[GiDlist] Simplification of Geometry

GiDList Archive 2013 - 2005
1

Is there any way to simplify the given geometry in the attachment. I have tried collapsing it and rebuilding the surfaces but this has been unsuccessful. What I am hoping to achieve is a simplified model with as few surfaces as possible so I can easily assign boundary conditions and loads. Any help would bee greatly appreciated.

Regards

R.M

UNIVERSITY OF CAPE TOWN This e-mail is subject to the UCT ICT policies and e-mail disclaimer published on our website at http://www.uct.ac.za/about/policies/emaildisclaimer/ or obtainable from +27 21 650 9111. This e-mail is intended only for the person(s) to whom it is addressed. If the e-mail has reached you in error, please notify the author. If you are not the intended recipient of the e-mail you may not use, disclose, copy, redirect or print the content. If this e-mail is not related to the business of UCT it is sent by the sender in the sender’s individual capacity.

-------------- next part --------------
An HTML attachment was scrubbed…
URL: http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121019/1b2b44df/attachment-0001.htm
-------------- next part --------------
A non-text attachment was scrubbed…
Name: Line geometry.jpg
Type: image/jpeg
Size: 130770 bytes
Desc: Line geometry.jpg
Url : http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121019/1b2b44df/attachment-0002.jpg
-------------- next part --------------
A non-text attachment was scrubbed…
Name: geometry1.jpg
Type: image/jpeg
Size: 113160 bytes
Desc: geometry1.jpg
Url : http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121019/1b2b44df/attachment-0003.jpg

2

Hi Rimon,
If your need is only on assign boundary condition, I encourage you to
work with layers, setting your geometry in appropriated layers, and
after, assign conditions over layers, instate of over entities.

Anyway, there are two tools inside GiD that can be used to “simplify
Geometry”, this function can be access by Geometry-Create-
1) Surfmesh: Elements of mesh must be triangles or quadrilaterals.
Mesh is converted by GiD into a set of surfaces, points and lines. The
geometric definition of surfaces is the mesh itself, but GiD treats them
as truly geometric entities. For example, these surfaces can be used as
the boundary of a volume, and a new mesh can be generated over them.
2) Geometry from mesh*:* This option converts all our mesh model (only
surface mesh, triangles and quadrilateral) to a geometry model,
obtaining a NURBS surfaces based definition. Creates a group of new
layers called “Reconstruction”, inside you will see two new layers: the
first “All Lines And Points” contains lines and point and the second
“Reconstructed Nurbs” the surfaces. If some surface couldn’t be
reconstructed it will appear a third layer called “SurfMeshes Not
Reconstructed” containing the remaining parts converted in Surfmeshes

Regards,
Adrià

El 19/10/2012 15:18, Rimon Muchehiwa escribió:
Is there any way to simplify the given geometry in the attachment. I
have tried collapsing it and rebuilding the surfaces but this has been
unsuccessful. What I am hoping to achieve is a simplified model with
as few surfaces as possible so I can easily assign boundary conditions
and loads. Any help would bee greatly appreciated.
Regards
R.M
### UNIVERSITY OF CAPE TOWN This e-mail is subject to the UCT ICT
policies and e-mail disclaimer published on our website at
http://www.uct.ac.za/about/policies/emaildisclaimer/ or obtainable
from +27 21 650 9111. This e-mail is intended only for the person(s)
to whom it is addressed. If the e-mail has reached you in error,
please notify the author. If you are not the intended recipient of the
e-mail you may not use, disclose, copy, redirect or print the content.
If this e-mail is not related to the business of UCT it is sent by the
sender in the sender’s individual capacity. ###

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
_http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist_
-------------- next part --------------
An HTML attachment was scrubbed…
URL: http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121022/76a948a8/attachment.htm

3

Bonjour,

Je suis en déplacement, retour le 24 octobre. Vos emails seront lus dans la mesure du possible.

En cas d’urgence, adressez votre message ici : info at mp-ingenieurs.ch

Eric Tonicello

=======================

Hi !

I’m out of office and I’ll be back on Wednesday October 24th.

Emails will possibly be read during this period.

In case of urgent matter, please send your message here : info at mp-ingenieurs.ch

Eric Tonicello

4

Dear GiD List,

I’m working on GiD 10.0 on Linux (Precise) and I have issues when generating a quadratic mesh.
I generate a quadratic mesh and save it. When I re-open the GiD file, the mesh is completely distorted and wrong.
Linear meshes work fine.

I had no issues on previous Linux versions.

Is this a known issue?

I don’t have the problem anymore on GiD11.0, but don’t have a license for that version.

Thank you very much for your support.

