Electrochemical Machining (ECM) is essentially a high-speed corrosion process. It harnesses corrosion to machine very hard metals, thin sections and complex shapes, without the tool ever touching the workpiece. The problem is that the shape does not directly determine the dimensions of the final product, which makes it hard to design the tool for accurate production.
Fluid flow, hydrogen evolution, Joule heating and temperature gradients all affect the final shape. As a result, an Edisonian approach is typically used for designing ECM tools, which go through multiple design iterations to achieve the desired shape on the workpiece.
Our solution is to model the process using High-Performance Computing (HPC), efficient numerical algorithms and multiphysics CFD packages. In this approach, one of the primary remaining challenges is the large computational grid deformation required to model tool motion. This work shows a novel technique where the ECM is modeled using an efficient re-meshing algorithm that allows large deformation without incurring low mesh metrics and/or numerical instabilities.
The numerical predictions agree fairly well with CAD design specifications. This methodology is computationally more efficient than overset methods, although there are round off errors that diminish the beginning with mesh refinement in high-deformation areas.
Leave a Reply