Computerized distortion minimization of laser beam welded complex components

激光焊接是一种快速、灵活的连接工艺。尽管能量沉积较低，但激光焊接会导致部件变形，从而导致不符合尺寸公差。因此，需要额外的加工步骤，如热或机械矫直。在众多研究工作中，基于模拟的调查有助于更好地理解扭曲，并增强对简单几何形状的成分扭曲的预测。然而，对于复杂的几何形状，为了使组件失真最小而对适当的工艺参数进行数值检测会由于计算成本增加而导致成本上升。此外，建立数值模型的任务包含许多非自动化的处理步骤，需要大量的用户专业知识。因此，这些复杂几何模型的大小需要一种新的方法来计算变形并在合理的时间内最小化变形。因此，本方案的目的是开发和验证一种新的计算机化和高效的方法来预测激光焊接复杂几何形状的合适的工艺参数，并使构件变形最小。这项建议的另一个目标是演示一种可能的方法，使单个建模步骤自动化，并简化用户对过程的控制。为了根据数值方法预测适当的过程参数，复杂的几何图形必须划分为多个子区域--类似于所谓的局部-全局-方法，但具有不同的连接操作。随后，对于残馀应力和构件变形，必须对这些分区进行数值分析。然后，可以将计算出的子区域的扭曲应用于计算整个复杂几何体的扭曲。将该方法与优化算法相结合，将有助于预测合适的工艺参数，使焊接几何形状的构件变形最小。

Laser beam welding is a fast and flexible joining process. Despite the low energy deposition, laser beam welding leads to component distortions and, thus, to non-compliance with dimensional tolerances. Therefore, additional processing steps like thermal or mechanical straightening are required. Simulation based investigations within numerous research works have contributed to a better understanding of distortions and to an enhanced prediction of component distortions of simple geometries. For complex geometries, however, the numerical detection of appropriate process parameters for minimal component distortion leads to rising costs due to increasing computational expenses. Moreover, the task to build the numerical model contains many non-automated processing steps and requires a lot of user expertise. The size of these complex geometry models therefore requires a new approach for the distortion computation and minimization within a reasonable time.Hence, the aim of the proposal is the development and validation of a new computerized and efficient method for the prediction of appropriate process parameters for laser beam welded complex geometries with minimal component distortion. An additional goal of this proposal is the demonstration of a possible way to automate the single modelling steps and to simplify user control of the procedureIn order to predict the appropriate process parameters based on numerical methods, a complex geometry has to be divided into subareas -- similar to the so called local-global-method, but with different joining operations. Subsequently, with regard to the residual stresses and the component distortion, these subareas have to be numerically analyzed. The computed distortions of the subareas can then be applied to compute the distortion of the whole complex geometry. The application of this method in combination with the use of optimization algorithms will facilitate the prediction of appropriate process parameters for minimal component distortion in welded geometries.