|About this Abstract
||7th World Congress on Integrated Computational Materials Engineering (ICME 2023)
||Designing Fatigue Resistance of Metallic Alloys with a Hybrid of Deep Learning and Micromechanics
||Anssi Laukkanen, matti lindroos, tom andersson, napat vajragupta, tatu pinomaa, sicong ren, abhishek biswas, tomi suhonen
|On-Site Speaker (Planned)
Fatigue remains a critical failure mechanism both of industrial and scientific interest. Fatigue testing is commonly a costly and time consuming exercise, which makes it difficult to establish microstructure to fatigue performance relationships. This is seen as an area where ICME driven "virtual fatigue testing" hybrid workflows consisting of physics- and data-driven modeling elements can support. We present a full field micromechanical approach to capture the effects of defects to fatigue performance of metallic alloys and steels. We utilize a high-throughput framework to derive a recurrent deep learning based surrogate to evaluate and act as a design tool for microstructural features improving resistance to fatigue. We demonstrate the approach on high strength steels and high entropy alloys using Bayesian workflows.