undulating foil, negative cell volume

koshurkoorkoshurkoor Member
edited September 28 in Fluids

I have to simulate the undulating hydrofoil. I am using symmetry boundary conditions (top, bottom, sides) to mimic the 2d effect on the hydrofoil. It undulates as per the udf when viewed in Display zone motion. However, i keep getting negative cell volume error at a particular time-step. I have tried to decrease the time-step, i have tried all different settings for the dynamic mesh. I am not sure what the error is, or how to improve! HELPP!


  • RobRob UKForum Coordinator

    It looks like the solver is squashing & stretching cells at the trailing edge. Why are you running 3d, and how did you choose the min and max sizes in the Mesh Methods panel?

  • I used the default values for min and max from the 'mesh length scale' box. I have, however, tried the original default parameters too. 

    My udf won't work on 2D. I am using the node motion based on simple beam deflection as the base for the oscillation in the trailing edge. 

    The only way i could make it work is by creating a 3D, while meshing - for the thickness, i choose the number of divisions = 1, and use the faces as symmetry to be able to mimic the 2D flow in the solver. I tried using periodic boundary conditions too. 

  • I have tried several values of remeshing parameters and believe thats why i am getting negative cell volume. How do i choose the right values?

  • RobRob UKForum Coordinator

    Have a look at the tutorials, but you'll need something around the size of the cells in the region of the moving blade.

  • Hello again,

    I replicated the same motion in a structured mesh. Why is the mesh deforming outside the domain? I have tried various settings of dynamic meshing but this keep happening. There is not enough information on this.

    Please, help!

  • RobRob UKForum Coordinator

    Redisplay using orthographic projection (there's an icon around the graphics window).

  • RobRob UKForum Coordinator

    It'll be linked to the mesh motion, check the zone is set to move and not the outer boundary.

  • I choose the top and bottom surfaces as stationary so that the nodes of that zone are excluded, but that really didn't help. The mesh still collapses.

    There are no tutorials available on sinusoidal motion using DEFINE_GRID_MOTION. I have tried all other logical ways to improve this.

  • RobRob UKForum Coordinator

    The problem is with the mesh motion & settings: the tutorials/videos cover the motion, the shape of that motion is irrelevant.

  • koshurkoorkoshurkoor Member
    edited October 26

    Unfortunately, there is only one example even in the latest manual, the one with cantilever beam, and that is not enough to learn from. I'm having hard time fixing this mesh motion. I created a smaller domain, tried to define zones, move it as a zone as well. None of that fixed this kind of problem. It does not have to do with the setting, but definitely the mesh motion.  Youtube covers mostly cgmotion which is not relevant to my case as my equations are much more complex in nature and can be imitated by define_grid_motion only. It maybe that something else needs to be added to the udf to make it work.

Sign In or Register to comment.