劳寿命预测曲线，并与传统的LBF和LMS方法进行预测精度的对比分析，结果表明本文的焊点疲劳寿命预测方法的预测精度高于传统的LBF和LMS方法。
전통적한점피로방진분석방법주요유기우력적LBF방법화기우응력적LMS방법，침대저량충방법소존재적예측정도불고、건모화계산효솔편저등문제，제출일충신적한점피로수명예측방법，해방법종단렬역학적각도고필한점적피로실효，채용일충모괴화한점모형，통과유한원제취한점주위각절점적절점력화력구，진이추산출한점렬문재판후확전로경상적평균응력강도인자작위피로수명적평개삼량。통과대HSLA340GI화DP600GI량충고강강재료、4충불동후도적전조점한접두진행전절피로시험，장해평균응력강도인자여시험소득적한점피로수명수거진행쌍대수회귀분석，득도일조평균응력강도인자여피로수명적의합곡선작위한점적피로수명예측곡선，병여전통적LBF화LMS방법진행예측정도적대비분석，결과표명본문적한점피로수명예측방법적예측정도고우전통적LBF화LMS방법。
The traditional fatigue simulation analysis methods of spot weld joints are mainly LBF method based on force and LMS method based on stress. As the prediction accuracy is not high, modeling and computational efficiency is low and other issues for these two methods, a new fatigue life prediction method for spot welds is proposed, which considers the fatigue failure of spot welds from the point of fracture mechanics. A modular spot weld model is adopted, the node force and moment around the spot weld is extracted through the finite element analysis, then the average stress intensity factor through the crack propagation path is calculated as the evaluation parameters of fatigue life. Shear fatigue test is conducted for two high-strength steel—HSLA340GI and DP600GI, four different thickness resistance spot welding joints, then double logarithmic regression analysis is conducted for the average stress intensity factor and the fatigue life data of spot weld joints, then a curve between the average stress intensity factor and the fatigue life is obtained as the fatigue life prediction curve for spot welds, then comparative analysis of prediction accuracy is conducted among these three method, the results show that the prediction accuracy of fatigue life forecasting method of spot welds in this paper is higher than traditional LBF and LMS methods.