Wu Sung-Heng, Tariq Usman, Joy Ranjit, Mahmood Muhammad Arif, Malik Asad Waqar, Liou Frank
Department of Mechanical Engineering, Missouri University of Science and Technology, Rolla, MO 65409, USA.
Intelligent Systems Center, Missouri University of Science and Technology, Rolla, MO 65409, USA.
Materials (Basel). 2024 Sep 3;17(17):4363. doi: 10.3390/ma17174363.
In directed energy deposition (DED), accurately controlling and predicting melt pool characteristics is essential for ensuring desired material qualities and geometric accuracies. This paper introduces a robust surrogate model based on recurrent neural network (RNN) architectures-Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Gated Recurrent Unit (GRU). Leveraging a time series dataset from multi-physics simulations and a three-factor, three-level experimental design, the model accurately predicts melt pool peak temperatures, lengths, widths, and depths under varying conditions. RNN algorithms, particularly Bi-LSTM, demonstrate high predictive accuracy, with an R-square of 0.983 for melt pool peak temperatures. For melt pool geometry, the GRU-based model excels, achieving R-square values above 0.88 and reducing computation time by at least 29%, showcasing its accuracy and efficiency. The RNN-based surrogate model built in this research enhances understanding of melt pool dynamics and supports precise DED system setups.
在定向能量沉积（DED）中，精确控制和预测熔池特性对于确保所需的材料质量和几何精度至关重要。本文介绍了一种基于循环神经网络（RNN）架构的强大代理模型——长短期记忆（LSTM）、双向LSTM（Bi-LSTM）和门控循环单元（GRU）。利用来自多物理场模拟的时间序列数据集和三因素、三水平的实验设计，该模型能够准确预测不同条件下熔池的峰值温度、长度、宽度和深度。RNN算法，特别是Bi-LSTM，显示出较高的预测精度，熔池峰值温度的决定系数R平方为0.983。对于熔池几何形状，基于GRU的模型表现出色，决定系数R平方值高于0.88，并且计算时间至少减少了29%，展示了其准确性和效率。本研究构建的基于RNN的代理模型增进了对熔池动力学的理解，并支持精确的DED系统设置。
