Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
My model seems too sensitive to the mesh
Posted 7 ago 2017, 15:05 GMT-4 Heat Transfer & Phase Change, Geometry, Mesh, Studies & Solvers Version 4.4 1 Reply
Please login with a confirmed email address before reporting spam
On one face I have a heat input, and on another I have convective heat transfer with various heat transfer coefficient distributions. The distributions are dependent on things such as heat flux, so it is a highly non-linear problem. I have meshed one end using quad elements and swept the mesh along the length of the geometry. These quad elements are further split into triangles on the face with convective heat transfer.
The issue that has me puzzled at the moment is the sensitivity of the model to the mesh size. If I split the geometry into 20 divisions lengthwise, for example, it will solve. If I refine the mesh to 40 divisions, it won't solve. If I further refine to 50, it will solve. 70 might not solve, but 75 might. Due to this irregularity I have a very difficult time getting a solution with an acceptable mesh. I don't really understand this. Does anyone have any ideas?
Thank you,
Bryan
The issue that has me puzzled at the moment is the sensitivity of the model to the mesh size. If I split the geometry into 20 divisions lengthwise, for example, it will solve. If I refine the mesh to 40 divisions, it won't solve. If I further refine to 50, it will solve. 70 might not solve, but 75 might. Due to this irregularity I have a very difficult time getting a solution with an acceptable mesh. I don't really understand this. Does anyone have any ideas?
Thank you,
Bryan
1 Reply Last Post 9 ago 2017, 13:24 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Just some guesses. Check your mesh quality. (Right-click mesh, statistics.) Some solvers and arrangements only work well with higher quality meshes. Also, consider changing the discretization (you may have to enable viewing that under preferences). Also, consider using a different solver.