[D.M]

Hi,
I have a problem with the dynamic mesh, and i don't know how to solve it , every thing is shown in the picture below.

(I have already tried everything like increasing or reducing spring factor, using size function in remeshing method and ... but nothing seems to fix this problem, i think the problem can be fixed if there is a method which increases the number of mesh nodes on the boundaries in each time step!)

By the way, the udf used for moving the top wall is DEFINE_GRID_MOTION.

plz help me.

[RaiderDoctor]

Wow, it looks like you are using two different meshing strategies here: tetrahedral on the profile at the moving wall, and hexahedral everywhere else. I don't think I've seen that before, would you mind posting how you did that?

As to your reduction in mesh quality. What you are experiencing is actually quite common, and it makes sense when you really think about it. You are basically taking a super-fine mesh, and then simply stretching it out which will create this much coarser mesh you see here. A simple way to combat this is to increase the number of elements at the meshing stage so that when the domain moves to its final position, the element quality is still acceptable.

If you have your heart set on remeshing strategies, however, try checking out this video (https://www.youtube.com/watch?v=aTIwmEyp3_8&t=187s at 4:06). While not exactly like your case, the node positions of the mesh are updated more aggresively so that the overall skewness is maintained (he goes in depth about which global remeshing strategies he uses in this vid. https://www.youtube.com/watch?v=eMId3eNhZSM). Try switching to diffusion-based smoothing. I personally have better luck with that.

Let us know if this works!

[D.M]

Originally Posted by RaiderDoctor

Wow, it looks like you are using two different meshing strategies here: tetrahedral on the profile at the moving wall, and hexahedral everywhere else. I don't think I've seen that before, would you mind posting how you did that?

As to your reduction in mesh quality. What you are experiencing is actually quite common, and it makes sense when you really think about it. You are basically taking a super-fine mesh, and then simply stretching it out which will create this much coarser mesh you see here. A simple way to combat this is to increase the number of elements at the meshing stage so that when the domain moves to its final position, the element quality is still acceptable.

If you have your heart set on remeshing strategies, however, try checking out this video (https://www.youtube.com/watch?v=aTIwmEyp3_8&t=187s at 4:06). While not exactly like your case, the node positions of the mesh are updated more aggresively so that the overall skewness is maintained (he goes in depth about which global remeshing strategies he uses in this vid. https://www.youtube.com/watch?v=eMId3eNhZSM). Try switching to diffusion-based smoothing. I personally have better luck with that.

Let us know if this works!

Hi,
I really don't know how to thank u, thanks for the reply , actually i had seen those videos on youtube, but the diffusion method which u recommended me to use really helped me, i coarsened the mesh and used diffusion method with diffusion parameter of 1 and it really worked out, thanks again for the reply.


and about the mesh that u asked for, it is pretty simple , i did it like the picture below, i created two parts (fluid1 & fluid2), and meshed them separately, one of the parts was meshed quad and the other part tri, and the line between the two parts was considered to be interior.