Fast prediction of thermal distortion in metal powder bed fusion additive manufacturing: Part 1, a thermal circuit network model
Fast prediction of thermal distortion in metal powder bed fusion additive manufacturing: Part 1, a thermal circuit network model
复制标题
DOI：
10.1016/j.addma.2018.05.023
复制
发表时间：
2018-08
影响因子：
11
通讯作者：
Hao Peng;M. Ghasri-Khouzani;S. Gong;R. Attardo;P. Ostiguy;Bernice Aboud Gatrell;Joseph Budzinski;Charles Tomonto;J. Neidig;M. Shankar;R. Billo;D. Go;David Hoelzle
中科院分区：
文献类型：
--
作者：
Hao Peng;M. Ghasri-Khouzani;S. Gong;R. Attardo;P. Ostiguy;Bernice Aboud Gatrell;Joseph Budzinski;Charles Tomonto;J. Neidig;M. Shankar;R. Billo;D. Go;David Hoelzle
增材制造（AM）工艺金属粉末床熔合（PBF）可以快速生产具有与锻造材料相当的机械性能的复杂零件。然而，PBF过程中积累的热应力会导致零件变形，可能导致零件不符合规格，并经常导致工艺故障。这篇手稿是两篇配套手稿中的第一篇，介绍了一种计算效率高的变形和应力预测算法，该算法被设计用于在集成到工艺设计优化例程中时大幅减少计算时间。在这第一稿中，我们介绍了一个热路网络（TCN）模型来估计PBF过程中的部分温度历史，PBF模拟中的一个主要计算瓶颈。在TCN模型中，我们通过将部件划分为热电路元件（TCE）来模拟通过部件和支撑结构的传导热传递，热电路元件由通过电阻器连接的热电容表示的热节点组成，然后以逐层的方式构建TCN以复制PBF过程。与传统的有限元法（FEM）热建模相比，TCN模型预测金属PBF AM部件的温度历史，计算速度快两个数量级以上，同时损失不到15%的准确度。配套手稿说明了如何将温度历史集成到热机械模型中，以预测热应力和变形。
The additive manufacturing (AM) process metal powder bed fusion (PBF) can quickly produce complex parts with mechanical properties comparable to wrought materials. However, thermal stress accumulated during PBF induces part distortion, potentially yielding parts out of specification and frequently process failure. This manuscript is the first of two companion manuscripts that introduce a computationally efficient distortion and stress prediction algorithm that is designed to drastically reduce compute time when integrated in to a process design optimization routine. In this first manuscript, we introduce a thermal circuit network (TCN) model to estimate the part temperature history during PBF, a major computational bottleneck in PBF simulation. In the TCN model, we are modeling conductive heat transfer through both the part and support structure by dividing the part into thermal circuit elements (TCEs), which consists of thermal nodes represented by thermal capacitances that are connected by resistors, and then building the TCN in a layer-by-layer manner to replicate the PBF process. In comparison to conventional finite element method (FEM) thermal modeling, the TCN model predicts the temperature history of metal PBF AM parts with more than two orders of magnitude faster computational speed, while sacrificing less than 15% accuracy. The companion manuscript illustrates how the temperature history is integrated into a thermomechanical model to predict thermal stress and distortion.