Abstract
Evaluating and representing intricate design, simulation, and process outcomes for organic geometries can present limitations on traditional computing devices (computers, phones, and tablets) and remains a formidable challenge. This deficit in recognizing critical manufacturability concerns may result in several design revisions and iterations. Virtual Reality (VR) presents a compelling alternative to conventional 3D environments, particularly when visualizing and analyzing complex organic geometries during the part design and manufacturing stages. Our work focuses on developing an immersive virtual environment (ImVR) to evaluate and visualize design and process parameters for additive manufacturing of complex shapes. We create visualization-centric designs that empower users to render surface and volume-based part models for design exploration to identify and evaluate Design for Additive Manufacturing (DfAM) features. These visualization-centric designs emphasize and enhance the detection of these DfAM features through corresponding visualized part models or through color plots shown to highlight these features. Further, we optimize the build using a genetic algorithm to formulate the best build orientation tailored to DfAM features. We create all the visualization-centric models for the different build orientations found optimally to showcase the impact of optimizing one DfAM feature on other corresponding parameters. On the process side, we implement a Digital Twin to visualize, control, and modify parameters in an FDM 3D printing process. The digital twin model mimics the real-time operation of the 3D printer, showing the generated print layer-by-layer. Bayesian Network, a predictive modeling algorithm, provides the recommended parameters for modifying the printing process. These modifications pertain to optimizing the ultimate tensile strength of the part model by varying bed temperature, speed, and layer height. Two case studies on complex geometries — a Topology-Optimized Bracket and a Lattice Wrench structure are presented to evaluate the capabilities of this toolkit. ImVR leverages inherent abilities of VR, such as multi-user collaboration and human-scale design, to simulate real-world interactions and facilitate a better analysis environment. Finally, an artificial intelligence-driven chatbot by ChatGPT is incorporated to assist the user while analyzing the part model and to provide real-time guidance on the virtual environment and the ImVR. We have trained the chatbot with documents and papers related to DfAM features and process parameters. It helps the chatbot understand the complex terminologies behind DfAM analysis and allows the chatbot to guide the user comprehensively. With multi-feature design and process integration in the immersive landscape, the ImVR toolkit provides distinct advantages of visualizing product design and manufacturing processes in a VR environment, leading to better evaluation of design and saving design iteration costs in the long term.
