Material modeling in Sheet Metal Forming Simulations : Quality comparison between comonly used material models

Abstract: In today's automotive industries, many different simulation programs are used to optimize parts before they come into production. This has created a market for complex material models to get the best possible approximation of reality in the simulation environment. Several industries are still using older material models that can’t give an acceptable accuracy for the materials currently in use as they are based on much simpler and older materials. The problem with material models is that there is no direct comparison between the material models which leads to several sheet metal forming companies still holding on to older models like Hill`48.   The purpose of this work is to create a comparison of sheet material models from a user perspective to be able to provide recommendations of material models. Different models will be tested for different materials and will be based on AutoForm's recommendations. AutoForm is a FEM based sheet metal forming simulation program used by large names in the automotive industry. These recommendations are Vegter2017, BBC2005 or Hill`48 for steel and Vegter2017, BBC2005 or Barlat`89 for aluminum.   This work is achieved by comparing experimental data from a Limiting Dome Height (LDH) test with a simulation of this test for all material models and then comparing the results. The data that will be compared consists of the major and minor strain in the sheet as well as the punch force. These parameters are chosen as they give an overview of the model’s applicability as well as accuracy. The test will be performed on all materials available in Volvo Cars material library to create a broader overview of all material models. The material models will also be evaluated depending on their user-friendliness by analyzing what types of data are required to perform a simulation.   The result from these tests showed that BBC 2005 should be recommended for aluminum and steel for companies that have access to biaxial data and for people who put optimization in focus. Hill`48 proved far too deviant in the results for steel and should not be used if other models are available. Vegter 2017 proved perfect for steel simulations as the result were great as well as the necessary material data can be obtained through standardized tensile tests. The result also showed that Vegter2017 should not be used for aluminum since the result was too deviant from the experimental data in aspect for both form approximation and strain magnitude. Barlat`89 gave accurate results with only data from a tensile test which makes it a preferred model when working with aluminum.   The conclusion from this work is that the choice of material model is very dependent on what conditions you have as very few industries have access to the tests required by the BBC 2005 model. Another conclusion may be drawn for Barlat`89 with aluminum and Vegter 2017 with steel as they can be preferred when working with a small timeframe as well as when few test data is available.