<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
 <channel>
  <atom:link href="http://www.comsol.no/community/forums/geometry/rss/feed.rss" rel="self" type="application/rss+xml"/>
  <title>Geometry - Latest Discussions - COMSOL Forums</title>
  <link>http://www.comsol.no/community/forums/</link>
  <description>Most recent forum discussions</description>
  <pubDate>Tue, 22 May 2012 08:49:33 +0000</pubDate>
  <image>
   <title>Geometry - Latest Discussions - COMSOL Forums</title>
   <url>http://www.comsol.no/shared/images/logos/comsol_logo.gif</url>
   <link>http://www.comsol.no/community/forums/</link>
  </image>
  <item>
   <title>How to increase the resolution of a geometry? (in my case a circle).</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28784/</link>
   <description>&lt;br /&gt;&#13;
Hej,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I have a solid block where I want to substract a really small part from the top surface (tries to be a sagging of 20 um). I am using a the difference booelan operation to do it.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
The volume to subract is a cylinder. My problem is that the cylinder is huge compared to the solid block. The circle (which I extrude to obtain the cylinder) is defined by a defined number of segments (or points).&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
In other software's there is the option to increase this number of points (segments) when defining a circle. I think it is called &amp;quot;resolution&amp;quot;.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
How can I do it in Comsol?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
(I was also thinking about defining a parabolic curve and extrude it, will I have the same problem?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thanks in advance.</description>
   <pubDate>Tue, 22 May 2012 08:49:33 +0000</pubDate>
   <guid isPermaLink="false">.1337676573.28784</guid>
  </item>
  <item>
   <title>How can I build  the structure with sinusoidal-shape interface in comsol?</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28756/</link>
   <description>Hi! I am a  novice and have a question that  how to construct a geometry object with the sinusoidal shape interface,e.g.sinusoidal grating in comsol 3.5a? Thank you very much.Here is my email:hujigang@hfut.edu.cn.</description>
   <pubDate>Mon, 21 May 2012 03:10:46 +0000</pubDate>
   <guid isPermaLink="false">.1337569846.28756</guid>
  </item>
  <item>
   <title>how to realize stent with comsol?</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28578/</link>
   <description>Hello to everybody,&lt;br /&gt;&#13;
I'm a student of Politecnico of Torino and i'm trying to realize a stent geometry with comsol (4.2a version).&lt;br /&gt;&#13;
Is there a method to realize it without using other cad as solidwork or autocad and so on?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Can someone helps me? I hope yes&lt;br /&gt;&#13;
I would be very happy for your help&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thanks for attention&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Michele</description>
   <pubDate>Mon, 14 May 2012 12:38:20 +0000</pubDate>
   <guid isPermaLink="false">.1336999100.28578</guid>
  </item>
  <item>
   <title>saving geometry as .mphbin file</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28550/</link>
   <description>Hi,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
New user.  How do save a geometry I created as '.mphbin' (or any other format) for future use?  Thanks.</description>
   <pubDate>Fri, 11 May 2012 12:01:33 +0000</pubDate>
   <guid isPermaLink="false">.1336737693.28550</guid>
  </item>
  <item>
   <title>Temperature Sensor Model</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28467/</link>
   <description>Hello,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I am designing a temperature sensor using an interdigitated capacitor.  Using the electrostatics modules, I was successfully able to design the sensor with the starting capacitance value I wanted.  However, regarding changing capacitance with temperature is where I run into problems.  Thus far, I have been using the joule heating and thermal expansion module.  Can anyone tell me if this is the correct module for my goal?  &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I have had some measure of success using it.  However I find that when I remove the surrounding rectangle with air associated with it in the positive z direction and silicon in the negative z direction, the study completes, but with that rectangle, I get an error of singular matrix/solution does not converge and cannot complete the study.  Why is this? &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Also, without the surrounding rectangle, and without the proper dielectric in place, it seems that I cannot calculate capacitance correctly, and the values are all very small, approaching zero.   &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I have attached my model.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I would greatly appreciate help understanding this, thank you.</description>
   <pubDate>Tue, 08 May 2012 21:11:51 +0000</pubDate>
   <guid isPermaLink="false">.1336511511.28467</guid>
  </item>
  <item>
   <title>Hiding geometry boundaries in graphics window?</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28459/</link>
   <description>Hi,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I have a model geometry that includes some sub-domains. I want to show the results from these domains, but I don't want their boundaries to be visible in the graphics window, so visually the model looks like only one domian. Does anyone know if and how this can be done? </description>
   <pubDate>Tue, 08 May 2012 12:46:42 +0000</pubDate>
   <guid isPermaLink="false">.1336481202.28459</guid>
  </item>
  <item>
   <title>electro-thermal analysis of micro-heeater</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28402/</link>
