- Forum
- Public Forums
- Finite Element Modelling - Code_Aster
- Contacts Between Parts in thermal analysis
Contacts Between Parts in thermal analysis
- BRUNEAUX Jerome
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
12 years 1 month ago #6913
by BRUNEAUX Jerome
Contacts Between Parts in thermal analysis was created by BRUNEAUX Jerome
Hi,
I'm currently doing a thermal analysis on a product which is made of 3 parts (a case, a PCB and an electronic component).
The case contains the PCB (almost cover all the surface) and the electronic component is directly in contact with the case through a hole in the PCB (see the attached picture)
Using Salome, I've made a mesh (from a compound geometry object which contains the three parts), then created groups from geometry to be able to set materials and temperatures of the electronic component.
To simplify things, I want the electronic component and the case to use the same material (Aluminium) and the PCB to use an other material.
The PCB and the electronic component are both in contact with the case, on the same face.
I've added a LIAISON_MAIL command to create the contact (otherwise aster fails with 'pivot presque nul' errors) :
The problem is that it seems the LIAISON_MAIL makes the material definition not working (I can"t see a difference between the PCB which have a very low lambda and the aluminium with a high lambda).
Does anyone have a clue of what I'm doing wrong?
Is there a better way to create contact between the case and the other parts ?
I've attached the comm file and a picture of the geometry. (Well, it seems I can't attach the files, so here is a link to these files :
dl.free.fr/kIAm6w9sO
)
If required, I can share the whole project (.hdf, and all the required files). I'm running the Salome_Meca 2013.1 version.
Regards, Jerome
I'm currently doing a thermal analysis on a product which is made of 3 parts (a case, a PCB and an electronic component).
The case contains the PCB (almost cover all the surface) and the electronic component is directly in contact with the case through a hole in the PCB (see the attached picture)
Using Salome, I've made a mesh (from a compound geometry object which contains the three parts), then created groups from geometry to be able to set materials and temperatures of the electronic component.
To simplify things, I want the electronic component and the case to use the same material (Aluminium) and the PCB to use an other material.
The PCB and the electronic component are both in contact with the case, on the same face.
I've added a LIAISON_MAIL command to create the contact (otherwise aster fails with 'pivot presque nul' errors) :
# Names of the groups has been changed to be more explicit for the topic
# The PCB and COMPONENT groups are volumes and the CASE_CONTACT_FACE is a face.
LIAISON_MAIL=(
_F(GROUP_MA_MAIT=('PCB','COMPONENT'),
GROUP_MA_ESCL='CASE_CONTACT_FACE',
),
),The problem is that it seems the LIAISON_MAIL makes the material definition not working (I can"t see a difference between the PCB which have a very low lambda and the aluminium with a high lambda).
Does anyone have a clue of what I'm doing wrong?
Is there a better way to create contact between the case and the other parts ?
I've attached the comm file and a picture of the geometry. (Well, it seems I can't attach the files, so here is a link to these files :
dl.free.fr/kIAm6w9sO
)
If required, I can share the whole project (.hdf, and all the required files). I'm running the Salome_Meca 2013.1 version.
Regards, Jerome
- Claus
-
- Offline
- Moderator
-
Less
More
- Posts: 670
- Thank you received: 34
12 years 1 month ago #6922
by Claus
Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
Replied by Claus on topic Re: Contacts Between Parts in thermal analysis
Using LIAISON_MAIL should not interfere with material definition. GROUP_MA_MAIT should only contain one vol.group.
Could you attach the .mess file?
If you want a rigid connection between the two parts, a compromise could be to mesh the whole thing as one, and assign materials to volume groups (but LIAISON_MAIL should work)
/C
Could you attach the .mess file?
If you want a rigid connection between the two parts, a compromise could be to mesh the whole thing as one, and assign materials to volume groups (but LIAISON_MAIL should work)
/C
Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
- BRUNEAUX Jerome
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
12 years 1 month ago #6923
by BRUNEAUX Jerome
Replied by BRUNEAUX Jerome on topic Re: Contacts Between Parts in thermal analysis
Hi,
Sorry not to have posted before, I've been working a lot on this problem.
Finally, I've been able to avoid using the LIAISON_MAIL by rebuilding the mesh properly in Salome (made a partition with the three elements, which result in a correct mesh where contacts areas share nodes). That way, I was sure the contact were good.
Then I changed the .comm file to make a temporal analysis and it clearly show what I expected (the heat propagates faster on the case than on the PCB).
This is now OK, but I realize, I moght have to switch to SYRTHES to be able to simulate a heatsink with fan.
Regards, Jerome
Sorry not to have posted before, I've been working a lot on this problem.
Finally, I've been able to avoid using the LIAISON_MAIL by rebuilding the mesh properly in Salome (made a partition with the three elements, which result in a correct mesh where contacts areas share nodes). That way, I was sure the contact were good.
Then I changed the .comm file to make a temporal analysis and it clearly show what I expected (the heat propagates faster on the case than on the PCB).
This is now OK, but I realize, I moght have to switch to SYRTHES to be able to simulate a heatsink with fan.
Regards, Jerome
- Claus
-
- Offline
- Moderator
-
Less
More
- Posts: 670
- Thank you received: 34
12 years 1 month ago #6932
by Claus
Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
Replied by Claus on topic Re: Contacts Between Parts in thermal analysis
Glad you got it sorted.
If you could post your analysis, perhaps along with a diagram of the loads, it would be really useful for new users to look at
/C
If you could post your analysis, perhaps along with a diagram of the loads, it would be really useful for new users to look at
/C
Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
- BRUNEAUX Jerome
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
12 years 1 month ago #6933
by BRUNEAUX Jerome
Replied by BRUNEAUX Jerome on topic Re: Contacts Between Parts in thermal analysis
Hi,
I was already thinking of doing a small tutorial for our company internal use, so if I have time to do it properly, then no problem for me to publish it if there is no 'confidential' material inside it. Otherwise, I will do it with 'hidden' geometry.
I'm also trying to use the Syrthes (and probably the Code_Saturn) softwares with the same geometry and as soon as I've got it working, I might also write a tutorial to make it work (I couldn't find much tutorials for Syrthes).
Regards, Jerome
I was already thinking of doing a small tutorial for our company internal use, so if I have time to do it properly, then no problem for me to publish it if there is no 'confidential' material inside it. Otherwise, I will do it with 'hidden' geometry.
I'm also trying to use the Syrthes (and probably the Code_Saturn) softwares with the same geometry and as soon as I've got it working, I might also write a tutorial to make it work (I couldn't find much tutorials for Syrthes).
Regards, Jerome
Moderators: catux
- Forum
- Public Forums
- Finite Element Modelling - Code_Aster
- Contacts Between Parts in thermal analysis
Time to create page: 0.142 seconds