含裂纹铝合金板单面修补结构疲劳裂纹扩展分析 *

Analysis of Fatigue Crack Propagation for Repaired Aluminum Alloy Plate Containing Crack with Single Patch

关键词:  斜裂纹前沿  垂直裂纹前沿  有限元模型  有效应力强度因子  疲劳寿命  疲劳裂纹扩展

Key words:  Skew crack front  Uniform crack front  Finite element model  Effective stress intensity factor  Fatigue life  Fatigue crack propagation

摘要:

为了研究单面修补结构疲劳裂纹的扩展规律，进行了玻璃纤维/环氧树脂复合材料单面修补含裂纹铝合金板的疲劳试验，并建立了基于斜裂纹前沿(UCF，Uniform Crack Front)和基于垂直裂纹前沿(SCF，Skew Crack Front)的三维有限元模型。利用数理统计方法和有限元模型得到了不同裂纹长度时的有效应力强度因子(ESIF，Effective Stress Intensity Factor)值，采用应力强度因子修正法从疲劳试验结果中获得了修补结构不同裂纹长度时应力强度因子(SIF，Stress Intensity Factor)值[Kexp]，并使用有限元模型的ESIF值和[Kexp]预测了修补结构的疲劳寿命。通过分析得到以下结论:同基于UCF有限元模型相比，基于SCF有限元模型的ESIF更接近于[Kexp]。在预测疲劳寿命时，基于UCF有限元模型预测结果误差较大，与试验中值寿命的最大误差可达到19%；而基于SCF有限元模型各ESIF预测结果与试验中值寿命的误差都在6%之内，与试验结果一致性较好。

Abstract:

In order to study the fatigue crack growth behavior of repaired plate with single-side composite patch， single-edge cracked aluminum alloy plates repaired with single-side glass fiber/epoxy resin composite patch were subjected to fatigue test， and two three-dimensional finite element models of repaired structures based on Uniform Crack Front (UCF) and Skew Crack Front (SCF) were established. By using the mathematical statistic method and finite element models， several Effective Stress Intensity Factors (ESIF) for different crack length were obtained， which were compared with Stress Intensity Factor (SIF) [Kexp] from experimental results by SIF correction method. The fatigue life of repaired structures were predicted by ESIF and[Kexp]. It is concluded that， compared with finite element model based on UCF， the value of SIF from finite element model based on UCF is more close to that from experimental results. When predicting the fatigue life of repaired structures， the biggest error between the results from finite element model based on UCF and the median fatigue life from experimental results is 19%， while the errors between the results from finite element model based on SCF and the median fatigue life from experimental results are within 6%， which illustrating that finite element model based on SCF is more suitable for predicting the fatigue life.