中文摘要：

传统投影法事先假定最大主应力为最大拉伸应力,且投影过程未考虑主应力拉压特征,为改进上述不足,提出了一种基于球面方向余弦族投影的车辆结构疲劳参量求解方法。构建了结构节点的方向余弦,搜索多工况下最大拉伸应力的方向。在疲劳最大应力和最小应力的求解过程中,始终保留主应力方向,确保其在投影后仍具备原有的拉压特性。以车辆三轴转向架焊接构架和轴箱体为例,分别采用传统投影法和提出的求解方法分析其疲劳强度。分析结果表明：对于焊接构架,提出的方法确定的节点254570最大拉伸应力较传统投影法高19.4%,2种方法得到的最小应力在27.8%的节点存在较大差异,一些节点最小应力甚至与提出方法的结果正负相悖;比较经修正的等效应力幅值,传统投影法在12%的节点上的计算结果较提出方法的结果低至少1 MPa,个别节点的等效应力甚至低34.73%,依此进行疲劳强度评定可能导致结论偏于危险;对于轴箱体,传统投影法得到的对称循环应力幅值可能偏离实际,某些节点的值甚至较提出的方法低45.32%或高51.23%,导致疲劳分析偏于危险或过于保守。可见,提出的方法确定了最大拉应力所在方向,考虑了主应力拉压特性,比传统投影法更为合理,用于疲劳分析更加可信。

英文摘要：

In the traditional projection method, the maximum principal stress was assumed to be the maximum tensile stress in advance, and the tension/compression properties of principal stress were not taken into account in the projection process. To improve the above inadequacies, a solving method of fatigue parameters based on the projection in spherical direction cosine group was proposed. A series of direction cosines around the points in structure were built, and the direction of the maximum tensile stress under multiple loading conditions was searched. In the process of calculating the maximum stress and the minimum stress, the directions of principal stress were retained, hence, the original tension/compression properties of stresses were preserved. The fatigue strength assessments of welded frame of three-axle bogie and the axle box body of vehicle were analyzed by using both the traditional projection method and the proposed solving method respectively. Analysis result indicates that for the welded frame, the maximum stress of node 254570 obtained by the proposed method is 19.4% higher than the value of the traditional projection method. The large differences of the minimum stresses deduced from two methods exist at 27.8% of nodes, the opposite results even exist in signs of the minimum stresses at several nodes. The revised equivalent stress amplitudes obtained by the traditional projection method are at least 1 MPa lower than the values obtained by the proposed method at 12% of nodes, and the values were even 34. 73% lower at individual node. Therefore, the subsequent fatigue strength assessment based on such fatigue parameters is unsafe. For the axle box body, the symmetrical cyclic stress amplitudes obtained by the traditional projection method deviate from the actual situation. The values of some nodes are even 45.32% lower or 51.23% higher than the values calculated by the proposed method, which will lead to dangerous or over- conservative results in fatigue analysis. The direction of the maximum tensile str