Surrogate Modeling of Melt Pool Thermal Field Using Deep Learning

Abstract

Powder-based additive manufacturing has transformed the manufacturing industryover the last decade. In Laser Powder Bed Fusion, a specific part is built in aniterative manner in which two-dimensional cross-sections are formed on top of eachother by melting and fusing the proper areas of the powder bed. In this process, thebehavior of the melt pool and its thermal field has a very important role in predictingthe quality of the manufactured part and its possible defects. However, the simulationof such a complex phenomenon is usually very time-consuming and requires hugecomputational resources. Flow-3D is one of the software packages capable of executingsuch simulations using iterative numerical solvers. In this work, we create threedatasets of single-trail processes using Flow-3D and use them to train a convolutionalneural network capable of predicting the behavior of the three-dimensional thermalfield of the melt pool solely by taking three parameters as input: laser power, laservelocity, and time step. The CNN achieves a relative Root Mean Squared Error of2% to 3% for the temperature field and an average Intersection over Union score of80% to 90% in predicting the melt pool area. Moreover, since time is included asone of the inputs of the model, the thermal field can be instantly obtained for anyarbitrary time step without the need to iterate and compute all the steps.

Keywords: Additive manufacturing, Laser powder bed fusion, Melt PoolTemperature, Convolutional Neural Network, Surrogate Model