Evaluation of in-plane edge stretchability under severe contact condition for third-generation advanced high-strength steel

Artificial Neural Network for Predicting Edge Stretchability in Hole Expansion Test With Gpa-Grade Steel

“…Figure 14(b) demonstrates the prediction results using HE and SET test datasets that the overestimation and underestimation problems are solved with representing exhibiting a 1.5% of HER difference. It is evaluated that training efficiency can be improved due to the similar enhanced mechanism of sheared edge stretchability although shearing method and stretching direction are different [8,9].…”

“…Figure 2 demonstrates the engineering strain-stress curves of two different GPa-grade steels with XF980 and TRIP1180. TRIP1180 is one of the conventional GPa-grade steels which is widely applied in BIW panels, and XF980 is categorized as a third-generation advanced high-strength steel exhibiting a good strengthductility combination [8][9].…”

“…However, it is not proper to apply for the advanced punching method such as humped bottom punch and twostage punching which show remarkable improvements in HER compared to the flat punch, since advanced punching methods can improve the HER by dramatically reducing the hardness at the bottom of fracture zone, although the maximum hardness is similar to the conventional punch as shown in Fig. 5 [8][9].…”

“…An ANN algorithm is applied to predict the edge stretchability of GPa-grade steels for various material hardening and punching conditions with 72 experimental data set. Additional 216 data-sets of SET test, which is one of edge stretchability methods by tensioning the tensile specimen with sheared induced on one side of edge, is applied to increase the training efficiency since the SET test is also affected by the material properties and material hardening in terms of strain hardening exponent and material hardening profile [8][9]. In a single data set, 23 input features for representing three parameters as listed in Table 2, which are corresponded with the output feature of the pre-damage strain.…”

“…It is difficult to predict edge cracking using numerical simulations with conventional forming limit curves since the onset of edge cracking occurs at a significantly lower strain than that of the forming limit curves, which causes significant delays in new product development [5][6]. Moreover, edge cracking is still being reported in the third-generation advanced high-strength steels which have been recently developed to obtain superior elongation while maintaining inherent strength [7][8][9]. For these reasons, the prediction of edge cracking is an enormous challenge for the wide application of GPa-grade steels in BIW structures.…”