   <description>sir,when i tried to simulate it it is showing some errror, and i could not  find the error .can u please help me to solve this structre..</description>
   <pubDate>Mon, 07 May 2012 04:07:32 +0000</pubDate>
   <guid isPermaLink="false">.1336363652.28402</guid>
  </item>
  <item>
   <title> Case distinctive periodic boundary condition</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28364/</link>
   <description>Hello,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I need a case distinctive periodic boundary condition (magnetic vector potential), depending on atan2(y,x).&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I thought about using a simple piecewise function for the magnetic vector potential on the wanted boundary. But I have no idea, how to tell COMSOL that I want the magnetic vector potential from another specific boundary as source (except using the COMSOL- given periodic boundary conditions). &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
magnetic vector potential on destination boundary should be the piecewise function:&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
argument __________________  expression&lt;br /&gt;&#13;
0 - pi/2 ____________________ Az  (on source boundary)&lt;br /&gt;&#13;
pi/2 - pi  ___________________  -Az  (on source boundary)&lt;br /&gt;&#13;
etc.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
The physical problem behind this is a simulation of a fourth electrical rotating machine.&lt;br /&gt;&#13;
When the rotor rotates out of the stator, the continuity condition changes to anti.&lt;br /&gt;&#13;
So the destination boundary is rotating all the time with the rotor and the source boundary stays still.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
It would be useful if there was a way to use Parameterized curves from the Post Processor as input funtions.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Any help?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thanks in advance</description>
   <pubDate>Fri, 04 May 2012 13:49:46 +0000</pubDate>
   <guid isPermaLink="false">.1336139386.28364</guid>
  </item>
  <item>
   <title>How to form Jacobian</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28341/</link>
   <description>I have measurement nodes in my geometry,I want to form a Jacobian matrix which has change in measurement value with respect to change in coefficient of pde at individual node position.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Please help me solving the problem. </description>
   <pubDate>Thu, 03 May 2012 18:34:50 +0000</pubDate>
   <guid isPermaLink="false">.1336070090.28341</guid>
  </item>
  <item>
   <title>Mesh query regarding open boundary FSI simulation</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28331/</link>
   <description>I've got a model in FSI which consists of two inlets to a central reservoir with a membrane on top. &lt;br /&gt;&#13;
When I use normal mesh settings, the mesh quality at the inlets is horrible. &lt;br /&gt;&#13;
Am I forgetting a boundary condition? Or is there something I can do regarding the mesh?&lt;br /&gt;&#13;
I apply simple open boundary conditions to the inlets. &lt;br /&gt;&#13;
See attached picture, I've filtered to show only the worst mesh quality, &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
any idea how to improve this mesh?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
</description>
   <pubDate>Thu, 03 May 2012 12:02:03 +0000</pubDate>
   <guid isPermaLink="false">.1336046523.28331</guid>
  </item>
  <item>
   <title>can comsol support multilayer shell?</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28316/</link>
   <description>Dear all,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I am trying to model a membrane using shell element because thickness is quite thin compared to length and width. But the membrane is lamated with three to five different materials and thickness at different location. It seems impossible to use 3D solid model  that will lead  too many mesh elements. Do anyone know how to do multilayer shell in comsol? </description>
   <pubDate>Thu, 03 May 2012 05:31:22 +0000</pubDate>
   <guid isPermaLink="false">.1336023082.28316</guid>
  </item>
  <item>
   <title>Integration over a filtered volume</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28254/</link>
   <description>Dear Support,&lt;br /&gt;&#13;
I kindly ask you to help me to solve the following problem.&lt;br /&gt;&#13;
I am able to plot a user-defined quantity with an element filter in a 3D domain: the logical expression I use for the filter of this quantity is :&amp;quot;quantity&amp;gt;0&amp;quot;.&lt;br /&gt;&#13;
Then, my 3D domain is &amp;quot;restricted&amp;quot; to the only elements where this quantity is larger than zero.&lt;br /&gt;&#13;
My question is: how can I do a volume integration in this &amp;quot;restricted&amp;quot; 3D domain?&lt;br /&gt;&#13;
I am not able to identify this restricted 3D domain as a specific entity, but I am only able to plot it.&lt;br /&gt;&#13;
Please, help me to fix this problem.&lt;br /&gt;&#13;
With my best regards,&lt;br /&gt;&#13;
Alessandro</description>
   <pubDate>Mon, 30 Apr 2012 15:19:58 +0000</pubDate>
   <guid isPermaLink="false">.1335799198.28254</guid>
  </item>
  <item>
   <title>STL Geometry &amp; Meshing</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/28251/</link>
   <description>I have been working on importing a geometry from an STL file; it's a thing 3-D object, where the length and width are on the order of 10^5 m and the thickness is 10^3 m. I originally had problems where the finalized geometry contained strange overlaps that came from some interpolation within COMSOL. I have finally figured out how to import without these problems.* Now I have another concern.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
