Learning Forum

[zoelle.wong]

Dear ANSYS Support Team, I'm trying to incorporate the sliding mesh technqiue with a more complicated setup (as seen below).  When I create my domain, I mesh zone 2 to be a region of structured cells and zone 2 to be a region of unstructured cells. This worked for me before with a simplified case. Unfortunately this time, I get an error from ANSYS's TUI panel saying "11: mpt_accept: error: accept failed: Bad file descriptor" and "unable to determine solver configurations (grid sections reached)" My understanding is this error message is related to a mesh that is too complex... is this correct? If so, what set-up would you recommend or would it better to change to a different dynamic meshing technique given the schematic below? I also tried creating a unstructured mesh with structured cells near the tulip (no separate zones) and separated the zones via marked registers in Fluent. When I tried this strategy, I wasn't able to create a mesh interface between the "structured" and "unstructured" cells. Thank you again so much!!  

[zoelle.wong]

[Rob]

What direction is the mesh sliding? Is this 2d, 2d-axi or 3d?

[zoelle.wong]

This is a 2D mesh, and zone 2 is sliding to the left. Velocity vector in UDF is specified as v = <-2.5,0,0> m/s

Eventually, I will have to make this a 3D mesh but the velocity vector would be the same

[Rob]

OK. So Moving Deforming Mesh. You'll need to look at layering (I suspect). It's not sliding mesh: that's for fixed mesh systems. 

[zoelle.wong]

Ah my mistake (thought they were interchangeable!) I'm trying to incorporate a layering mesh with the set up above, which worked previously for a case without curved edges. Before, I meshed a grid where one domain had structured cells and the other had unstructured cells; the two domains were joined by a mesh interface.

When I did this procedure for the geometry illustrated above, I received the error:

"11: mpt_accept: error: accept failed: Bad file descriptor" and "unable to determine solver configurations (grid sections reached)"

I'm couldn't find what this error means in the documentation 

[Rob]

You need layering to allow the zone to move - I've not seen that error before but it's likely Fluent not liking the grid you're proposing. Have a look at the in-cylinder (IC) examples as you're modelling a valve or piston motion.