v5.2 Beta - mDBC issues
Hi. Following the 6th Dual SPH workshop, I have tried very hard including the past 4 days to expand my simulations with mDBC from 2D (which work well) to 3D but I have a few problems and questions which are unresolved even after reading the guides, the workshop mDBC keynote talk by Alex and other forum posts, checked the 'normals', etc
- what is <distanceh value="3.0" comment="Maximum distance (H*distanceh) to compute normals data (default=2)". Why in some examples is 2.0 and others is 3.0 and what is the impact?
- in <initialise> for piston cases, what is <maxdisth v="0" comment="Maximum distance to boundary limit. It uses H*maxdisth (default=2)" />?
- When creating various elements e.g. obstacles, boundary floor, boundary walls, etc, what should be the order and why? I understand that water is created last to fill the void, but with the other elements, if I create an element which overlaps a previous one, does it overwrite the particles in the lattice?
I want to replicate a physical model tank with a piston and a complex seawall and/or obstacle. My simulations in 2D run fine, but when expanding to 3D there are issues with particles leaving the domain or falling through the solid boundary. The 2D profile is complex therefore to transform to 3D my idea was to extrude in the 'y' direction and add two side walls which run in the 'x' direction to contain the water. This is why I failed to adapt any of the examples that come in the v5.2 Beta
- 07_WavesCylinder. Note that the base case without any modification has water particles falling down the side of the ramp, not many but a few. When I try to introduce an obstacle like the example in the 3D Dam Break there are particles falling down through the obstacle. Also, the example extrudes the boundary (or beach) in the 'x' direction and in my case I need to extrude in the 'y' direction because the structure is complex so I can't follow the method.
- 06_WaveTank. I tried first creating an .stl file and adding the side walls. Then I tried to create the bottom and structure using the extrude function or drawboxes, but in all cases with side walls. I followed and implemented the process with great care to ensure that the boundary was at Dp/2, either towards the inside or not, inverting normals for the side walls so they point at the fluid, etc etc, and nothing works.
- my objective and process. maybe there is something wrong there
- one image showing an error because particles are not where they should be after only a couple of seconds. This was from an attempt using an .stl file for the geometry and adding side walls
- example of particle generation where I tried to match an example from dsph and associated xml file which has particles leaving the domain after less than 1 second (I had to zip to be able to upload)
Just to let everyone know, that what has worked is to have the whole structure as .stl, including the side walls. Now there are no particles leaving the domain, no issues with particles disappearing down the sides of the tank, all the normals look good.
I am happy that works better now
It is great but I wonder did you build all the structures into one stl. document or in seperated stl. documents because I look at the shared animation that there is only the structure but not the side walls.
could youn please share yout xml and stl file with us.
Many THX in advandce