Negative volume in dynamic mesh

    • omid1988

      I am modeling the motion of a box within a 3D domain. I have 30 inflation layers around the box. A UDF is defined to apply the constant velocity motion. I tried to separate inflation layers and apply the UDF to both box and its inflation layers while the mesh is getting updated.

      I set the timestep so that it corresponds to box motion for the quarter size of the mesh outside the inflation layer. Therefore, it's not moving a lot at each time step. However, I get a negative volume error at the first time step.

      I set smoothing(Spring) with default values and remeshing based on the mesh info, for the dynamic mesh.

      Could you let me know what causes this negative volume?

      Is there a way to see cells with negative volumes in Fluent?




    • Rob
      Ansys Employee

      Mesh quality may help with visualisation, skewness will be 1 for negative volume cells.  What's causing the motion?

    • omid1988

      Sure, I will check it. The problem with visualization is that I have about 5 million mesh nodes and it's hard to clearly see cells in 3D, but I will do my best to check.
      There are actually two boxes that have a constant velocity of 6 [m/s] with opposite direction (getting close to each other), and incoming flow is 18 m/s. I want to see flow patterns around the boxes and find the forces applied on them.

    • Karthik R


      Please create a cell register to visualize the negative cells. Or as suggested by Rob, please visualize the bad cells using this option.




    • omid1988

      When I click on "Cell Registers" it's blank and I couldn't create anything.

      I could use contour/mesh/cell equiangle skew to visualize skewness on different surfaces but it's not ideal.

    • omid1988

      Thanks Rob and Karthik!

      I am working on another mesh with inflation layers and not getting negative volume. 
      Before running simulation, I separated the inflation layers and applied the motion to the inflation layers the same as the box. Maybe, the cell separation did not worked properly for the first case I was working on.

      Now I get divergence problem. I will create a separate post for that.

Viewing 5 reply threads
  • You must be logged in to reply to this topic.