Predicting transient response using data-driven models for ball-impact simulations
This study investigates the application of machine learning (ML) models for predicting transient responses in ball-impact elastodynamics simulations. We focus on the canonical problem of ball impact on laminated structures, which captures essential physics while maintaining computational tractabilit...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2024-01-01
|
| Series: | Journal of Physics Communications |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/2399-6528/ad9f1f |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841551097947226112 |
|---|---|
| author | Ross Pivovar Fei Chen Raghunath Katragadda Vidyasagar Ananthan |
| author_facet | Ross Pivovar Fei Chen Raghunath Katragadda Vidyasagar Ananthan |
| author_sort | Ross Pivovar |
| collection | DOAJ |
| description | This study investigates the application of machine learning (ML) models for predicting transient responses in ball-impact elastodynamics simulations. We focus on the canonical problem of ball impact on laminated structures, which captures essential physics while maintaining computational tractability. Novel contributions include: (1) development of a temporal multi-resolution strategy for stable long-time predictions, (2) systematic comparison of U-Nets and Fourier Neural Operators as spatial ML kernels, and (3) demonstration of accurate non-local metric predictions across full time-horizons. Using a synthetic dataset of 6500 impact scenarios, we achieve 3.5–8% prediction accuracy while providing 10,000x speedup compared to traditional FEM simulations. The proposed methodology enables rapid virtual prototyping for impact-resistant design optimization. |
| format | Article |
| id | doaj-art-cd7a36b81a7346d187490b0dc8ecd846 |
| institution | Kabale University |
| issn | 2399-6528 |
| language | English |
| publishDate | 2024-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Journal of Physics Communications |
| spelling | doaj-art-cd7a36b81a7346d187490b0dc8ecd8462025-01-09T16:36:07ZengIOP PublishingJournal of Physics Communications2399-65282024-01-0181212500510.1088/2399-6528/ad9f1fPredicting transient response using data-driven models for ball-impact simulationsRoss Pivovar0https://orcid.org/0000-0002-6379-1589Fei Chen1https://orcid.org/0009-0009-1378-7643Raghunath Katragadda2Vidyasagar Ananthan3https://orcid.org/0000-0003-0262-5429Advanced Computing, AWS, 12 W 39th St, New York, NY, 10018, United States of AmericaAdvanced Computing, AWS, 3075 Olcott St, Santa Clara, CA 95054, United States of AmericaProduct Integrity, Lab126, 1100 Enterprise Way, Sunnyvale, CA 94089, United States of AmericaAdvanced Computing, AWS, 2205 7th Ave, Seattle, WA 98121, United States of AmericaThis study investigates the application of machine learning (ML) models for predicting transient responses in ball-impact elastodynamics simulations. We focus on the canonical problem of ball impact on laminated structures, which captures essential physics while maintaining computational tractability. Novel contributions include: (1) development of a temporal multi-resolution strategy for stable long-time predictions, (2) systematic comparison of U-Nets and Fourier Neural Operators as spatial ML kernels, and (3) demonstration of accurate non-local metric predictions across full time-horizons. Using a synthetic dataset of 6500 impact scenarios, we achieve 3.5–8% prediction accuracy while providing 10,000x speedup compared to traditional FEM simulations. The proposed methodology enables rapid virtual prototyping for impact-resistant design optimization.https://doi.org/10.1088/2399-6528/ad9f1ftransientelastodynamicsFNOCNNsdata-driven modelsML |
| spellingShingle | Ross Pivovar Fei Chen Raghunath Katragadda Vidyasagar Ananthan Predicting transient response using data-driven models for ball-impact simulations Journal of Physics Communications transient elastodynamics FNO CNNs data-driven models ML |
| title | Predicting transient response using data-driven models for ball-impact simulations |
| title_full | Predicting transient response using data-driven models for ball-impact simulations |
| title_fullStr | Predicting transient response using data-driven models for ball-impact simulations |
| title_full_unstemmed | Predicting transient response using data-driven models for ball-impact simulations |
| title_short | Predicting transient response using data-driven models for ball-impact simulations |
| title_sort | predicting transient response using data driven models for ball impact simulations |
| topic | transient elastodynamics FNO CNNs data-driven models ML |
| url | https://doi.org/10.1088/2399-6528/ad9f1f |
| work_keys_str_mv | AT rosspivovar predictingtransientresponseusingdatadrivenmodelsforballimpactsimulations AT feichen predictingtransientresponseusingdatadrivenmodelsforballimpactsimulations AT raghunathkatragadda predictingtransientresponseusingdatadrivenmodelsforballimpactsimulations AT vidyasagarananthan predictingtransientresponseusingdatadrivenmodelsforballimpactsimulations |