[iphenix]

When I run an immersed case, the solver give me the following warning:
| ********* WARNING ********* |
| There are 1517893 interpolation points ( 33.42%) that are not |
| properly mapped onto a source domain. This will likely introduce |
| incorrect interpolated values for the CEL callback on the immersed |
| boundary. If part of the immersed solid moves out of the fluid |
| domain, please check the problem setup carefully. If part of the |
| immersed boundary overlaps with fluid boundary, please break the |
| immersed boundary into two or more boundaries, and use the CEL |
| callback on a proper immersed boundary. If this is due to a slight |
| difference in mesh, please increase the expert parameter "octree |
| bndbox tolerance".
************************************************** ******

How I deal with the warning, ignore the warning or modify my model?
do the warnings affect the accuracy of calculation result?

[ghorrocks]

My understanding is immersed solids do not handle properly where the immersed solid overlaps a wall boundary. So use the results with caution, check the results carefully before trusting them.

[iphenix]

Thank you for your answer.

[urosgrivc]

It is usualy way beter to do imersed solid cases vith rigid body if posible.
there you can put inflation layers and other important mesh features for enhanced wall treatment.
Imersed solid is quite (well wery) bad at values near or on the wals usualy nothing usefull comes out unles extremly fine global mesh is used, if your main result is a wariable on or near the wall try to make your simulation better.
What is your case about?

[iphenix]

Originally Posted by urosgrivc

It is usualy way beter to do imersed solid cases vith rigid body if posible.
there you can put inflation layers and other important mesh features for enhanced wall treatment.
Imersed solid is quite (well wery) bad at values near or on the wals usualy nothing usefull comes out unles extremly fine global mesh is used, if your main result is a wariable on or near the wall try to make your simulation better.
What is your case about?

My case is to solve the force of a immersed solid in a fluid.
There is an external flow and an internal flow that flows through the immersed solid at the same time.

[ghorrocks]

An image would help.

[urosgrivc]

And the error probably tryes to say that your solid has left the fluid domain.
this can easily hapen if conditions are vrong or constrains arent corect, wery easily infact.
first few iterations-timesteps are very critical for only slightly constrained imersed solids. Forces which work on the body can be badly converged and huge numbers may accelerate the body out of the domain, check your results for what has happened. It totaly depends on the problem but physical timescale comes in handy in some cases like boat hull on a water surface if simulation is steady state. And yes a picture and explanation will help us help you

[abubakarizhar]

Dear

did you resolve this issue . Im trying to simulate vortex induced vibrations of a cylinder with immersed solid. Although ive used a fine mesh in fluid around immersed solid even then the error comes saying that 90% of nodes are not properly mapped. This warning doesnt show up when boundary tracking is turned off. Anyone has idea?

[urosgrivc]

abubakarizhar try to use rigid body instead

[abubakarizhar]

Dear
I am already using rigid body. I am saying that warning message doesnot show up when boundary model is off but when modified forcing function is used warning comes with almost 90% of nodes not properly mapped right from the first iteration. one more thing my rigid body position is ok and its well inside the fluid domain then why this warning.

[abubakarizhar]

| ********* WARNING ********* |
| There are 6904 interpolation points ( 81.19%) that are not |
| properly mapped onto a source domain. This will likely introduce |
| incorrect interpolated values for the CEL callback on the immersed |
| boundary. If part of the immersed solid moves out of the fluid |
| domain, please check the problem setup carefully. If part of the |
| immersed boundary overlaps with fluid boundary, please break the |
| immersed boundary into two or more boundaries, and use the CEL |
| callback on a proper immersed boundary. If this is due to a slight |
| difference in mesh, please increase the expert parameter "octree |
| bndbox tolerance". |
| Boundary Patch Name : RR |
+--------------------------------------------------------------------+

how can I increase the octree bndbox tolerance?

[ghorrocks]

urosgvic is recommending you use moving mesh rigid bodies, not immersed solids. I do not agree with his suggestion, for this application immersed solids appears to be adequate and is much easier to apply than moving mesh rigid bodies.

I think the error message you are getting is because you have a large number of nodes of the immersed solid lying outside the fluid domain. You should check your model than the immersed solid lies entirely inside the fluid domain.