A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Numerical methods and mathematical models of Elmer
Post Reply
Sheeppowa
Posts: 7
Joined: 19 Jul 2019, 15:54
Antispam: Yes

A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by Sheeppowa »

Good morning everyone,

To train myself to master Elmer, I have in mind, before carrying out simulations for my own work, to find simulation results obtained within the framework of this article : https://link.springer.com/article/10.10 ... 010-0310-7

For those who cannot download it and as I do not think it is allowed to distribute this article for free, I summarize in pictures the modelling (geometry, boundary conditions) as well as the results obtained by the authors.

Their work consists in carrying out an electrical-thermal_mechanical simulation of a system from Gleeble which is a test bench allowing tensile/compression or other tests to be carried out at temperature via a heating by Joule effect.

I am only interested here for the moment in electrical-thermal coupling, if I can already obtain similar results, the mechanical aspect can be considered.

The system consists of two copper jaws that clamp a circular rod of the material to be studied. You will find their modeling under the attached figure.

The authors also indicate the boundary conditions for electrical modelling, they are:
- a current phi_imp on the omega_s_gf2 boundary of one of the jaws
- of an impressed current J_impulse on the omega_s_gf1 boundary of the other jaw
- a condition expressing the imperfect electrical contact between the copper jaws and the cylindrical omega_s_s_sg material

For thermal simulation, they dare:
- An imposed flow q_impulse zero on the surfaces omega_s_g_es, omega_s_g_is of the jaws as well as the surface s_es of the cylindrical bar
- a condition expressing the imperfect thermal contact between the copper jaws and the cylindrical omega_s_s_sg material
- a thermal transfer between the cylindrical material and the environment in omega_s where h_th_eff is equal to 2000 Wm-2K-1

In the same image, you will find the results obtained for the electrical potential, the current density and the distribution of the temperature in the bar between the jaws.

To move on to the part about me, I first used the Salomé software to model the same geometry, allowing me to discern the three bodies as well as each of the external surfaces to apply the boundary conditions, you can find the modeling in UNV file that elmer can read.

As a result, I was able to define the boundary conditions as close as possible to those defined in the article, but I could not define the electrical and thermal boundary conditions at the junction between the jaws and the omega_s_s_g rod.

You will find in a second image file the mesh size used as well as the result obtained for the electrical potential and the current volume. Unfortunately, the temperature shows no sign of change . The.sif file is also available.

I would therefore, after having presented the context and the results ask some questions :

- When adding the unv file to elmer, it shows me 14 surfaces, while the external surfaces are 9, it means that there are 5 more surfaces but I don't know what they refer to, how do I know?

- Concerning the calculation of the temperature, perhaps it is also necessary to change the way the calculation is carried out and not just look at the boundary conditions.

Thank you very much,

Have a good day.
Attachments
Resume_article.png
(335.6 KiB) Not downloaded yet
Resume_values.png
(172.26 KiB) Not downloaded yet
case.sif
(4.33 KiB) Downloaded 219 times
Sheeppowa
Posts: 7
Joined: 19 Jul 2019, 15:54
Antispam: Yes

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by Sheeppowa »

Hello everyone,

So, I manage to find something.

First, Salome' s unit is the meter so we have to add a "coordinate scaling = 1.0e-3" to have the correct dimensions.

The previously inaccessible borders are in fact the contact interface between the copper jaws and the specimen.

The temperature obtained may seem relatively correct, the iso-temperatures obtained take the form of the iso-temperatures obtained in the article, nevertheless the temperatures reached here are not sufficiently high compared to the article.

It would seem that the difference must come from the modeling of the interface between the jaws and the specimen, it would seem logical that this part of the specimen is also subjected to the Joule effect, which is not the case given the results, so this can influence the temperature values obtained...
Attachments
case.sif
File use for modeling
(5.22 KiB) Downloaded 223 times
Data.jpg
Data showing the potential, joule heating, volume current, temperature and iso-temperature
(378.05 KiB) Not downloaded yet
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by kevinarden »

The interface has to share the same nodes, if the mesh is the same size at the interface than you may have coincident nodes that have to be merged. I would investigate the interface to make sure they share the same nodes. Or investigate Chapter 9 of the Elmer Solver manual - Enforcing shared conditions between boundaries.
Sheeppowa
Posts: 7
Joined: 19 Jul 2019, 15:54
Antispam: Yes

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by Sheeppowa »

Hello Kevinarden, thank you for your reply.

I'll check all that, as I know, the interface was made of the two different materials so maybe the nodes are shared.

I will search in that chapter.

Thank you again for your reply !
mzenker
Posts: 1999
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by mzenker »

Hi,

since you use Salomé, you can solve the problem of duplicated interfaces at mesh generation: Make a partition of the geometry you want to mesh and mesh that one, as described here: http://www.elmerfem.org/elmerwiki/index ... e_to_Elmer.
That description is a bit outdated, but still valid. The step of recreating the geometry groups in the mesh is now simplified by the "create groups from geometry" entry in the context menu.

HTH,
Matthias
Sheeppowa
Posts: 7
Joined: 19 Jul 2019, 15:54
Antispam: Yes

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by Sheeppowa »

Hi Matthias,

This is exactly the link I read to create my mesh. I have created the partition on the geometry set made of the three different solid and I add thee groups made of the boundaries including the interface between the specimen and the grips.

After the mesh, I create the groups of faces (each group is a boundary) and three groups of volumes.

I also refine the mesh on the specimen. But maybe I made wrong
Attachments
Mesh.JPG
Salome mesh
(109.8 KiB) Not downloaded yet
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by kevinarden »

Salome has a coincident node check and a coincident node merge option. It my be worth checking.
Sheeppowa
Posts: 7
Joined: 19 Jul 2019, 15:54
Antispam: Yes

Re: A Coupled Electrical–Thermal–(Mechanical) Modeling of Gleeble Tensile Tests

Post by Sheeppowa »

Oh ok, I will checked that so.

Thank you, I come back when I find.
Post Reply