This project brings together analytical, computational and testing capabilities of three universities for the development, validation, and software implementation of a multi-scale method for thermo-mechanical response of open-cell Aluminum foams, from linear behavior to complete crushing. The University of California Los Angeles team will develop a suite of high-fidelity simulation modules that will yield the thermo-mechanical response of open-cell Aluminum foams at the micro-scale. The Purdue University team will build the essential multiscale framework with which a complete structure (comprising not only foams) can be efficiently analyzed. The Illinois Institute of Technology team will conduct the experiments that are needed to calibrate and validate the proposed modeling approaches. These experiments will feature custom instruments and unique measurements that will enable the probing of thermo-mechanical behavior of open-cell foams with requisite detail.
The over-arching goal of this research project is to facilitate widespread use of Aluminum foams as load-bearing structural members, by providing a validated simulation tool to design engineers. The potential societal benefit of this enabling technology is to help reduce energy consumption in a multitude of transportation systems that range from automobiles and railroad cars to elevators. The project will also feature educational, outreach and training activities at participating institutions that are integrated with the proposed research program. These activities will engage and educate graduate and undergraduate students in the growing area of multiscale scale mechanics, and create awareness and stimulate interest in STEM (Science, Technology, Engineering and Mathematics) disciplines among young high school students, and incoming freshmen.
This project has been conducted in collaboration with three participating institutions; UCLA, Purdue and Illinois Institute of Technology (IIT) with the major objective of developing a new multiscale framework capable of accurately analyzing the crushing behavior of open cell aluminum foams. These solid foams, which are essentially a 3D network of bone-shaped ligaments, are lightweight materials with promising applications in mitigating impact/blast energy such as in car bumpers and various protective components. The objective of IIT team is to conduct a variety of experiments on foam ligaments, parent material and aluminum foam that will also be used to validate and calibrate the micro-mechanical models, and to provide feedback on the results obtained from simulations conducted by UCLA and Purdue teams. Experimental research undertaken by IIT team is driven by the challenges related to the accurate description of ligament morphology, ligaments’ mechanical behavior, and material response under severe loading conditions characterized by large strains, high strain rates and thermoplastic coupling. In this perspective, IIT team has designed and conducted a series of experiments on both the foam ligaments and the bulk aluminum alloy from which the foam is made. Morphological characterization studies conducted at IIT on aluminum foam ligaments have provided a unique database from which mathematical models have been developed to describe the variation of ligament geometry along its length. This is an important first step in constructing realistic foam models starting from the ligament scale. This model is ready to be used in any foam model that attempts to capture the geometry of ligaments in a more realistic way. Microstructural investigation of ligament cross sections has also revealed that, unlike bulk aluminum parent material, ligaments may have casting defects such as pores and inclusions that weakens and causes a large margin of variation in mechanical properties, which must be factored in to improve the fidelity of modeling work. These experimental results have been shared with UCLA team to guide their geometric ligament models. Cantilever bending and tension experiments have been designed and conducted to analyze the material properties of ligaments and to compare it to that of the bulk aluminum alloy. Geometric characterization of the ligaments mentioned above has provided the necessary mathematical tool in analyzing the results of ligament-scale experiments and extract the relevant material properties of ligaments from load-displacement curves. Results showed that average ligament properties are close to that of bulk material, and constitutive models developed from bulk material testing can be used for ligaments. However, casting defects and limited number of grains across ligament cross section cause a significant variation in the mechanical properties of ligaments, which must be factored in both the modeling work and the evaluation of results. Another important outcome of IIT research is the construction of a rate and temperature dependent constitutive model, for the first time, for the parent material (Al 6101-T6 alloy). To this end, a large number of experiments have been conducted at both quasi-static and dynamic loading rates in a wide range of temperatures. Split Hopkinson pressure bar (SHPB) technique has been used to achieve impact loading conditions at both room temperature and elevated temperatures using a custom made heating chamber. Based on the experimental data a modified Johnson-Cook (MJC) material model has been constructed to predict the flow stress of parent material in a wide range of loading conditions in a seamless manner. The classical JC model is one of the most common constitutive models used in commercial FE codes, but it doesn’t always lend itself to accurate representation of rate?dependent thermoplastic behavior of materials. This is mainly because of the decoupled nature of strain hardening, rate sensitivity and thermal softening terms in the model. Therefore, we have modified JC model by coupling strain and strain rate hardening response with thermal effects to capture experimentally observed constitutive behavior of 6101-T6 alloy. Upon the completion of model development and determination of model parameters, our findings have been shared with UCLA team to be embedded in stress-resultant plasticity based model at the ligament scale that they have developed. This constitutive model is also ready to be used in any foam modeling effort where strain rate effects and thermoplastic heating (and resulting thermal softening) effects due to severe localized strains cannot be ignored, such as in dynamically loaded foam crushing events commonly encountered in impact energy mitigation applications.
