Hi Markus,
I set up a webpage with several meshes, see
http://cell-in-fluid.fri.uniza.sk/en/content/oif-espresso
You can download the zip files and use them.
The difference between the sphere and the ball is this one: sphere
is a 2D surface triangulated with triangles and each edge is
common with exactly two triangles. Ball is a 3D mesh with
tetrahedra, so also the inside of the sphere is triangulated.
However, in the files provided, only two kinds of data is given:
1. the positions of the mesh points (points are on the surface of
the sphere as well as inside of the sphere, almost uniformly
distributed.)
2. the incidences of the triangles in the mesh.
Notice, that no information is given on which four triangles form
a tetrahedron. We consider the mesh simply as a composition of
points, triangles and edges. This results however in restricted
use of elastic forces in the OIF framework. With these balls, you
are NOT allowed to use the bending, because there are not exactly
two triangles common for one edge. Also, you can NOT use the
volume and the global surface preservation. So for this type of
balls, you can use the stretching (it is defined for each edge)
and local area conservation only, because this is defined for each
triangle.
I hope the explanation is clear,
Ivan
Dňa 25.2.2014 17:16 Wink, Markus wrote / napísal(a):
Hi
Ivan,
Anyway, what kind of
meshes do you need? I could provide you with some spheres,
balls or red-blood-cells
So far, I am interested in simple geometries
like spheres. So it would be great, if you could provide me
one of those. Anyway, what is the difference between balls
and spheres? By “ball”, do you mean structures like a bucky
ball?
Greetings Markus
Von:
Ivan Cimrak [mailto:address@hidden]
Gesendet: Dienstag, 25. Februar 2014 15:32
An: Wink, Markus; address@hidden
Betreff: Re: AW: [ESPResSo-users] OIF, deviation
in "origin"-command
Hi Markus,
We use gmsh (http://geuz.org/gmsh/)
software for producing triangulations. There are however
many others, commercial or non-commercial, just type "mesh
generation tools" in your favorite search engine. Anyway,
what kind of meshes do you need? I could provide you with
some spheres, balls or red-blood-cells.
Ivan
Dňa 25.2.2014 15:25 Wink, Markus wrote / napísal(a):
Hi,
One more comment, your
objects should be numbered consecutively, starting from 0
Thanks a lot for the command on the
numbering. That’s a very good point!
When an OIF object
deforms, the origin definitely does not need to coincide
with the say center of the mass.
That is true, but so far no integration has
be done, so no deformation should have taken place.
Simply, the
triangulation mesh given in the
$fileNodes file is not centered around the
"expected" center of the sphere.
Probably, the mesh generator did not produced balanced
triangulation, which is in fact not a mistake.
That is what I thought. Anyway, I have never
produced a triangulation on my own. Is there any
recommendation of software/literature… you could give?
Greetings
Markus Wink
Dear Markus,
The switch "origin" in the oif_object_analyze command,
computes the average of the coordinates of the current
positions of the mesh points. When an OIF object deforms,
the origin definitely does not need to coincide with the
say center of the mass.
There is no wrong calculation in your script. Simply, the
triangulation mesh given in the
$fileNodes file is not centered around the
"expected" center of the sphere. Probably, the
mesh generator did not produced balanced triangulation,
which is in fact not a mistake.
One more comment, your objects should be numbered
consecutively, starting from 0. I strongly recommend to
keep this in mind. Normaly, the software should complain
when you do not do this, there will be a patch on this
soon.
Ivan
Dňa 25.2.2014 15:02 Wink, Markus wrote / napísal(a):
Dear
all,
I have noticed a quite strange behavior in
the command “origin” for the OIF-implementation.
Probably somebody can help me out with that. In my
script, I have put a sphere in a channel at a specific
position. Afterwards I wanted to check, which result the
command “origin” gives. The interesting parts of my
scripts are:
oif_create_template template-id 0
nodes-file $fileNodes triangles-file $fileTriangles ks
10 kb 10 kv 50
oif_add_object object-id 1 template-id 0
origin 15 15 15 rotate 0 0 0 part-type 1 mass 1
set x [oif_object_analyze object-id 0
origin]
puts $x
So I put the object with its origin @ 15 15
15; command “origin” prints out: 15.008655470737903
15.02980503816793 14.995383511450385
No integration of the system has be done so
far.
I can think of two reasons for the
deviation:
1)
The source code is correct and the
deviation results because of rounding errors
2)
The template I have used (I used the
sphere_nodes/triangles.dat found on git hub) is somehow
not correct. Actually, I have noticed, that the points
given do not exactly fulfill the relation for a sphere
(x^2+y^2+z^2=r^2). So it could be that the deviation
origins from that artefact.
3)
The calculations performed by the command
“origin” has some error. Unfortunately, I could not
locate, where the calculation is performed in the source
code.
Does anyone have a clue?
Greetings
Markus Wink
|