Industry experts Steven Sheng of General Motors, and Xinran Xiao, a professor at the Michigan State University, discuss their research into the crash simulation of lightweight automotive materials
Steven Sheng, formability engineer, General Motors
Please tell us about your research into the prediction of fracture in warm forming magnesium alloys.
I have developed a method for representing the forming limit for warm forming magnesium alloy sheet material. In this method, the strain rate and temperature effect are presented concisely by a single parameter of the Zener-Hollomon parameter. Thus, the forming limit can be represented as a forming limit surface.
Traditionally, forming limit curves (FLCs) are used to identify fracture/necking failure in sheet metal forming. At elevated temperatures, the magnesium alloys exhibit strong rate sensitivity and thus many FLCs are needed to represent different forming conditions. Those FLCs are not only costly to obtain but also inconvenient for use in FEM simulation.
As only one surface is needed, this method could reduce the efforts of laboratory testing and provide a convenient tool in the development of thermal forming processes. The usefulness of the forming limit surface is demonstrated by predicting fracture in warm forming magnesium alloy AZ31B in an isothermal conical cup test and non-isothermal round cup draw test.
What are the biggest hurdles to overcome in implementing CAE and FEA (finite element analysis) in the safety testing of lightweight materials?
Improving the accuracy of simulation results, which are affected by many factors such as material modeling, mesh quality, algorithm accuracy etc. For product development, a multi-objective optimization solver, which can be adapted with different CAE solvers, is also needed.
Xinran Xiao, professor of mechanical engineering, Michigan State University
Please tell us about your work in modeling of thin-walled composites for crash simulation.
Thin-walled tubes, used in the front rails of a vehicle chassis, are the primary energy absorbing structures in vehicles. In a frontal impact, the front rails are subjected to axial crash load, and axial crash of tubes is the benchmark problem to gauge the capability of crash simulations. The axial crash simulation of composite tubes is very challenging. These can absorb two or more times the energy of steel rails of the same weight. The superior energy absorption of composites is attributed to the extensive damage and failure of the materials. To simulate the behavior of composites with damage and failure, however, is difficult. Normally, the constitutive models are valid to the point where the material reaches its peak load carrying ability. They are not adequate to describe the behavior of materials with significant damage.
Previously, to match the simulation with experimental results, analysts had to adjust a number of parameters. The problem is that a set of parameters that are suitable for one geometry may not work for others. After having tested these material models myself, I realized that the lack of consideration of irreversible strains in a continuous damage mechanics-based model is a major problem in axial crash simulation. On the other hand, with the exception of a few, most vehicle structures will not be subjected to axial loading. The inability to predict an axial crash should not prevent the use of composites in these applications.
In a presentation at the Modeling, Simulation and Crash Safety Congress 2016, I will discuss the requirements for composite material models for different types of vehicle structures. I will also present the key developments in axial crash simulation of composite tubes. We are working on coupled damage-plasticity models for composites. The improved material models together with a new shell-beam method for thin-walled structures have resulted in promising improvement in axial crash simulations of composite tubes.
What are the most important hurdles to overcome in implementing CAE and FEA in the safety testing of lightweight materials?
With high fidelity FEA simulations, different design options can be evaluated through virtual testing before they are built. Virtual design and validation is critical to car makers. If the simulation technology is not ready for a material, extra tests will have to be conducted. This is a hurdle for the use of composites in crash critical components.
For crash simulations, we still have some gaps for metallic materials. As mentioned, we need to differentiate the types of vehicle structures. In many cases, we can bridge the gaps with the existing technology. The constitutive models for crash simulations inevitably have many parameters. Some of them are difficult to measure. If we have experimental results covering enough load cases, parameter optimization may help us to get the set of parameters which provides the best correlation for the load cases tested. I did some work several years ago and it is effective. On the other hand, one must know that these values are not necessarily valid for a new load case.
What are some of the implications of improved material models and more accurate safety simulations to the manufacture of lightweight vehicles?
FEA simulation has changed the landscape of automotive design. The car manufacturer will not turn back the wheel to a test-intensive design procedure as it were before computers. To increase the use of composites in crash critical structures, we have to be able to predict the crashworthiness of the structure as we do for metal parts. Good material models and robust and accurate safety simulations are critical to vehicle lightweighting. The technology for the crash simulation of composite structures has improved a lot. Ideally, these new developments need to be examined carefully in a coordinated effort. Individual teams can only carry the research to a certain extent.
Both Steven Sheng and Xinran Xiao will be presenting their research at the Modeling, Simulation and Crash Safety Congress 2016, which takes places January 26-27, 2016, and will be held at the Grand Ballroom, Detroit Marriott Troy Hotel, Detroit, Michigan, USA.
