Geometry distance between two points

sys999 · January 16, 2011, 14:11

Hello,

unfortunately I have a big problem with the distance between two points.

For example I create these two points (0.1 0 0) and (0.4 0 0). When I'm now measuring the distance between them Icem shows me "0.30000000447034".

When I'm importing a geometry Icem changes the distances too.

My mesh should be very accurate.

Can anybody help me? Thx...

la7low · January 16, 2011, 21:57

Hi,
Ive never made mesh for micro flows as probably this is your case (or almost), right?
but I think you should bother with settings/model in Icem and decrease the tolerances there. Maybe thats for just surfaces and will not help at all, but I am quite sure you need to change settings for such cases. I recommend you to look at the help about settings...and also the import settings if there are any about those..
Also, the measure distance tool (ruler icon) might take the point where you clicked at and not the actual coordinate of the point.

Funny though that even 1e-7 order of accuracy is not enough for you. How many cells do you want to squeeze on a 0.3 long edge?

sys999 · January 17, 2011, 04:12

I have to mesh a given geometry - a bladed disk - for a Modal FEA in Ansys. For that the mesh has to be rotationally symmetric.
I've thought the units in Icem are meters? Because the model I want to import is about 0,9 meters long.

la7low · January 17, 2011, 15:44

The units can be whatever in Icem, you will specify it later in the solver: e.g. Fluent-scale. Or maybe when you output the mesh to cerain formats, you can specify the unit of length you used for the geometry/mesh.

PSYMN · January 17, 2011, 17:44

You can always scale the model... Instead of working in meters or millimeters, you could work in nano meters, etc...

sys999 · January 18, 2011, 07:43

The point is that i need a accurate rotationally symmetric mesh.
If I mesh a disk in this way in ICEM and export it to Workbench (Fe Modeler -> Modal Analyses) the Eigenvalues aren't the same as in a perfect rotationally symmetric mesh which I can create in Ansys Mechanical.
Though the mesh created in Icem seems good to me...
That means that Icem makes a somehow imprecise mesh.

What's the reason for this?
Is it possible to make it perfectly rotationally symmetric in Icem?

picture of the disk:
http://s7.directupload.net/images/110118/23grm4ex.png

PSYMN · January 30, 2011, 18:18

Interesting... I would like to investigate this further. Perhaps I can get your basic example file? I wonder if it matters how you got to FE modeler. What middle format did you use? Did you do the periodicity check in ICEM CFD?

January 16, 2011, 14:11	Geometry distance between two points	#1
sys999 New Member Bla Blubb Join Date: Dec 2010 Posts: 11 Rep Power: 15	Hello, unfortunately I have a big problem with the distance between two points. For example I create these two points (0.1 0 0) and (0.4 0 0). When I'm now measuring the distance between them Icem shows me "0.30000000447034". When I'm importing a geometry Icem changes the distances too. My mesh should be very accurate. Can anybody help me? Thx...

January 16, 2011, 21:57	micro mesh	#2
la7low Member Join Date: May 2010 Posts: 44 Rep Power: 15	Hi, Ive never made mesh for micro flows as probably this is your case (or almost), right? but I think you should bother with settings/model in Icem and decrease the tolerances there. Maybe thats for just surfaces and will not help at all, but I am quite sure you need to change settings for such cases. I recommend you to look at the help about settings...and also the import settings if there are any about those.. Also, the measure distance tool (ruler icon) might take the point where you clicked at and not the actual coordinate of the point. Funny though that even 1e-7 order of accuracy is not enough for you. How many cells do you want to squeeze on a 0.3 long edge?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Create geometry from mesh points in ICEM-CFD	David Arthur	Main CFD Forum	3	October 25, 2014 12:19
Produce geometry from points	ashtonJ	ANSYS Meshing & Geometry	5	November 30, 2010 15:27
[blockMesh] BlockMeshmergePatchPairs	hjasak	OpenFOAM Meshing & Mesh Conversion	11	August 15, 2008 07:36
vitual _ real	deneb	FLUENT	3	January 22, 2007 04:31
Adaptive optimization of geometry	Jan Rusås	CFX	1	September 16, 2000 02:38

January 17, 2011, 04:12		#3
sys999 New Member Bla Blubb Join Date: Dec 2010 Posts: 11 Rep Power: 15	I have to mesh a given geometry - a bladed disk - for a Modal FEA in Ansys. For that the mesh has to be rotationally symmetric. I've thought the units in Icem are meters? Because the model I want to import is about 0,9 meters long.

January 17, 2011, 15:44		#4
la7low Member Join Date: May 2010 Posts: 44 Rep Power: 15	The units can be whatever in Icem, you will specify it later in the solver: e.g. Fluent-scale. Or maybe when you output the mesh to cerain formats, you can specify the unit of length you used for the geometry/mesh.

January 17, 2011, 17:44		#5
PSYMN Senior Member Simon Pereira Join Date: Mar 2009 Location: Ann Arbor, MI Posts: 2,663 Blog Entries: 1 Rep Power: 47	You can always scale the model... Instead of working in meters or millimeters, you could work in nano meters, etc...

January 18, 2011, 07:43		#6
sys999 New Member Bla Blubb Join Date: Dec 2010 Posts: 11 Rep Power: 15	The point is that i need a accurate rotationally symmetric mesh. If I mesh a disk in this way in ICEM and export it to Workbench (Fe Modeler -> Modal Analyses) the Eigenvalues aren't the same as in a perfect rotationally symmetric mesh which I can create in Ansys Mechanical. Though the mesh created in Icem seems good to me... That means that Icem makes a somehow imprecise mesh. What's the reason for this? Is it possible to make it perfectly rotationally symmetric in Icem? picture of the disk: http://s7.directupload.net/images/110118/23grm4ex.png

January 30, 2011, 18:18		#7
PSYMN Senior Member Simon Pereira Join Date: Mar 2009 Location: Ann Arbor, MI Posts: 2,663 Blog Entries: 1 Rep Power: 47	Interesting... I would like to investigate this further. Perhaps I can get your basic example file? I wonder if it matters how you got to FE modeler. What middle format did you use? Did you do the periodicity check in ICEM CFD?