[jlefevre76]

Hey forum users!

I'm a student working on my MS in mechanical engineering and a professor asked me for some help modeling free convection for a complex system.

Imagine a cylindrical disk, the top surface is exposed to quiescent air, the sides and bottom are more or less adiabatic (for now that's how we plan to model it). The disk is exposed to radiation from a laser that effectively gives it a volumetric heat generation as a function of r,z, and t.
( = f(r,z,t) )

As a general rule, I start with getting a simpler problem to solve before I do the complex one, so I started with solving the steady state problem for constant volumetric heat generation. This one is solving just fine and giving realistic results (correct order of magnitude).

Next, using the same constant volumetric heat generation, I tried to solve the transient problem, but results weren't making any sense. From what I could tell, no heat was escaping into the air this time, and the flow field made no sense.

Anyway, I just wanted to save myself trouble and see if anyone knows if it's true that fluent doesn't support transient conduction? From what I've read on the forums and for lack of details on the subject in the user manual, I'm guessing this is the case. I know transient conduction will work with a shadow surface (designating a thickness and material properties for a surface), but for a solid volume, is it possible to model transient conduction while solving the fluid field on one of the surfaces?

(by the way, the coincident surfaces of the fluid and solid volumes are coupled, that's why I assume the steady state was working)

So, my two main questions are:

1. Is transient conduction possible in fluent for a solid volume?
2. Does Adina, or XAnsys 11.0 support solving a transient conduction problem simultaneously with a transient convection problem?

Thanks guys!

[jlefevre76]

*Sniff, sniff* Well, I solved it, not that anyone cares...

It was an issue with the pressure inlet and outlet boundary conditions. I was using a cylindrical system and my direction of expected flow was poorly defined with a cartesian vector. Once I fixed the vector definitions, it knew which way the flow was supposed to go and yielded a realistic result. I also found it's better to solve it steady state laminar first, then do the transient problem. I'm not sure how I'll get around this for the transient source term, but I'll keep playing with it.

[syavash]

Originally Posted by jlefevre76

*Sniff, sniff* Well, I solved it, not that anyone cares...

It was an issue with the pressure inlet and outlet boundary conditions. I was using a cylindrical system and my direction of expected flow was poorly defined with a cartesian vector. Once I fixed the vector definitions, it knew which way the flow was supposed to go and yielded a realistic result. I also found it's better to solve it steady state laminar first, then do the transient problem. I'm not sure how I'll get around this for the transient source term, but I'll keep playing with it.

Dear Jeremy,

I could not overcome the same transient heat conduction problem. I have two fluid zones and a solid zone between them (2-D). I already have two shadow walls in Fluent which are coupled between adjacent fluid/solid zones. lower fluid is cold and upper one is hot. I would expect to see the heat conduction between two fluids through solid zone but nothing happens after too many time steps. would you tell me how to resolve it?!

tnx
Syavash