Best regards,
Axelle

5

It’s not a known problem, it the first time that I heard about it,

In any case we are not going to fix and release more 10.x versions,
if it works well with GiD 11.x maybe it also works with the last 10.2.1d
developer version (using your current 10.x licence)
dowload it at http://www.gidhome.com/download/developer-versions

If the problem appear again with GiD 11.x let us know to fix it.

Enrique Escolano

----- Original Message -----
From: “Vire, Axelle” a.vire at imperial.ac.uk
To: gidlist at listas.cimne.upc.edu
Sent: Wednesday, October 24, 2012 6:00 PM
Subject: [GiDlist] Quadratic mesh on Ubuntu and GiD10.0


Dear GiD List,
I’m working on GiD 10.0 on Linux (Precise) and I have issues when
generating a quadratic mesh.
I generate a quadratic mesh and save it. When I re-open the GiD file, the
mesh is completely distorted and wrong.
Linear meshes work fine.
I had no issues on previous Linux versions.
Is this a known issue?
I don’t have the problem anymore on GiD11.0, but don’t have a license for
that version.
Thank you very much for your support.
Best regards,
Axelle

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
_http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist_

6

Dear Enrique,

It seems fine with 10.2.1d.

Many thanks for your help.
Axelle


On 25 oct. 2012, at 11:07, Enrique Escolano wrote:

It’s not a known problem, it the first time that I heard about it,
In any case we are not going to fix and release more 10.x versions,
if it works well with GiD 11.x maybe it also works with the last 10.2.1d
developer version (using your current 10.x licence)
_dowload it at http://www.gidhome.com/download/developer-versions_
If the problem appear again with GiD 11.x let us know to fix it.
Enrique Escolano
----- Original Message -----
From: “Vire, Axelle” a.vire at imperial.ac.uk
To: gidlist at listas.cimne.upc.edu
Sent: Wednesday, October 24, 2012 6:00 PM
Subject: [GiDlist] Quadratic mesh on Ubuntu and GiD10.0

Dear GiD List,
I’m working on GiD 10.0 on Linux (Precise) and I have issues when
generating a quadratic mesh.
I generate a quadratic mesh and save it. When I re-open the GiD file, the
mesh is completely distorted and wrong.
Linear meshes work fine.
I had no issues on previous Linux versions.
Is this a known issue?
I don’t have the problem anymore on GiD11.0, but don’t have a license for
that version.
Thank you very much for your support.
Best regards,
Axelle

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
_http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist_

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
_http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist_

7

Hi

I have GiD results in binary format from two solvers. To validate the
solver, I want to check if the two solvers produced the same results.
Therefore I want to dump the results in each file and compare it node by
node. Can you give me some hints to do that in GiD?

BR
Giang Bui
-------------- next part --------------
An HTML attachment was scrubbed…
URL: http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121107/723b7cb9/attachment.htm

8

1- If do you feel more comfortable, you can convert the binary files to ASCII from GiD (Files-Export-Post information-ASCII files…),
and then compare the text files if they are very similar with some diff tool (e.g. tkdiff, Ramdebugger of GiD include this tool)

2- Probably both results will be very differents, then a text comparison will be unusable.
You can read both results in GiD (must have different result name, to not overwrite), and create a new derivated result sustracting them.
Window-Create result… select the first result, the operator - , the second result, provide the new name and press Apply,
then you will be able to graphically see how different the results are.

Enrique Escolano
----- Original Message -----
From: Hoang Giang Bui
To: gidlist at listas.cimne.upc.edu
Sent: Wednesday, November 07, 2012 4:11 PM
Subject: [GiDlist] compare results of gid post files


Hi

I have GiD results in binary format from two solvers. To validate the solver, I want to check if the two solvers produced the same results. Therefore I want to dump the results in each file and compare it node by node. Can you give me some hints to do that in GiD?

BR
Giang Bui



\


\

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist
-------------- next part --------------
An HTML attachment was scrubbed…
URL: http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121107/3754ba4a/attachment.htm

9

On 11/07/12 17:22, Enrique Escolano wrote:
1- If do you feel more comfortable, you can convert the binary files
to ASCII from GiD (Files-Export-Post information-ASCII files…),
and then compare the text files if they are very similar with some
diff tool (e.g. tkdiff, Ramdebugger of GiD include this tool)
2- Probably both results will be very differents, then a text
comparison will be unusable.
You can read both results in GiD (must have different result name,
to not overwrite), and create a new derivated result sustracting them.
Window-Create result… select the first result, the operator - ,
the second result, provide the new name and press Apply,
then you will be able to graphically see how different the results
are.
Enrique Escolano
----- Original Message -----
From: Hoang Giang Bui mailto:hgbk2008 at gmail.com
To: gidlist at listas.cimne.upc.edu
mailto:gidlist at listas.cimne.upc.edu
Sent: Wednesday, November 07, 2012 4:11 PM
Subject: [GiDlist] compare results of gid post files
Hi
I have GiD results in binary format from two solvers. To validate
the solver, I want to check if the two solvers produced the same
results. Therefore I want to dump the results in each file and
compare it node by node. Can you give me some hints to do that in GiD?
BR
Giang Bui
------------------------------------------------------------------------

