Very nice video TheAspear! I am trying to simulate fracturing myself and have a quick question on getting started: How are you modeling the mesh? I've tried looking at software packages like netgen/gmsh, but I'm not sure how to build a data structure for the mesh file that they output. Do I just store a series of 4-point tetrahedron and somehow find which ones are neighbors? I would appreciate any direction!
Hi @skyflashings -- I am not familiar with netgen/gmsh. These simulations were done using a combination of ABAQUS, FRANC3D (a crack growth prediction and remeshing code), and python scripts that I wrote to sort of interface between the two. I have uploaded the python scripts, support modules, a tutorial, and some example input files (to give you an idea of the file format that ABAQUS uses). The link to my website is in the video description above. Good luck. :)
More detailed information (including computer codes/scripts) regarding the method we used to perform the 3D elastic-plastic tearing simulations can be found in the following document located on NASA's technical report server: "Surrogate Modeling of High-Fidelity Fracture Simulations for Real-Time Residual Strength Predictions", January 2011, Spear et al.
In this video, I'm only showing the evolving mesh, no contours to represent fields like stress or strain. In some of my other videos I actually have the contour turned "on" to show these fields.
@lordpoee I just uploaded the python scripts, support modules, and a tutorial on my website for doing these kinds of simulations (cfg.cornell.edu/~ads259/index_files/RecentResearch.htm). You need access to ABAQUS and FRANC3D (and a lot of patience :) to actually run the scripts, but you can always read through the python scripts and get a general sense of the computational work flow involved in using adaptive remeshing and material state mapping for 3D crack growth simulations.
@lordpoee Thanks for your comment! The circular region represents a particular template of finite elements surrounding the crack front. While the finite element mesh, itself, is artificial, it is used to calculate the stress fields in the structure, which are real. The region of material inside the circular template is where stress concentrates due to the discontinuity of the crack. So, if I understand your question correctly, then the answer is yes, but in slightly different terminology. :)
@TheAspear It would also be interesting to see a similar simulation using differentially hardened carbon steel as the source. Are you able to simulate different materials?
Very nice video TheAspear! I am trying to simulate fracturing myself and have a quick question on getting started: How are you modeling the mesh? I've tried looking at software packages like netgen/gmsh, but I'm not sure how to build a data structure for the mesh file that they output. Do I just store a series of 4-point tetrahedron and somehow find which ones are neighbors? I would appreciate any direction!
skyflashings 4 months ago
Hi @skyflashings -- I am not familiar with netgen/gmsh. These simulations were done using a combination of ABAQUS, FRANC3D (a crack growth prediction and remeshing code), and python scripts that I wrote to sort of interface between the two. I have uploaded the python scripts, support modules, a tutorial, and some example input files (to give you an idea of the file format that ABAQUS uses). The link to my website is in the video description above. Good luck. :)
TheAspear 4 months ago
More detailed information (including computer codes/scripts) regarding the method we used to perform the 3D elastic-plastic tearing simulations can be found in the following document located on NASA's technical report server: "Surrogate Modeling of High-Fidelity Fracture Simulations for Real-Time Residual Strength Predictions", January 2011, Spear et al.
TheAspear 1 year ago
In this video, I'm only showing the evolving mesh, no contours to represent fields like stress or strain. In some of my other videos I actually have the contour turned "on" to show these fields.
TheAspear 1 year ago
@TheAspear Could you post your input data and stress formulas?
lordpoee 1 year ago
@lordpoee Like say, are you using the Holloman equation?
lordpoee 1 year ago
@lordpoee I just uploaded the python scripts, support modules, and a tutorial on my website for doing these kinds of simulations (cfg.cornell.edu/~ads259/index_files/RecentResearch.htm). You need access to ABAQUS and FRANC3D (and a lot of patience :) to actually run the scripts, but you can always read through the python scripts and get a general sense of the computational work flow involved in using adaptive remeshing and material state mapping for 3D crack growth simulations.
TheAspear 4 months ago
@TheAspear Cool, I'll check it out now :)
lordpoee 4 months ago
@TheAspear hmmm (cfg.cornell.edu) does not appear to be valid...
lordpoee 4 months ago
@lordpoee Ok, I just put the link to my website in the video description above.
TheAspear 4 months ago
@TheAspear Okay cool, thanks again!
lordpoee 4 months ago
I take it the circular shape toward the edge of the tear symbolizes the focal point of the pressure?
lordpoee 1 year ago
@lordpoee Thanks for your comment! The circular region represents a particular template of finite elements surrounding the crack front. While the finite element mesh, itself, is artificial, it is used to calculate the stress fields in the structure, which are real. The region of material inside the circular template is where stress concentrates due to the discontinuity of the crack. So, if I understand your question correctly, then the answer is yes, but in slightly different terminology. :)
TheAspear 1 year ago
@TheAspear It would also be interesting to see a similar simulation using differentially hardened carbon steel as the source. Are you able to simulate different materials?
lordpoee 1 year ago