Wrong Thermal Simulation

Hello,

I've done a simple test case to show up the difficulty i have to get a thermal simulation of a two blocks stacked.
Two blocs are joined together and AFFE_CHAR_THER/FLUX_REP sources are applied at the horizontal junction. The blocs have the same MATERIAL properties.

The top and the bottom faces of the respective blocs have the same AFFE_CHAR_THER/ECHANGE.
After Aster solving and after seeing the Cut-planes (100'th sample), we see that there is a frontier at the junction. The temperature should diffuse in both top and down direction.

I attach the project to this post to better understand the test.

To summarize, when there are two geometrical blocs and planar sources,
There is a discontinuity in the mesh solving; even the two blocs have the same material (TOUS = OUI). The main question is, does there is a good practice to join two blocs with planar sources and heat exchanges?

Cordially<br /><br />Post edited by: stephane, at: 2011/08/23 20:37

I cannot attach the file.

regards<br /><br />Post edited by: stephane, at: 2011/08/23 20:46

Attaching the Python scripts