S. W. Cheung, Y. Choi, H. K. Springer, T. Kadeethum
{"title":"Data-scarce surrogate modeling of shock-induced pore collapse process","authors":"S. W. Cheung, Y. Choi, H. K. Springer, T. Kadeethum","doi":"10.1007/s00193-024-01177-2","DOIUrl":null,"url":null,"abstract":"<div><p>Understanding the mechanisms of shock-induced pore collapse is of great interest in various disciplines in sciences and engineering, including materials science, biological sciences, and geophysics. However, numerical modeling of the complex pore collapse processes can be costly. To this end, a strong need exists to develop surrogate models for generating economic predictions of pore collapse processes. In this work, we study the use of a data-driven reduced-order model, namely dynamic mode decomposition, and a deep generative model, namely conditional generative adversarial networks, to resemble the numerical simulations of the pore collapse process at representative training shock pressures. Since the simulations are expensive, the training data are scarce, which makes training an accurate surrogate model challenging. To overcome the difficulties posed by the complex physics phenomena, we make several crucial treatments to the plain original form of the methods to increase the capability of approximating and predicting the dynamics. In particular, physics information is used as indicators or conditional inputs to guide the prediction. In realizing these methods, the training of each dynamic mode composition model takes only around 30 s on CPU. In contrast, training a generative adversarial network model takes 8 h on GPU. Moreover, using dynamic mode decomposition, the final-time relative error is around 0.3% in the reproductive cases. We also demonstrate the predictive power of the methods at unseen testing shock pressures, where the error ranges from 1.3 to 5% in the interpolatory cases and 8 to 9% in extrapolatory cases.</p></div>","PeriodicalId":775,"journal":{"name":"Shock Waves","volume":"34 3","pages":"237 - 256"},"PeriodicalIF":1.7000,"publicationDate":"2024-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Shock Waves","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s00193-024-01177-2","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"MECHANICS","Score":null,"Total":0}
引用次数: 0
Abstract
Understanding the mechanisms of shock-induced pore collapse is of great interest in various disciplines in sciences and engineering, including materials science, biological sciences, and geophysics. However, numerical modeling of the complex pore collapse processes can be costly. To this end, a strong need exists to develop surrogate models for generating economic predictions of pore collapse processes. In this work, we study the use of a data-driven reduced-order model, namely dynamic mode decomposition, and a deep generative model, namely conditional generative adversarial networks, to resemble the numerical simulations of the pore collapse process at representative training shock pressures. Since the simulations are expensive, the training data are scarce, which makes training an accurate surrogate model challenging. To overcome the difficulties posed by the complex physics phenomena, we make several crucial treatments to the plain original form of the methods to increase the capability of approximating and predicting the dynamics. In particular, physics information is used as indicators or conditional inputs to guide the prediction. In realizing these methods, the training of each dynamic mode composition model takes only around 30 s on CPU. In contrast, training a generative adversarial network model takes 8 h on GPU. Moreover, using dynamic mode decomposition, the final-time relative error is around 0.3% in the reproductive cases. We also demonstrate the predictive power of the methods at unseen testing shock pressures, where the error ranges from 1.3 to 5% in the interpolatory cases and 8 to 9% in extrapolatory cases.
期刊介绍:
Shock Waves provides a forum for presenting and discussing new results in all fields where shock and detonation phenomena play a role. The journal addresses physicists, engineers and applied mathematicians working on theoretical, experimental or numerical issues, including diagnostics and flow visualization.
The research fields considered include, but are not limited to, aero- and gas dynamics, acoustics, physical chemistry, condensed matter and plasmas, with applications encompassing materials sciences, space sciences, geosciences, life sciences and medicine.
Of particular interest are contributions which provide insights into fundamental aspects of the techniques that are relevant to more than one specific research community.
The journal publishes scholarly research papers, invited review articles and short notes, as well as comments on papers already published in this journal. Occasionally concise meeting reports of interest to the Shock Waves community are published.