Learning Forum

[Christian Neuer]

Hi everybody,

I have a coupled field (CF) simulation with thermal and structural as selected physics.

Previously I had a transient thermal and a transient structural analysis system linked, with the disadvantage that the structural analysis only started once the thermal analysis was completed. Since I have one material with a very high expansion rate (more than tenfold) I had to "adjust" the lambda values for the expansion I'd expect at a certain temperature.

Then I tried to use a CF to calculate both physics in each timestep. I thought I had to use the "real" lambda values now since my material has now the correct expansion (in this case the correct thickness) for the calculation of the heat conductance. Looking at my simulation results I saw temperatures that were way to high on the "cold" side. My suspicion was that the expansion at a certain timestep in my simulation wasn't taken into consideration for the calculation of the heat conduction. To confirm this I tried another run with the SCTE close to glass (=way lower) and I got the exact same temperature on the "cold" side again.

So my question: Am I doing sth. wrong (e.g. missing a check somewhere) or isn't the CF analysis able to take variing lenghts/thicknesses during a simulation into consideration when calculating the heat conduction?

Thank you,

Christian

[danielshaw]

Do you have Large Deflection ON?

[Christian Neuer]

 

Are you saying that CF only considers the increase in thickness for the calculation of the thermal resistance if Large Deflection is ON? I have it turned off right now, because I thought large deflection only applies to problems like fishing rods or diving boards.

My problem is more like the pic I inserted, I have a multilayered fire-resistant glass which consists of two glass panes (blue) and an intumescent layer in between (grey). Do you think turning on large deflection is the solution?

Thanks for your answer!

 

 

[peteroznewman]

Large Deflection means nodal locations are updated as the solution proceeds, so you need that turned on to get the updated thickness of the model as the solution proceeds.

[Christian Neuer]

Thank you very much, that solves my problem!

[Erik Kostson]

Thanks for that - we will close this post as it might be of use to others.

Erik