Application of the finite-element method to a design of optimized tool geometry for the O.S.U. formability test

Abstract

A new, plane-strain, sheet-formability test (the O.S.U. Formability Test, OSUFT) has been recently proposed, and it has shown many improvements over the limiting dome height (LDH) test. However, the prototype tool geometry was initially determined arbitrarily for the experiment so that an enhancement of the tool geometry was made with dual purpose: to design the tool geometry to generate consistent planestrain state up to failure under various lubrication states and different testing materials and, at the same time, to make the testing equipment cost as low as possible so that the test may be readily available for small- and medium-scale stamping companies. The latter demands a compact tool geometry to minimize the required press capacity, while the former requires wider blanks that increase the punch load. Considering these conflicting conditions, computer simulation technique using three-dimensional finite-element method was introduced, rather than performing numerous die tryouts, to design the optimal tool geometry from simulative trial and error. By reducing the size of the entire tool and controlling the width-to-length ratio of the blank, an enhanced tool geometry was found that generates stable plane-strain state up to failure and still features low required load capacity for materials with r-values up to 2.0, friction coefficient ranges of 0.15 to approximately 0.35, and thicknesses up to 1.5 mm. The bending-dominant failure due to smaller radii of the tool was avoided. Comparison of LDH simulation showed that the enhanced configuration of the test will produce more proportional strain path and larger plane-strain area near the predicted failure region. It was also predicted that the testing results will be less sensitive to the lubrication state on the tool surface and the material anisotropy of the sheet, which will contribute to a better repeatability of the test. Experiments revealed that the optimized tool showed significantly less scatter in measurement compared with that of the LDH and the original O.S.U. formability tests.