Numerical and Experimental Study of Elastic Recovery in Deep Drawing of Low Carbon and Galvanized Steel Sheet Conical Products

Abstract

Deep Drawing is an important sheet metal forming process. It has numerous applications in different industrial fields. Due to improvements in technological techniques, modern industries focused on products dimensional accuracy especially for the products that have inclined walls since it may suffer from the effect of elastic recovery or what is known as springback. Springback is considered a major defect besides thinning, earing, and other deep drawing defects. It may cause serious deviation from demand dimensions particularly for the products inclined by certain angle such as conical shells or products. This study investigates the effect of wall inclination angle, punch velocity, type and thickness of the sheet material on elastic recovery behaviour. Two types of sheet metal, low carbon (AISI 1008) and galvanized steel sheets, of 110 mm diameters circular blanks at 0.9 and 1.2 mm thickness are formed by tooling set (punch, die, and blank holder). Conical dies were used to execute the experimental work and numerical having inclination angles at 70°, 72°, and 74° where, the punch velocity was 100, 150, and 200 mm/min. Numerical simulation was conducted using ABAQUS 6.14 where dynamic explicit solver was used to preform formation of conical products and static or standard solver to predict the behaviour of springback effect. A comparison between the experimental and numerical simulation was conducted. The results show that elastic recovery decreased with increasing of die wall angle, punch velocity, and sheet or blank thickness while increased with increasing of metal yielding stress. The springback factor rose by 0.003 to 0.007 with die angle increasing by 0.002 to 0.006 with punch velocity and thickness increasing, while it decreased by 0.001 to 0.003 with increasing of yielding point. The numerical simulation results show same tendency and high agreement with experimental results with a maximum discrepancy of 4%.