A hybrid hint-based and fuzzy comprehensive evaluation method for optimal parting curve generation in injection mold design

Optimal parting curve contributes to simplify mold structure, reduce the cost of mold manufacturing and injection production. As it depends on a series of technique, such as undercut feature recognition, mold piece region recognition, parting curve assessment, and etc., optimal parting curve generation is still a complicated and time-consuming task. A hybrid hint-based and fuzzy comprehensive evaluation method is proposed in this paper to recognize mold piece region and generate optimal parting curve for 3D CAD model in injection mold design. The hint topological regions are innovatively defined and classified into three categories: concave-edge region, inner loop region, and isolated surface. Based on this, potential mold piece region is defined and recognized by the surface visibility. Exact region is recognized from the potential mold piece region by the hint derived from adjacency relationship of topological regions. Inexact region is recognized from these topological regions that are not exact regions through their adjacent information. Candidate mold piece region set is then constructed by the combination of exact region and inexact region. Boundary edges of each candidate mold piece region are extracted to form the parting curve scheme set. By using the parting curve evaluation model which is based on fuzzy comprehensive evaluation method, the parting curve with the best assessment result is selected as the optimal scheme. Finally, case study is provided to validate the feasibility and effectiveness of the proposed approach.