Methodology for intelligent injection point location based on geometric algorithms and discrete topologies for virtual digital twin environments

Abstract

This article presents an innovative methodology for locating injection points in injection-molded parts using intelligent models with geometric algorithms for discrete topologies. The first algorithm calculates the center of mass of the discrete model based on the center of mass of each triangular facet in the system, ensuring uniform molten plastic distribution during mold cavity filling. Two sub-algorithms intelligently evaluate the geometry and optimal injection point location. The first sub-algorithm generates a geometric matrix based on a two-dimensional nodal quadrature adapted to the part's bounding box. The second sub-algorithm projects the nodal matrix and associated circular areas orthogonally on the part's surface along the demolding direction. The optimal injection point location is determined by minimizing the distance to the center of mass from the first algorithm's result. This novel methodology has been validated through rheological simulations in six case studies with complex geometries. The results demonstrate uniform and homogeneous molten plastic distribution with minimal pressure loss during the filling phase. Importantly, this methodology does not require expert intervention, reducing time and costs associated with manual injection mold feed system design. It is also adaptable to various design environments and virtual twin systems, not tied to specific CAD software. The validated results surpass the state of the art, offering an agile alternative for digital twin applications in new product design environments, reducing dependence on experts, facilitating designer training, and ultimately cutting costs