CE Seminar by Pınar Acar

Time: 10:00
Location: ENG 208

Speaker: Pınar Acar
Title: Multi-Scale Materials Design: Recent Challenges and Future Vision
Date:  26 April, 2017
Time: 10:00-11:00
Place: ENG 208
The area of multi-scale design and optimization of materials has been garnering a considerable interest due to the increasing needs for high performance materials in electronics, energy and structural applications, and extreme environments. The research on these materials and their manufacturing routes will potentially extend in future to higher order representations for modeling and multi-scale design of composites, ceramics and other materials for extreme environments such as hypersonics and aerothermoelasticity applications, fabrication of adaptive thermal response materials, energetic composites in fuel cells, thermal energy harvesting in satellites and materials for green energy applications by maintaining the concept of 3D design and optimization under uncertainty.
The multi-scale modeling starts from the microstructural level with quantification of the structure using probabilistic descriptors. The computational scheme links the micro-macro scales through volume-averaged relations to calculate the engineering material properties. These properties are sensitive to microstructural variations, which are introduced by the experimental imperfections. The quantification of uncertainties is a statistical process, and requires large-scale test data, which might be absent or intentionally avoided in some cases due to high operational costs of measurements. In such cases it is more practical to replace them with small-scale measurements. To predict the spatial and temporal evolutions of a microstructure using small-scale data, we developed a stochastic reconstruction framework based on Markov Random Field (MRF) approach. We implemented an analytical UQ algorithm to model the uncertainty propagation to the microstructure design and material properties, instead of the traditional computational algorithms. Incorporating this approach in an automated optimization framework, we significantly reduced the computation time, effectively captured the outcomes of stochasticity in microstructure design and simplified achieving multiple material designs with optimized material properties. We also derived a reduced order-modeling (ROM) scheme to represent the texture evolution in different deformation processes. With this approach we not only identified the optimum microstructure design but also the optimal processing route to manufacture a material with optimized properties and/or texture.
Pinar Acar is a PhD candidate in Aerospace Engineering Department of University of Michigan, where she works as a graduate student research assistant. She received her BSc and MSc degrees in Astronautical and Aeronautical Engineering from Istanbul Technical University, where she also worked as a teaching assistant for 3 years. During her time in Istanbul Technical University, she participated in numerous tasks including NASA and NATO projects. Furthermore, she led a team in designing the first satellite de-orbiting system in Turkey. Pinar joined the University of Michigan in 2013 to earn her PhD, and since then she has been working on multi-scale materials design, design under uncertainty, microstructures, optimization, uncertainty quantification and reduced order modeling. During the course of her PhD, she developed various computational methods and optimization algorithms for studying the behavior of multi-scale models under uncertainties. Pinar is the author and co-author of more than 30 journal and conference papers as well as 2 book chapters for NATO Science and Technology Organization.
Among her many awards, Pinar is the recipient of the prominent international Amelia Earhart Fellowship, which is awarded annually to only a few talented women around the world in aerospace sciences, the University of Michigan Rackham PhD Fellowship and the Scientific and Technologic Research Council of Turkey’s Graduate Student Fellowship.