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
_http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist_

\

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
_http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist_
Hi

I realized that dumping data from binary file is a viable solution for
my case although it’s not easy to do in the first place. Dumping data
from ASCII file was not good due to some reasons:

  • A simulation generates some hundreds data files (depends on number
    of steps). Each data files is then compared with a reference result from
    simulation with a reliable solver.
  • ASCII data file grows bigger when the number of d.o.f is large. For
    instance, one of my problem with 5000 nodes generates 14MB for one ASCII
    data file. Performs numerous simulation of large scale may consume
    tremendous memory.

I want an automatic mechanism to compare the results since there were so
many to do it manually. As you mentioned, compare text files using diff
tool does not make sense at all.

BR
Giang Bui

-------------- next part --------------
An HTML attachment was scrubbed…
URL: http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121115/3f3d61d9/attachment.htm

10

I realized that dumping data from binary file is a viable solution for my case although it’s not easy to do in the first place. Dumping data from ASCII file was not good due to some reasons:

  • A simulation generates some hundreds data files (depends on number of steps). Each data files is then compared with a reference result from simulation with a reliable solver.
  • ASCII data file grows bigger when the number of d.o.f is large. For instance, one of my problem with 5000 nodes generates 14MB for one ASCII data file. Performs numerous simulation of large scale may consume tremendous memory.

I want an automatic mechanism to compare the results since there were so many to do it manually. As you mentioned, compare text files using diff tool does not make sense at all.

Off course an ASCII is bigger that its equivalent binary compressed file, and probably a little slower to be read

  • About handling hundreds of files: you can write all time steps in a single file.

How do you define that two results are “equivalent”? It is your own decision to consider similar results as equal.
Numerical values and the kind of results written by two different programs will be different, (e.g. how to automatically map results with different names)

You must write your own program to compare files as you want, the ASCII format allows you to read it easily than a complicated binary compressed format.

you can also run GiD in batch mode to read results, and write some representative data to be compared (e.g. an image of a contour fill, a graph of some point, some integral value, etc)

----- Original Message -----
From: Hoang Giang Bui
To: gidlist at listas.cimne.upc.edu
Sent: Thursday, November 15, 2012 5:37 PM
Subject: Re: [GiDlist] compare results of gid post files


On 11/07/12 17:22, Enrique Escolano wrote:

1- If do you feel more comfortable, you can convert the binary files to ASCII from GiD (Files-Export-Post information-ASCII files…),
and then compare the text files if they are very similar with some diff tool (e.g. tkdiff, Ramdebugger of GiD include this tool)

2- Probably both results will be very differents, then a text comparison will be unusable.
You can read both results in GiD (must have different result name, to not overwrite), and create a new derivated result sustracting them.
Window-Create result… select the first result, the operator - , the second result, provide the new name and press Apply,
then you will be able to graphically see how different the results are.

Enrique Escolano
----- Original Message -----
From: Hoang Giang Bui
To: gidlist at listas.cimne.upc.edu
Sent: Wednesday, November 07, 2012 4:11 PM
Subject: [GiDlist] compare results of gid post files


Hi

I have GiD results in binary format from two solvers. To validate the solver, I want to check if the two solvers produced the same results. Therefore I want to dump the results in each file and compare it node by node. Can you give me some hints to do that in GiD?

BR
Giang Bui



\

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist



\

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist

Hi

I realized that dumping data from binary file is a viable solution for my case although it’s not easy to do in the first place. Dumping data from ASCII file was not good due to some reasons:

  • A simulation generates some hundreds data files (depends on number of steps). Each data files is then compared with a reference result from simulation with a reliable solver.
  • ASCII data file grows bigger when the number of d.o.f is large. For instance, one of my problem with 5000 nodes generates 14MB for one ASCII data file. Performs numerous simulation of large scale may consume tremendous memory.

I want an automatic mechanism to compare the results since there were so many to do it manually. As you mentioned, compare text files using diff tool does not make sense at all.

BR
Giang Bui



\


\

GiDlist mailing list
GiDlist at listas.cimne.upc.edu
http://listas.cimne.upc.edu/cgi-bin/mailman/listinfo/gidlist
-------------- next part --------------
An HTML attachment was scrubbed…
URL: http://listas.cimne.upc.edu/pipermail/gidlist/attachments/20121115/f1c43b3a/attachment.htm