Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability

Saved in:
Bibliographic Details
Published in:Геофизический журнал
Date:2010
Main Authors: Galvan, B., Miller, S.
Format: Article
Language:English
Published: Інститут геофізики ім. С.I. Субботіна НАН України 2010
Online Access:https://nasplib.isofts.kiev.ua/handle/123456789/101315
Tags: Add Tag
No Tags, Be the first to tag this record!
Journal Title:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Cite this:Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability / B. Galvan, S. Miller // Геофизический журнал. — 2010. — Т. 32, № 4. — С. 49-50. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1859717050704855040
author Galvan, B.
Miller, S.
author_facet Galvan, B.
Miller, S.
citation_txt Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability / B. Galvan, S. Miller // Геофизический журнал. — 2010. — Т. 32, № 4. — С. 49-50. — англ.
collection DSpace DC
container_title Геофизический журнал
first_indexed 2025-12-01T08:12:57Z
format Article
fulltext ��������� ��� ���������������������� �� ��� ��!"#�$%&'�("�%()�#*+#�' #(&"�&��&,#��-�%()� �)#..'(/ Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability B. Galvan, S. Miller, 2010 Bonn University, Bonn, Germany Fluid flow in the earth is controlled by fracture networks that evolve in response to far field stress, local stress perturbations, and the pressure state of the fluid within them. The se-processes are very important for many geophysical systems, including earthquakes and volcanoes. Modelling the underly- ing physics is challenging because the time scales involved, from the elastic wave speed of crack growth to pressure diffusion and flow, make the se-prob- lems numerically cumbersome. Our approach to model the se-processes is to couple the elastic- plastic response of the solid porous matrix to a pres- sure dependent (nonlinear) diffusion model for the Fig. 1. CPU-GPU Poro-elasticity: 800 400points, grid size of 25 m. One day of simulation. Computation time: 42 min. Fig. 2. Poro-elasto-plastic CPU (MATLAB): 150 100 points. Grid size is133 min X and 90 min Y. One day of simulation. Computation time: about 7 hours. fluid flow. Changes in the fluid pressure introduce changes in stresses in the porous media, which may lead to either hydro-facture or shear fracture 0� ��������� ��� ���������������������� /#�)-(% '1%.�+,#(� #(%2 within the solid. See page forces, forces related by gradients of pore pressure, can also promote delo- calized crack formation. Many models of fracture propagation have been developed using finite ele- ments or other numerical methods in order to over- come the high deformation of the grid, however costly re-meshing algorithms are necessary to accurate model the evolving crack. The complete model, non- linear diffusion and poro-elasto-plasticity, is very computationally expensive. GPU technology allows high resolution modelling and easy implementation of explicit finite difference methods in an efficient way. We have taken advantage of many-cores GPU technology together with CPU and developed a high- resolution fully explicit finite difference model of nonlinear diffusion coupled with the mechanical re- sponse of poro-elasto-plastic medium (Fig. 1). In our algorithm, we can model both the propagation of previously de?ned fractures and fracture genera- tion and growth in response to the evolving stress field. Our model includes shear and tensile cracking, which plays a dominant role in the hydraulic proper- ties of the poro-elastic media as well as changes in the rheological properties. High resolution 2D simu- lations are presented showing the hydro-mechani- cal evolution of systems driven by high pressure sources at depth, such as some aftershock se- quences and with application to volcano-mechanics. Using CPU-GPU approach numerical resolution can be increased to more than three times and compu- tational time is decreased as much as at enth-com- pared with the CPU alone approach (Fig. 2).
id nasplib_isofts_kiev_ua-123456789-101315
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 0203-3100
language English
last_indexed 2025-12-01T08:12:57Z
publishDate 2010
publisher Інститут геофізики ім. С.I. Субботіна НАН України
record_format dspace
spelling Galvan, B.
Miller, S.
2016-06-02T14:24:23Z
2016-06-02T14:24:23Z
2010
Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability / B. Galvan, S. Miller // Геофизический журнал. — 2010. — Т. 32, № 4. — С. 49-50. — англ.
0203-3100
https://nasplib.isofts.kiev.ua/handle/123456789/101315
en
Інститут геофізики ім. С.I. Субботіна НАН України
Геофизический журнал
Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
Article
published earlier
spellingShingle Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
Galvan, B.
Miller, S.
title Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
title_full Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
title_fullStr Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
title_full_unstemmed Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
title_short Ancient CPU-GPU simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
title_sort ancient cpu-gpu simulation of evolving fracture networks in a poro-elasto-plastic medium with pressure-dependent permeability
url https://nasplib.isofts.kiev.ua/handle/123456789/101315
work_keys_str_mv AT galvanb ancientcpugpusimulationofevolvingfracturenetworksinaporoelastoplasticmediumwithpressuredependentpermeability
AT millers ancientcpugpusimulationofevolvingfracturenetworksinaporoelastoplasticmediumwithpressuredependentpermeability