The intent of this is to build a Java file that I can run within a Fortran model with a variety of geometries, so I ideally want something that is very general and not gui-based. The import geometries will always have a top surface, a bottom surface, and some side faces that are perpendicular to the y-axis. &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
If you look at the model, you can see that the imported geometry still contains all of the faces defined within the stl file. I cannot get rids of this segmentation, which is the problem now for meshing and for setting boundary conditions.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Is there a way to combine all of these segments together? They are all treated as one domain within the geometry. I have tried converting to solid and a variety of other processes dealing with the geometry. But it seems that when I go to mesh, the mesh is over each small unit and really takes quite a bit of time to get going. &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Second question, is there a way to write in the face number when I create the geometry. Since I want something that can be generalized, but I will always know the number of faces/facets, numbering them would allow me to easily group them into the surface necessary for the boundary conditions that I want to apply.   &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Unfortunately, my file is too large. I'll try to figure this out so that i can attach it. Probably clearing the mesh will solve that problem.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
*Just for your information, I have attempted a variety of other methods to bring in this geometry, including an interpolation function (from a file) with a block &amp;amp; parametric surfaces for the geometry. I never got this to work successfully to make something 3D, though it worked for either the top or bottom just fine (see the mention of the interpolation in the first paragraph). I've also gone through many, many permutations of stl files, and this seems to be the best. </description>
   <pubDate>Mon, 30 Apr 2012 13:30:18 +0000</pubDate>
   <guid isPermaLink="false">.1335792618.28251</guid>
  </item>
  <item>
   <title>High pressure around rotating domain boundary</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/27791/</link>
   <description>Hi, and thanks for reading.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I am modelling an impeller in a turbulent vessel (2d for now). I have an ALE moving mesh rotating with the impeller. When I compute the model the areas of highest pressure and velocity are at the boundary between the moving and the fixed mesh - regardless how large I set the rotating domain, this pattern always follows it.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Has anyone got any advice on what I am doing wrong? The file is 180MB, but if it would help I can attach it.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Many thanks,&lt;br /&gt;&#13;
Andrew</description>
   <pubDate>Fri, 13 Apr 2012 10:10:27 +0000</pubDate>
   <guid isPermaLink="false">.1334311827.27791</guid>
  </item>
  <item>
   <title>Extruding to 3d</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/20215/</link>
   <description>Hi,&lt;br /&gt;&#13;
2 questions: &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
1)  I am creating a ring in 2d, and it seems perfectly circular, but when I extrude it to 3d, it suddenly gets slightly jagged sides and becomes a sort of polygon. I tried to fillet it in 2d and then extrude, but it makes no difference, the 3d still has edges. Anyone know how to fix this?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
2)  Is there any way to put a 2d face on a 3d object? For example, on a cylinder, I would like to put a small circle on its top face to model a laser's heat deposition. I think this would let me impose heat transfer conditions on the boundary. Other suggestions are also welcome.&lt;br /&gt;&#13;
Thanks</description>
   <pubDate>Wed, 29 Jun 2011 18:00:15 +0000</pubDate>
   <guid isPermaLink="false">.1309370415.20215</guid>
  </item>
  <item>
   <title>import complex 3D geometry internal error code 900 / boolean operation failed</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/19033/</link>
   <description>Hi everyone,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I am trying to import a complex 3D geometry into COMSOL 4.0a.  I've created it using CATIA (R19). The file format to be imported is .stp. Also, I do have access to the CAD import tool. The model consists of four domains that, once united, have a cylindrical shape. The interfaces between the domains, however, are very complex and irregular. I can import each domain separately (and form a union for each domain individually). However, once i try to form a union between all four domains I obtain the following error message: &amp;quot;Boolean operation failed&amp;quot;. In a different thread I read that this might be due to the repair tolerance. Hence, I played with the import tolerances. Now, however, I obtain another error message: &amp;quot;Internal error in geometry processing internal error code: 900&amp;quot;. &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
My questions:&lt;br /&gt;&#13;
===========&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
1. I am dealing with a coupled fluid structure problem (using &amp;quot;acsl&amp;quot; toolbox), is it 100% necessary to form a union or will an assembly do the trick as well?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
2. If I need to form a union, why do I get the outlined error messages and what can I do to fix the problem (I am almost 100% sure that my geometry is correct and that the domain interfaces are error free)&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thanks you very much in advance&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Martin</description>
   <pubDate>Mon, 06 Jun 2011 13:10:25 +0000</pubDate>
   <guid isPermaLink="false">.1307365825.19033</guid>
  </item>
