I am trying to create a channel for particles to move through.
I would like to rotate a region and using union to define a complex region similar to the one in the image, any idea how to define it, the region surrounded by the black lines?
I have the points that form these edges through a python file, which I want to use to write the LAMMPS file and create the data file.
As should be evident from the lack of responses, your description is too vague to give any specific suggestions, particularly since you require a specific method.
What is specifically needed:
the black lines do not enclose a region but are just two sets of parallel lines. So what exactly is the region you want to define?
what do you want to use the region for? creating atoms? creating a wall potential?
you say you want to build a “channel”. But then what kind of a channel would that be? How would be the openings be treated? How would periodic boundaries be handled? or non-periodic boundaries?
The images below are examples of a full channel. For now, I am thinking the opening and ending would be connected via periodic boundary.
I can define the regions using region block if the angles in the channel are 0 and 90 degrees. So I thought I could continue with it, I am open to other ideas on how to build it. I thought it would be easy if I could define the regions to create atoms and then use them with fix wall/region to confine them to the channel.
Fix wall/region cannot handle such complex constructs. Specifically where there are overlaps of regions or edges where regions connect, there are cases where forces would be ill-defined or it is difficult to decide between inside and outside.
Thus, you would have to use particles to represent the channel. This can be done with regions and create_atoms and delete_atoms in multiple setps or you can build an STL file and use that with create_atoms.
Hello,
With the STL file and create_atoms, is it possible to create a complex geometry (like strews) in LAMMPS and let it rotate?
I know we can use LIGGGHTS to do this, but LIGGGHTS uses a very old version of LAMMPS (Nov 2013), so there are lots of limitations for simulation.
I would like to find someway to swift to LAMMPS, however the main issue is the moving and complex mesh.
I am very looking forward to hearing from you; thank you in advance.
This is a new question is at best only vaguely related to the discussion that you are adding it to. This is called “hijacking a thread” and considered a violation of online forum etiquette and thus something to NOT do in online forums. Doing this for multiple threads is even worse since people may not realize that your question has already been answered in a different thread.
Please see the forum guidelines post to understand some of the best practices we expect here and thus to avoid blunders like this in the future.