LAUSR

The micro-sandwich sheets with metallic/polymer fibrous core have been pointed out as one of the most promising technological solutions to the automotive industry. However, due to the lack of understanding of certain fundamentals related with the mechanical behavior of micro-sandwich sheets during forming processes, the transfer and scale-up of this promising technology to industry has been limited. A challenging aspect in the experimental characterization of these materials is related with the unknow properties of the composition of the core which consist of metallic (polymer) fibers and adhesive. In general, the suppliers of monolithic metal sheets also make available the respective datasheets with the mechanical and chemical properties. These datasheets use to refer the minimum, maximum or a specific tolerance range (dependent of the grade material) to the mechanical properties. However, the micro-sandwich sheets are not provided with this mechanical and chemical data, at least, for all layers. Therefore, it is missing a simple and robust methodology to supply the mechanical properties of the total micro-sandwich sheet to the industry. Furthermore, there is no study about the different numerical approaches available in the commercial stamping softwares to modelling and simulate micro-sandwich materials. In this work, a strategy to deduce the unknown mechanical properties of the fibrous core from symmetric or asymmetric micro-sandwich sheets, i.e., with the same or different skin thickness, is presented. For true stress-strain curves and anisotropy purposes, uniaxial tensile tests in 3 different directions, according ISO 6892-1:2009 standard, were performed. Total micro-sandwich specimens and skin specimens were tested. The mechanical properties of the core were deduced from micro-sandwich and skin’s mechanical properties. Based on this data, the constitutive model was established. Additionally, 6 different Nakazima geometries were punched according to ISO 12004 for formability assessment. Experimental Forming Limit Curves (FLC) and principal strains data were recorded during the tests thanks to a high-resolution cameras and software GOM ARAMIS. Finally, the experimental mechanical tests were virtually reproduced with both commercial codes AutoForm R5.2 and PAM-STAMP 2G 2015.1. Thus, different modelling strategies and Finite Element Method (FEM) approaches were compared. The excellent agreement between numerical and experimental results demonstrate the accuracy of the applied methodology.