Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.14028 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912335122137088 |
|---|---|
| author | Maor, Ron Hansen, Lars Jerolmack, Douglas Goldsby, David |
| author_facet | Maor, Ron Hansen, Lars Jerolmack, Douglas Goldsby, David |
| contents | Rock and ice are ubiquitous geologic materials. While apparently solid, they also exhibit fluid behavior under stress - a property termed viscoelasticity. Viscoelastic convection of Earth's mantle drives tectonic plate motion with consequences for earthquakes and sea-level rise, while viscoelastic deformation of ice controls glacier flow and the flexure of icy moons. For crystalline materials, "flow laws" describing bulk rheology can be derived from understanding microstructural dynamics such as crystal-defect migration. Common geologic materials like ice and olivine have grain sizes and crystal orientations that evolve with strain; this complexity precludes a first principles approach. Here we use a Bayesian inference method to learn the connection between microstructure and flow in ice and olivine, from fits to experimental data of these materials undergoing steady-state deformation and forced oscillations. We demonstrate that this method can constrain a nonlinear viscoelastic model for each material, that is capable of capturing both steady and transient dynamics and can also predict dynamics for data it was not trained on. Our results may improve geodynamic models that rely on parameterized constitutive equations, while our approach will be useful for experimental design and hypothesis testing. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_14028 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning the nature of viscoelasticity in geologic materials with MCMC Maor, Ron Hansen, Lars Jerolmack, Douglas Goldsby, David Geophysics Computational Physics Data Analysis, Statistics and Probability Rock and ice are ubiquitous geologic materials. While apparently solid, they also exhibit fluid behavior under stress - a property termed viscoelasticity. Viscoelastic convection of Earth's mantle drives tectonic plate motion with consequences for earthquakes and sea-level rise, while viscoelastic deformation of ice controls glacier flow and the flexure of icy moons. For crystalline materials, "flow laws" describing bulk rheology can be derived from understanding microstructural dynamics such as crystal-defect migration. Common geologic materials like ice and olivine have grain sizes and crystal orientations that evolve with strain; this complexity precludes a first principles approach. Here we use a Bayesian inference method to learn the connection between microstructure and flow in ice and olivine, from fits to experimental data of these materials undergoing steady-state deformation and forced oscillations. We demonstrate that this method can constrain a nonlinear viscoelastic model for each material, that is capable of capturing both steady and transient dynamics and can also predict dynamics for data it was not trained on. Our results may improve geodynamic models that rely on parameterized constitutive equations, while our approach will be useful for experimental design and hypothesis testing. |
| title | Learning the nature of viscoelasticity in geologic materials with MCMC |
| topic | Geophysics Computational Physics Data Analysis, Statistics and Probability |
| url | https://arxiv.org/abs/2504.14028 |