CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Negative Volume - TET Mesh Compression (http://www.cfd-online.com/Forums/fluent/74664-negative-volume-tet-mesh-compression.html)

ChesterP April 6, 2010 05:42

Negative Volume - TET Mesh Compression
 
Hi everybody,

I'm doing 3D in-cylinder simulations.

When the piston head is moving from BDC to TDC and reaches the cut-off distance I first have skewness issues and then at around 297-298 negative volumes appearing.

So there are problems only for the compression of the tetrahedral mesh.

I tried to vary :
- Spring Constant Factor between 0 and 0.5;
- Boundary Node Relaxation between 0.2 and 1;
- Convergence Tolerance between 0.001 and 0.0001;
- Number of Iterations between 20 and 100;

I remesh at each time step with a crank angle step of 0.05.

Thanks in advance

yubaibai88 April 6, 2010 06:17

Quote:

Originally Posted by ChesterP (Post 253305)
Hi everybody,

I'm doing 3D in-cylinder simulations.

When the piston head is moving from BDC to TDC and reaches the cut-off distance I first have skewness issues and then at around 297-298 negative volumes appearing.

So there are problems only for the compression of the tetrahedral mesh.

I tried to vary :
- Spring Constant Factor between 0 and 0.5;
- Boundary Node Relaxation between 0.2 and 1;
- Convergence Tolerance between 0.001 and 0.0001;
- Number of Iterations between 20 and 100;

I remesh at each time step with a crank angle step of 0.05.

Thanks in advance

hi chesterp:
I't so glad to see you again
maybe you can try to change the parameter in remeshing like min length scale and max length scale.
do not set them too much adjacent to the min and max length scale in your domain

yubaibai88 April 10, 2010 01:34

2 Attachment(s)
Quote:

Originally Posted by ChesterP (Post 253305)
Hi everybody,

I'm doing 3D in-cylinder simulations.

When the piston head is moving from BDC to TDC and reaches the cut-off distance I first have skewness issues and then at around 297-298 negative volumes appearing.

So there are problems only for the compression of the tetrahedral mesh.

I tried to vary :
- Spring Constant Factor between 0 and 0.5;
- Boundary Node Relaxation between 0.2 and 1;
- Convergence Tolerance between 0.001 and 0.0001;
- Number of Iterations between 20 and 100;

I remesh at each time step with a crank angle step of 0.05.

Thanks in advance

hi chesterp:
I think I've find out where your problem is.
but first I want say sorry for the previous answer to you for it's not right.and second is that I can only demonstrate it in 2D simulation.

now it's my opinion:
in the cold flow simulation,I follow all what the tutorials told me to do without set the piston stroke cutoff 0.01m.when start mesh motion,I gained the mesh as prestent in picture 1.it is obviously that the mesh adjacent to the cylinder boundary is very bad.
then I creat a new dynamic mesh zone.cylinder-side-tri,which is the boundary of the triangular zone.after mesh motion,the mesh I gained is showen in picture 2.it rapidly improved after setting a new deforming boundary zone.
note that the motion I defined is from TDC to BDC.
after the pistion reach BDC.I delete the deforming boundary zone defined in TDC.then start mesh motion,finally negative volume appears when piston reach the cutoff distance.
that's all I do to experiment the appearing of negative volume.sorry to say I do it in 2D simulation,but boldly can I guess where your problem is.maybe it is because you did not define a deforming boundary zone which is the boundary of your tetrahedral zone.fluent assumes it wall if you don't define it.when the piston moving up,the boundary nodes are compressed and finally negative volume appears.
hope it help

ChesterP April 11, 2010 02:05

Quote:

Originally Posted by yubaibai88 (Post 254040)
hi chesterp:
I think I've find out where your problem is.
but first I want say sorry for the previous answer to you for it's not right.and second is that I can only demonstrate it in 2D simulation.

now it's my opinion:
in the cold flow simulation,I follow all what the tutorials told me to do without set the piston stroke cutoff 0.01m.when start mesh motion,I gained the mesh as prestent in picture 1.it is obviously that the mesh adjacent to the cylinder boundary is very bad.
then I creat a new dynamic mesh zone.cylinder-side-tri,which is the boundary of the triangular zone.after mesh motion,the mesh I gained is showen in picture 2.it rapidly improved after setting a new deforming boundary zone.
note that the motion I defined is from TDC to BDC.
after the pistion reach BDC.I delete the deforming boundary zone defined in TDC.then start mesh motion,finally negative volume appears when piston reach the cutoff distance.
that's all I do to experiment the appearing of negative volume.sorry to say I do it in 2D simulation,but boldly can I guess where your problem is.maybe it is because you did not define a deforming boundary zone which is the boundary of your tetrahedral zone.fluent assumes it wall if you don't define it.when the piston moving up,the boundary nodes are compressed and finally negative volume appears.
hope it help

Thank you very much for your help. :)
I already defined the wall as deforming for my tetrahedral mesh so it is not the issue I think :(

When the piston reach the cutoff distance and that the tetrahedral mesh should be then smoothed and remeshed it is like the process is not done quickly enough compared to the motion of the piston. Hence negative volumes are appearing. I already chose a small crank angle step (0.05) and remeshed at each step. I will try to run a simulation with a smaller rotation speed of the engine and I will tell you what.

Have a nice day

yubaibai88 April 11, 2010 02:49

Quote:

Originally Posted by ChesterP (Post 254131)
Thank you very much for your help. :)
I already defined the wall as deforming for my tetrahedral mesh so it is not the issue I think :(

When the piston reach the cutoff distance and that the tetrahedral mesh should be then smoothed and remeshed it is like the process is not done quickly enough compared to the motion of the piston. Hence negative volumes are appearing. I already chose a small crank angle step (0.05) and remeshed at each step. I will try to run a simulation with a smaller rotation speed of the engine and I will tell you what.

Have a nice day

all right
if the negative volume still appears,sent your cas/dat to me.I will try to check.
good luck


All times are GMT -4. The time now is 06:59.