NSF Award
2332264
The accurate, precise, and comprehensive representation of shape details of manufactured components is critical for the design, control, and optimization of next-generation, cyber-based manufacturing systems. However, in many cases, the design intent encoded in the Computer-Aided Design (CAD) model cannot be applied directly to the as-manufactured part because the physical part never perfectly matches the geometry of the designed part. Therefore, linking measurements back to the design is challenging. This project establishes a systematic and generalizable mapping methodology between as-designed models and as-built geometries in the presence of large deviations. The methods developed as part of the project define mappings that leverage the design information in the context of the as-built part, which is otherwise impossible with simple reverse engineering. This work will facilitate data mapping from different resources and modalities in manufacturing and across the product lifecycle. Hence this project will enable digital twin systems and Manufacturing 4.0 that rely on connectivity and information exchange between digital models and physical parts. In addition, the integration of education and outreach will broaden participation in manufacturing and train the next generation of engineers and researchers with state-of-the-art CAD modeling, manufacturing, and data analytics expertise. <br/><br/>The overarching goal of this research is to establish a systematic mapping framework between as-designed models and the as-built geometries to be used for applications in next-generation manufacturing systems. The specific research objectives of this project include: (1) creating a rough mesh fit of the as-built point cloud that represents the exterior of the as-built part to enable a mesh-based mapping between as-built and reference geometries. The approach of this project looks at the entire object as a unit and its natural deformations to accommodate large deformations. (2) Fitting the CAD model to the as-built geometry while preserving the design topology information embedded in the original CAD. The approach of this project makes conscious choices to retain the topology of the reconstructed shape. (3) Establishing a systematic mapping between as-designed models and the as-built geometries. The team introduces the idea of persistent mapping that uses a defined procedure based on identified features or landmarks to parameterize the object surface piece by piece. Finally (4), Evaluating the effectiveness and demonstrating the usefulness of these new methods for specialized case studies in additive manufacturing and nondestructive evaluation. The project will facilitate the digital integration of design and manufacturing within larger product lifecycle ecosystems, especially in sectors with requirements for process monitoring and quality assurance, e.g., additive manufacturing and nondestructive evaluation.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
