topic Re: Heat transfer through insulation in Inventor Nastran Forum

Heat transfer through insulation

Anonymous — Tue, 13 Oct 2015 04:22:22 GMT

Is there a way to set up a simulation so that it shows me what temperature may I expect on the opposite surface of a 50" x 50" piece of insulation? I only have temperature (1800F), insulation thickness (6"), thermal conductivity (at 1800F = 0.03 in*lbf/(s*in*F) ), and ambient temp. (80F) given. I have been trying to figure it out for many hours today, no luck so far..

I would appreciate any advise

Re: Heat transfer through insulation

david_cordova — Tue, 13 Oct 2015 05:08:00 GMT

You would need to do a steady state heat transfer analysis. It sounds like you have most of the inputs you need. You would need to apply a temperature load of 1800 deg on one face, and probably convection on the opposite face using an ambient temp of 80 deg and the appropriate convection coefficient. You may need to define boundary conditions on the other faces. If you don't define boundary conditions to the other faces they will behave as perfect insulators and you will essentially have a 1D heat transfer problem.

Re: Heat transfer through insulation

Anonymous — Tue, 13 Oct 2015 14:14:02 GMT

Hi David,

Thank you for a bit of guidance, I tried what you proposed and the problem I encounter is that when I set up convection and solve the simulation the "colder" side of the block (the one with convection on it) is completly copying the convection temperature. So if I set up convection ambient at 80F one side of my block is 1800 and the other one is 80 which is not realistic.. from my experience it should easly be around 160F or more

Do you know what could be the problem?

Re: Heat transfer through insulation

david_cordova — Tue, 13 Oct 2015 18:07:11 GMT

It sounds like the analysis ran successfully so the setup is correct however the results aren't what you are expecting so something is likely off with the inputs. I would double check your material properties, convection coefficient, and the units for these inputs.

Re: Heat transfer through insulation

Anonymous — Tue, 13 Oct 2015 22:55:12 GMT

It looks like it was convection coeffticient, thank your for pointing me into the right direction. 🙂 Now I will have to find the appropriate one, for 80F, natural convection environment. Is there any value you could propose from your experience?

Re: Heat transfer through insulation

Andrew.Sartorelli — Wed, 14 Oct 2015 10:28:41 GMT

Hi MP81,

If you are just using natural convection you could use 3.21e-6 W/(mm^2-C), this is the value we use in Simulation Mechanical for air at 20C with free convection.

______________________________________________________________

If my post answers your question, please click the "Accept as Solution" button. This helps everyone find answers more quickly!

Re: Heat transfer through insulation

Anonymous — Wed, 14 Oct 2015 16:52:57 GMT

It looks like my initial excitment that the problem is solved was premature, similar issue came back.

Andrew, I used your coefficient (thanks for that) and I converted it to Btu/(sec*in^2*F) which gave me the value of 0.000001091. The part I am trying to simulate now is made of mild steel (I have attached the part along with the in-cad simulation files).

Coefficient and thermal conductivity values should be correct.

The problem remains the same, on one side of the part I applied temperature of 1800F and to the front face I applied convection (20C ambient).

The front face sustain the ambient temperature, it's like it doesn't resond to the applied temp. Its hard to believe that exposed to the temperature that high, steel wouldn't heat up more. Also, doesn't matter how high the temperature I apply is (I tried 3000F), it changes nothing in relation to min temp. of the face with applied convection. The heat doesn't seem to spread.

The only thing that changes something is if I turn my conductivity coeff. very low, or thermal conductivity very high.

Do you have any suggestion what am I doing wrong?