PSI - Issue 35

İbrahim Yelek et al. / Procedia Structural Integrity 35 (2022) 51 – 58 Yelek et al./ Structural Integrity Procedia 00 (2019) 000 – 000

52 2

1. Introduction Hemming is a forming process that is applied by folding the edge regions of the sheet metal on itself. The hemming process is a bending process that is applied intensively for purposes such as joining more than one piece as Hamedon et al. (2014) investigated in their work, obtaining a more durable structure in the edge areas, and improving the visuality of the structure. It is used in many manufacturing sectors such as automotive, white goods, electrical panels. The Hemming process is generally divided into two as conventional die hemming and roller hemming and also there are different types of hemming operations such as robot-assisted roller hemming, laser assisted roller hemming and hemming with electromagnetic forming processes as stated by Dewang and Sharma (2019) in their research. In the hemming process with the conventional die, the edges of the sheet are folded by a die. There are different types such as 90° bend, 135° bend, pre -hemming, open hemming, closed hemming as it is seen in Hamedon et al. (2014) research. In the hemming process with roll, the automated robot can complete the hemming operation on a special route by moving the hemming roll. As in every metal forming process, cracking or tearing problems may occur on the shaped parts during this process. In order to detect such problems during design, the finite element method, as used by Le Maoût et al. (2010), is frequently used. Numerical calculations are of great importance especially to determine the suitability of material selection for the process. To make the numerical calculations correctly, the modelled process should be transferred to the finite element environment as close to real life. For this reason, as reported by Le Maoût et al. (2010), determining the basic forming properties of the sheet material such as its strength, elastic and plastic behaviour and anisotropic property significantly increase the accuracy percentage of numerical calculations compared to the real scenario. Accordingly, in order to use the true-stress true-strain graphs obtained from the tensile test results of the material up to the maximum stress and to model the hardening behaviour after this point to breaking point in numerical calculations, the post-necking stress-strain curve should be determined with mathematical equations or iterative studies, as Zhao et al. (2016) did in their work. In numerical calculations, fracture models are used, as suggested by Bai and Wierzbicki (2010), to examine the fracture cases that occur due to high plastic deformations on the material during the process. Bai and Wierzbicki (2010) used this model for ductile metals based on the Mohr-Coulomb fracture criterion, which is widely used for modelling the correct mechanical behaviour of materials such as crack-free rock and soils with the help of hydrostatic pressure and lode angle. By the way, Mohr and Marcadet (2015) provided a more accurate estimation of fracture for biaxial loading using the Hosford-Coulomb model they presented. Also, they stated that the direct use of a Mohr-Coulomb criterion systematically underestimates the stress up to fracture for biaxial loading. The Hosford Coulomb fracture model is used for ductile fracture estimation according to stress triaxiality, strain to fracture and lode angle parameters. With the aim of capturing variable stress triaxiality values during the experimental work, Roth and Mohr (2016) conducted mechanical tests with different geometries such as tensile test specimen with a hole or notched tensile test specimens by using different test machines. With help of these proposed experiments, they managed to keep the stress triaxiality value constant from the beginning of the test until the onset of the fracture point. Similarly, to determine the ductile fracture behaviour of structural steels multi-axially, Kõrgesaar et al. (2018) revealed the limits of the material by using different test specimens. After determining the ductile fracture limits for different stress states of the material, they used equivalent strain values for the generation of fracture model in finite element analysis to examine the penetration response of stiffened panels used in shipbuilding. On the other hand, Kõrgesaar (2019) compared the numerical data with the experimental data of the damages that occurred in ship accidents using different fracture criteria. For bending and hemming type of processes, Sarkar (2020) worked on the possible causes in detail regarding the cracks that occur in the parts undergoing the bending process. Oxide compounds that occur in the inner structure of the material during production create pollution and pave the way for crack formation. For this reason, the pre determination of the minimum bend radius of the material to be used ensures that such events are prevented before production. Based on these studies in literature, the material that is damaged on the bending surface during the hemming process has been investigated with numerical calculations in this study. In the hemming process, DX51D grade material is used for panel production. According to EN 10346:2015 standard, for DX51D grade, there is no restriction for yield stress. Because of this situation, DX51D productions can contain a large range of stress values for yield stress depending on raw material quality and production process parameters. Due to the large range of yield