| description abstract | Abstract. Additive manufacturing (AM) technology enables the creation of complex objects and has significant application potential across various fields. However, as a layer-by-layer construction process, AM is influenced by several factors, with thermal strain being particularly critical in impacting manufacturing performance and product quality. During printing, uneven heat distribution creates significant temperature gradients between layers, resulting in considerable thermal-induced distortion of the object. Therefore, it is crucial to develop an inverse shape optimization method to ensure that the fabricated object closely approximates the desired shape. To achieve this goal, a thermal-induced distortion-driven inverse design method is proposed in this work. In the proposed method, the Hausdorff distance is applied as a metric to quantify the differences between the fabricated object and the desired shape, where the thermal-induced distortions are simulated using the inherent strain method. To ensure the accuracy of the predicted distortions, the initial thermal strain of the printing material is experimentally calibrated. Since the calculation of the shape boundary and the simulation are nondifferentiable, the genetic algorithm is applied to search for a potential global optimal solution. Additionally, a strategy is proposed to reduce the dimensionality of the design variables, and a strategy is introduced to prevent self-intersection during the optimization. To further improve the efficiency of the proposed method, a layer-wise optimization strategy is introduced to generate the initial shape for the genetic algorithm. The effectiveness and robustness of the proposed method have been evaluated on several typical 3D models. Specifically, two models are physical printed and measured, which demonstrates the reliability and validity of our method. | |
