[didiean]

Hello, everyone~Thanks for looking into it~
I am doing some calculation on a deforming airfoil~
As it is known, Fluent can use boundary layer spring method to deform boundary layer during a moving-deforming mesh simulations.
For 2D calculation, I used structured grid and it works well.The 10%c rear of the airfoil could flap between -10deg~10deg. 90%c of the airfoil is stationary.
Then, I extruded the 2D grid to form a 3D grid, and used the modified UDF almost the same with the 2D UDF. But after the rear flapped about 3.4deg towards the suction side, negative grids showed up. I wonder why this happens. Could the boundary layer spring method be used well for 3D calculation? Or, do I have to do some modification when I deal with 3D calculation?

Some detailed pictures are listed as followed.
Picture 1 shows the negative grids location——the same relative positions across the spanwise. We can look into it in picture 2.
For the right showed in picture 4, we can see that it seems that the grids outside the deforming airfoil are pulled towards one direction. For the left showed in picture3, as the skewness of the first grid is too big, the first layers are pushed back, and the negative grids show up.
It seems that grids around the airfoil are pulled anticlockwise as the 10%c rear(on left of the minus grids) of the airfoil flaps anticlockwise.
But for the 2D calculation, grids near the stationary part of the airfoil seems unchanged.

I am very very confused. Could anybody help me out? Any suggestions are appreciated.
Thank you all.

[didiean]

Hope I have described my questions clearly. If not, please ask me in detail~thank you.

[didiean]

In the UDF manual, there is an example of a cantilever beam which I refer to. In the UDF example, there is one line of sentence as follows:

SET_DEFORMING_THREAD_FLAG(THREAD_T0(tf));

This sentence is used to active the deforming flag on adjacent cell zone, which means that the cells adjacent to the deforming wall will also be deformed, in order to avoid skewness.

But I can't find any description of SET_DEFORMING_THREAD_FLAG. Is there anybody familiar with this?
Do I have to modify it for 3D calculation?
It seems that if this sentence doesn't work for 3D, skewness occurs. But I am not sure whether it is the reason and how to modify it.

[didiean]

For 2D calculation, there seems to be little influence on the grid development to delete this sentence. So this may not be the reason for 3D grid fail.
Any suggestion? Call for help~please

Originally Posted by didiean

In the UDF manual, there is an example of a cantilever beam which I refer to. In the UDF example, there is one line of sentence as follows:

SET_DEFORMING_THREAD_FLAG(THREAD_T0(tf));

This sentence is used to active the deforming flag on adjacent cell zone, which means that the cells adjacent to the deforming wall will also be deformed, in order to avoid skewness.

But I can't find any description of SET_DEFORMING_THREAD_FLAG. Is there anybody familiar with this?
Do I have to modify it for 3D calculation?
It seems that if this sentence doesn't work for 3D, skewness occurs. But I am not sure whether it is the reason and how to modify it.

[didiean]

Has anyone ever used the boundary layer spring method for 3D calculation？Please help me~~
As the User Guide says, the spring-based smoothing method on non-triangular or non-tetrahedral cell zones is turned off by default. But why is the boundary layer influenced by spring constant factor when I haven't activated spring-on-all-shapes?

[didiean]

Has anyone used spring dynamic method for 3D structured model? I am still stuck~
Call for help~

[Rahul123]

I m modelling a 2d dynamic mesh problem.Cells adjacent to boundarys only deform.Can u tell me how to deform the entire domain?

[didiean]

Originally Posted by Rahul123

I m modelling a 2d dynamic mesh problem.Cells adjacent to boundarys only deform.Can u tell me how to deform the entire domain?

You can define the entire domain as Deforming type.

[Rahul123]

I have done that but only the cell layer adjacent to the rotating boundary is deforming.I dont knw wat the problem is

[Rahul123]

Its a rotating square 2d problem.I have divided the mesh into two zones.Square is surrounded by a circular mesh zone which is rotating with the cylinder.The left over domain is made deforming but as I said the cells adjacent to the circular zone is only deforming which I guess is causing the divergence after some time steps(150-200).What should I do?I m using smoothing and remeshing

[didiean]

Originally Posted by Rahul123

Its a rotating square 2d problem.I have divided the mesh into two zones.Square is surrounded by a circular mesh zone which is rotating with the cylinder.The left over domain is made deforming but as I said the cells adjacent to the circular zone is only deforming which I guess is causing the divergence after some time steps(150-200).What should I do?I m using smoothing and remeshing

I haven't quite followed you. Maybe, you could attach some pictures.

[Rahul123]

[ATTACH]mesh2.jpg[/ATTACH]
these are the pics

[Rahul123]

only 1 cell is deforming...others are intact

[didiean]

Originally Posted by Rahul123

only 1 cell is deforming...others are intact

I am not sure about the reason. Try to modify the parameters of remeshing.

[zheli_piglet]

hello didiean,

Did you finally manage to figure out how to set deforming flag on adjacent cell zone? I've met with the same problem? It will be nice if you could offer me some hints...thanks a lot!

Julie

[didiean]

I couldn't recall how I solve it. I might have set some parts of the domains or boundaries as rigid. Hope it helps.

Originally Posted by zheli_piglet

hello didiean,

Did you finally manage to figure out how to set deforming flag on adjacent cell zone? I've met with the same problem? It will be nice if you could offer me some hints...thanks a lot!

Julie