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.
Recalculate without re-meshing?
Posted Feb 26, 2024, 10:44 p.m. EST Mesh, Studies & Solvers, Physics Interfaces Version 6.2 2 Replies
Please login with a confirmed email address before reporting spam
Hi all,
I'm sure there's a way to do this, and I've been searching, but cannot figure out the right keywords to figure it out.
I have a model for which I'm sweeping a few different geometric parameters, which makes the simulation quite long. I need to run this same set of sweeps twice, one with an applied voltage at one location, and the second with it at a different location.
Currently, I have this set up with a single component geometry and two separate physics setus. The physics setups are identical other than the placement of the applied voltage. Then I have three studies defined: a parametric sweep for each physics, and then a third study that just runs the other two sequentially.
In this setup, the same set of meshes are generated twice, once for each of the parametric sweeps. But this seems quite silly to do, since the meshes are the same. Is there a way to either save the meshes and re-use them in the second sweep, or just "update" the first sweep after it's done with the re-applied voltage, or maybe something else to cut down on the time? There must be a way to not need to recalculate all the meshes again (hopefully).
Thanks.
Attachments:
Please login with a confirmed email address before reporting spam
Update: the issue was a misunderstanding of "Study" and "Step". Each Study generates a mesh (or many meshes, if you add a parametric sweep), and then performs each Step on that mesh. Each Step may be assigned to a different Physics interface, which can have different boundary conditions. The resolution to my confusion was just to move the second Study into a second Step within the first Study - so one Study with two Steps (and a parametric sweep).
Please login with a confirmed email address before reporting spam
Hi Ben, thank you for your insight.
I have structurally almost the same situation and looking for how to make a parametric sweep having some variations in boundary conditions for my model.
Unfortunately I still can not manage to use your recipe because if I have a parametric study (=study with a parametric sweep) and two steps where one step has "Modify model configuration for study step" option on to modify boundary conditions, I can not access the data from step 1. It's only the data calculated for step 2 that are accessible after the study is done.
Could you please share how you get access to the dataset of the first step?
Best regards, Albert