Hi-
I have a question about the fix wall/region. The documentation says that convex corners on regions are handled fine if the region is a primitive one, but not if it is a union or intersection. If I write my own region geometry code to create a region with convex corners, will these corners be handled well by fix wall/region?
Thanks, Steven Strong
Hi-
I have a question about the fix wall/region. The documentation says that
convex corners on regions are handled fine if the region is a primitive one,
but not if it is a union or intersection. If I write my own region geometry
code to create a region with convex corners, will these corners be handled
well by fix wall/region?
there really isn't much that can be said beyond: if in doubt, make a test.
axel.
Thanks,
I have a related question. There seems to be some confusion with regards to how fix wall/region handles atoms on the inside/outside of the region. The documentation says “Note that the region volume can be interior or exterior to the bounding surface, which will determine in which direction the surface interacts with particles”. It doesn’t specify which way the surface normal points, but I would suspect that the surface normal points away from the interior of the region. Later on, the documentation says “you must insure that r is always > 0 for all particles in the group, or LAMMPS will generate an error. This means you cannot start your simulation with particles on the region surface (r = 0) or with particles on the wrong side of the region surface (r < 0).” I can’t find any documentation about which side of the surface is the wrong side, but going with my above assumption, that the surface normal points out, that would mean that the wrong side of the surface is on the inside of the region. This agrees with an error message I’ve gotten, which says “ERROR: Particle on or inside surface of region used in fix wall/region”. However, that error message seems to be generated in the wrong case. For example, see the attached test script, which places one particle at the origin and makes a region that does not include the origin. The particle is not inside or on the surface of the region, but I still get that error message.
Thanks, Steven
in.test (384 Bytes)
Fix wall/region doesn’t actually know anything about the region
geometry. It relies on several methods, provided by the
region, that give it the info it needs to push off particles.
So if you write your new region correctly, it should just work.
Look at region_block.cpp for an example that has to handle corners.
If you can figure out how to make regions intersects or unions work
better/smarter for those methods that fix wall/region uses, send us some code for it!
Steve
Thanks,
I have a related question. There seems to be some confusion with regards to
how fix wall/region handles atoms on the inside/outside of the region. The
documentation says "Note that the region volume can be interior or exterior
to the bounding surface, which will determine in which direction the surface
interacts with particles". It doesn't specify which way the surface normal
points, but I would suspect that the surface normal points away from the
interior of the region. Later on, the documentation says "you must insure
that r is always > 0 for all particles in the group, or LAMMPS will generate
an error. This means you cannot start your simulation with particles on the
region surface (r = 0) or with particles on the wrong side of the region
surface (r < 0)." I can't find any documentation about which side of the
surface is the wrong side, but going with my above assumption, that the
surface normal points out, that would mean that the wrong side of the
surface is on the inside of the region. This agrees with an error message
I've gotten, which says "ERROR: Particle on or inside surface of region used
in fix wall/region". However, that error message seems to be generated in
the wrong case. For example, see the attached test script, which places one
particle at the origin and makes a region that does not include the origin.
The particle is not inside or on the surface of the region, but I still get
that error message.
i guess that is a question on how the code (or rather its author)
interprets the "side in" and "side out" flag of the region with
respect to where the wall is.
the current implementation of fix wall/region seems to interpret that
"side in" means that the *wall* would be on the inside of the region
and thus only a particle within that region would experience the wall
potential. that would invert the logic of whether a particle that is
outside the region as it applied to other commands. so the way how you
interpret the situation would require to use "side out" for the
region.
not sure if that could be called a bug. i find it a bit surprising as
well, but if you look at it from the perspective of the wall there is
some sense to it.
axel.
Ok, thanks. Do you think an update to the documentation could make the side in/out issue more clear?
-Steven
Side in/out applies to everything a region is used for. The setting simply
toggles what is considered “inside” the region. I.e. the interior of
a sphere or the exterior.
When you create particles “inside” a region, they are either in the
interior or exterior. When the region boundary is used as a wall,
the particles have to always be “inside” the region, whether
that is interior or exterior. Thus the normal used for the region
boundary toggles with in/out. The surface
doesn’t interact with particles in the other direction, b/c
the particles can’t be there. There is no ambiguity about
the distance from the surf to a particle, b/c the particle
is always “inside”.
All of that is consistent with this statement in the doc:
“Note that the region volume can be interior or exterior
to the bounding surface, which will determine in which direction the surface
interacts with particles”
Please suggest something for the doc page to make it more clear.
Steve
Program Description
(Program not available for streaming.) What would it be like to go inside the mind of an animal? Now, the revolutionary science of animal cognition is revealing hard evidence about how animals understand the world around them, uncovering their remarkable problem-solving abilities, and exploring the complexity of their powers of communication and even their emotions. In the three-hour special “Inside Animal Minds,” NOVA explores these breakthroughs through three iconic creatures: dogs, birds, and dolphins. We’ll travel into the spectacularly nuanced noses of dogs and wolves and ask whether their reliance on different senses has shaped their evolution. We’ll see through the eyes of a starling in flight and test the tool-using skills of one of the smartest birds, the crow. We’ll listen in as scientists track dolphins in the Caribbean and elephants on the African savannah, trying to unlock the secrets of animal communication. As we discover how researchers are pushing the animal mind to its limits, we’ll uncover surprising similarities to—and differences from—the human mind.
(Program not available for streaming.) What would it be like to go inside the mind of an animal? Now, the revolutionary science of animal cognition is revealing hard evidence about how animals understand the world around them, uncovering their remarkable problem-solving abilities, and exploring the complexity of their powers of communication and even their emotions. In the three-hour special “Inside Animal Minds,” NOVA explores these breakthroughs through three iconic creatures: dogs, birds, and dolphins. We’ll travel into the spectacularly nuanced noses of dogs and wolves and ask whether their reliance on different senses has shaped their evolution. We’ll see through the eyes of a starling in flight and test the tool-using skills of one of the smartest birds, the crow. We’ll listen in as scientists track dolphins in the Caribbean and elephants on the African savannah, trying to unlock the secrets of animal communication. As we discover how researchers are pushing the animal mind to its limits, we’ll uncover surprising similarities to—and differences from—the human mind.
Maybe not a change to the documentation then, but a change to the error message “ERROR: Particle on or inside surface of region used in fix wall/region”. As you say, when the region boundary is a wall, the particles must be inside the region, so it would make more sense to me if the error message said “ERROR: Particle outside surface of region…”.
Thanks, Steven
good suggestion - I changed the error message.
Steve