A Digital Twin for Bead Based Additive Manufacturing Processes : Application to Large Format Polylactic Acid Additive Manufacturing
Please login to view abstract download link
Bead-based additive manufacturing are processes in which large parts are built by sequentially assembling elongated beads. They cover a wide range of materials including metals, polymers, ceramics and concrete. However, few numerical techniques are both fast enough to enable real-time process control and capable of predicting the large-scale effects of process parameters. Numerical efficiency is crucial for optimization loops in order to leverage the geometric flexibility allowed by digital fabrication. A digital twin encompassing the FastScan thermal model [1, 2] weakly coupled with the QuadWire mechanical model [2]. The digital twin has been tailored to polymer-based large format additive manufacturing (LFAM). LFAM main challenges are the intrinsic anisotropy of the process and the slow cooling rate of the material. Large deformation occurs due to the reduced stiffness of polymers at high temperatures, coupled with the build-up of residual stresses from thermochemical shrinkage during cooling. This can cause both geometric (misalignment of the printing head and the structure) or structural (layer delamination, warping and debonding) failure. The experimental validation of the digital twin on thin-walled structure using infrared thermometry and digital image correlation is presented in [4]. This contribution focuses on using the digital twin to determine process parameters (e.g. dwell time, substrate temperature, robot velocity) to avoid excessive residual stresses and geometric miss-match between the nozzle and the printed part. Hence, the proposed digital twin represents a rapid numerical tool for design and optimizing manufacturing conditions and improving the quality and manufacturability of LFAM parts.