Issue with Surface Separation in Gmsh
I am writing to report a bug I encountered while using Gmsh for my modeling needs. Additionally, I would greatly appreciate your assistance in resolving this issue.
To provide context, I am currently working on creating a simple model of a borehole surrounded by rocks, where I aim to simulate the behavior of electrical potentials within this geological structure. Specifically, I am attempting to place five electrodes inside the borehole to apply a potential difference between the injection electrode and the edges of my model (i.e., the edges of the rocks). Subsequently, I intend to measure the potentials in the other electrodes.
In my endeavor, I have utilized Gmsh to generate the geometry for my model. To create the borehole, I have employed two circles nested within each other, with the intent of separating their surfaces and extruding them to form the cylinder representing the borehole. However, I have encountered a limitation within Gmsh that impedes my progress.
The issue arises when attempting to set a significant difference in radius between the inner and outer circles, specifically when the radius of the surrounding rocks exceeds approximately 99 times the radius of the borehole/electrode. Beyond this threshold, Gmsh fails to accurately separate the surfaces, resulting in interference where the surface of the larger circle encompasses that of the smaller one. This limitation severely restricts my ability to create the desired geometry for my simulation, as I require a much greater disparity in radius between the borehole and the surrounding rocks for numerical reasons (e.g., on the order of 1000 times).
Attached, I have uploaded my .geo file (simple-2D-borehole-rock-model.geo), illustrating the geometry setup and the parameter values used. When R_rock is set to a value exceeding approximately 99 times R_borehole, the interference issue occurs, compromising the integrity of the model.
I have attempted alternative approaches, such as utilizing cylinders in the OpenCASCADE environment. However, I have encountered challenges in extruding them and integrating the electrodes within the borehole section, rendering this solution ineffective. I have uploaded another .geo file (Final_model.geo) to show you how I expect my model to look. Please check it for reference.
Given the importance of this modeling task and the limitation encountered, I kindly request your assistance in addressing this bug. Any guidance, workaround suggestions, or potential updates to Gmsh that could resolve this issue would be immensely helpful and greatly appreciated.
Thank you for your attention to this matter