[KevinMcTaggart]

I'm trying to predict forces on STL objects in steady flow. As a test case, I've modified Tutorial 5 (circular pier) by vertically extending the domain from -0.2 m to 0.6 m, and using an STL model of the vertical cylinder from 0.0 m to 0.4 m elevation. After some effort, I've determined that the pressure on the bottom of the cylinder is half of what it should be when using defaults of X 33 1 and X 40 3. A pressure probe suggests to me that the hydrostatic pressure in the domain is generally correct. Function calls include sixdof_gc::forces_stl calling interpolation::ccipol4, which calls lint4. I noticed that the call to lint4 uses a vertical weight wc of 0.5, which leads me to believe that the pressure within the STL cylinder could be 0.0, causing the interpolated pressure on the surface to be half its actual value.

[valgrinda]

Within 6DOF_gc, the pressure is reconstructed also inside the structure using ghost cells, so there the forces should be ok. You are right though, the point probe will include a "0" for pressure values inside the structure. I will change that for the next release (also for the probe line).

We have now a new and updated 6DOF version including mooring based on direct forcing (see 6DOF_df_...cpp files), see these open-access papers for more information:
https://doi.org/10.1016/j.apor.2020.102410
https://doi.org/10.1002/fld.4894

This method will give you improved force calculation for floating bodies.

[KevinMcTaggart]

Thank you Hans. Are there input options for obtaining the correct pressure, rather than half, on the bottom of the cylinder? I believe that the cylinder bottom is at a cell boundary, which could perhaps be a factor. For drag force, the result is more encouraging. Drag on Reef3D STL cylinder is 2.78 N, compared to 2.52 N for Divemesh S 36 vertical cylinder in same flow, with force evaluated using Reef3D P 81. Note that I used smaller cell size (half) for Reef3D STL case.