|
[Sponsors] |
How to speed up simulations even further, and is my simulation setup correct? |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
July 24, 2024, 07:35 |
How to speed up simulations even further, and is my simulation setup correct?
|
#1 |
New Member
johnny simson
Join Date: Nov 2023
Posts: 18
Rep Power: 2 |
Ive tried simulating a flapping door on ansys fluent, here's the current mesh im using. Its a partially structured mesh? The immediate box/face surrounding the flapping door is unstructured. And the other faces are structured as you can see in the image.
I have also attached the mesh independence test that I've been running and i wanted to get some validation from you guys. Please ignore the refinement number mentioned in the picture attached. here's the idea behind those numbers: 1) I set the number of flaps to be recorded as 2, and from those 2 flaps and the angular frequency, i found the flow time as 45 seconds. 2) Since i couldn't find the appropriate size of an average element in my mesh, i went with the face meshing value i specifically gave for my bounding box (the face that contains the slit/door) 3) From that value, i found the appropriate time step size for my 10m/s inlet 4) from the appropriate time step size value, i found the number of time steps required but doing this: time step size * number of time steps = 45 Now here's where the trouble begins, I set my convergence as 10e-04, and it quickly converged. Same for 10e-05. So I'm stuck with 10e-06 and its been 2 days and still nothing, i kind of have to speed things up. Now do I use the values of the cl cd from my lower convergence case and use those values for the mesh independence test? Or wait until my 10e-06 one gives me a convergence so that i could use that instead? I have 5 more cases to go with, and idk how to speed things up. I really want something to show by this week? Is there any way to speed it up? reduce the number of cases im running? Idk what its going on. I feel so lost. I have a case with 9 million time steps, i feel like im doing something terribly wrong here. Please correct my stupidity, im open to hear it. |
|
Tags |
converge cfd, flapping motion, mesh 2d, mesh independence study, solution converged |
|
|