New methodology for demoldability analysis based on volume discretization algorithms

In this paper, a new approach for moldability analysis is presented. A new algorithm based on a discretization method using storage arrays is developed. The discretization method consists of slicing a CAD part model and analyzing 2D intersection profiles by scanning lines along the parting direction and side demolding directions. The intersection points obtained are classified according to the value of their coordinates and assigned to different areas of the mesh. A hierarchical classification in terms of demolding is also proposed. A set of classification criteria for transforming a discrete points mesh into the desired demolding map is introduced. The stored boundary points are evaluated through a new reallocation algorithm of points in mesh in order to compare each point with the remaining points included in the higher order hierarchy demolding zone. The algorithm then reassigns boundary points to a set of hierarchically classified mold zones, and calculates the parting line subsequent to the parting direction. The proposed method allows the moldability analysis to be performed on any system, and no other internal information of the solid is required. Solid geometry remains stored in arrays for further work. This volume discretization method improves the efficiency of the process since it is defined by a much more simple conception by means of massive data storage arrays whose elements are the coordinates of the part boundary points. The resulting demoldability map is displayed on the same CAD part analyzed, providing designers with an evaluation tool which assists them in taking decisions about the mold design.

► Discretization algorithm allows working with the part geometry stored in arrays. ► We proposed a hierarchical classification in terms of demolding. ► Classification criteria transform discrete mesh into demolding map. ► Boundary points are reassigned to a set of hierarchically classified mold zones. ► Moldability is performed on any system, no other solid internal information required.