  <item>
   <title>Create arbitrary distribution of circles  via Livelink for Matlab</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/18356/</link>
   <description>Hallo Comsol Community&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
First of all, since i am very new to this discussion forum please excuse any 'unclearness' in the way i will try to express my problem.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I am currently using the matlab livelink feature of comsol 4.1 for creating a 2D geometry that consists of arbitrarily posistioned circles.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
The way i am trying to do this is by the following matlab script:&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
model = ModelUtil.create('Model');&lt;br /&gt;&#13;
geom1 = model.geom.create('geom1',2);&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
index=1&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
x=randi([-100 100],30,1)&lt;br /&gt;&#13;
y=randi([-100 100],30,1)&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
index&lt;br /&gt;&#13;
for i=1:1:size(x,1)&lt;br /&gt;&#13;
        c(i)= geom2.feature.create((strcat('c',i)),'Circle');&lt;br /&gt;&#13;
        c(i).set('r',1);&lt;br /&gt;&#13;
        c(i).set('pos',[x(i) y(i)]);&lt;br /&gt;&#13;
     geom2.run;&lt;br /&gt;&#13;
       index=index+1&lt;br /&gt;&#13;
end&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
geom2.run;&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
mphgeom(model,'geom2');&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
1) I really do not understand why each geometric element has to be under a different variable name.Right now i am incorporating all my circles inside the matrix called c and i do not know if this is correct.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
2) If i run the same script with &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
x=randi([-100 100],40,1)&lt;br /&gt;&#13;
y=randi([-100 100],40,1)  instead of&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
x=randi([-100 100],30,1)&lt;br /&gt;&#13;
y=randi([-100 100],30,1)  &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
then i get an error on the geom2.run function that i also cannot explain.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
</description>
   <pubDate>Tue, 17 May 2011 19:03:08 +0000</pubDate>
   <guid isPermaLink="false">.1305658988.18356</guid>
  </item>
  <item>
   <title>How to define two surface reactions having same flux or reaction rate?</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/18143/</link>
   <description>Hi Everyone,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
     I am using Comsol 4.1 doing a simple simulation on a 1D geometry. There two simple reactions are happening on both end of the line. On the left, the reaction is A to B. On the right, the reaction is B to A. The reaction rate constant is different though. Whenever reaction A to B happens on the left, the reaction B to A will happen on the right. I just simply using the Transport in Diluted species to have the reaction happen. However, I do not know how to make the connection between these two reactions.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
This has bothered me for a few day. I am new to Comsol. Anyone can help me out with this? &lt;br /&gt;&#13;
Really appreciate.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thank you in advance,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Jay&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
</description>
   <pubDate>Thu, 12 May 2011 13:10:40 +0000</pubDate>
   <guid isPermaLink="false">.1305205840.18143</guid>
  </item>
  <item>
   <title>Modelling nanotube</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/17987/</link>
   <description>Dear all,&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I would like to simulate the influence of fiber orientation of CNT (Carbon Nano Tube) on some properties, like thermal conductivity and mechanical properties.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Please feel free to contact me if further information is needed. Any suggestions will be appreciated.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thank you in advance for your time and your consideration.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Best regards,</description>
   <pubDate>Mon, 09 May 2011 00:37:05 +0000</pubDate>
   <guid isPermaLink="false">.1304901425.17987</guid>
  </item>
  <item>
   <title>Modelling solar radiation / black body radiator?</title>
   <link>http://www.comsol.no/community/forums/geometry/thread/13396/</link>
   <description>Hello everybody.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I would like to model accurately the thermal behavior of a flat plate solar collector.&lt;br /&gt;&#13;
(from the top to the bottom : a glass cover, an air layer, a radiation absorbeur and a thermal insulation).&lt;br /&gt;&#13;
 &lt;br /&gt;&#13;
I did a first model (2D and 3D) using an input energy flux on the top boundary of the glass cover. But I'm not really satisfied with this configuration : &lt;br /&gt;&#13;
It seems that in this case, Comsol is working in conduction/convection only, and did not take into account that the incident flux type is radiative. I would like to traduce the fact that the glass cover is opaque to the IR radiation and transparent to the UV radiation.&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
I've tried to model a black body radiator with the surface to surface radiation module (heat transfer)... but it didn't work. &lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Does someone know how I could manage that?&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Thanks a lot for your answers&lt;br /&gt;&#13;
&lt;br /&gt;&#13;
Fabrice&lt;br /&gt;&#13;
</description>
   <pubDate>Tue, 25 Jan 2011 16:24:03 +0000</pubDate>
   <guid isPermaLink="false">.1295972643.13396</guid>
  </item>
 </channel>
</rss>

