[hatef]

Hey all,

How do I assign time dependent material properties?
Your help will be much appreciated!

Thanks,
Hatef

[Lance]

well, this will work:
1000 [kg/m^3]*t*1[s^-1]
then just assign the expression to your material.

[hatef]

Thanks for your reply, I actually tried to use that, but it doesn't work on its own! I guess CFX generates a table for temp. and pres. dependent material properties before the actual numerical iteration starts in order to reference them during simulation. I was wondering if I can manually generate a table for time!?

[Lance]

yep, use a user function. See e.g.
http://www.cfd-online.com/Forums/cfx...onditions.html

[hatef]

I have done time dependent boundary condition before, that works fine. But when it comes to defining a new material with time dependent properties, it doesn't like it!

[Lance]

OK, i see - maybe CFX wont allow time-dependent material properties for some reason. What are you trying to achieve? Maybe it can be done in another way?

[Birnensuppe]

Hallo Lance,

I am also interested in time dependent material properties.I simulate electric arcs and I am struggling at the moment with the implementation of the temperature dependent density of my fluid. The simulation crashes after 4 iteration, the simulation runs fine with a static density. My thought was that perhaps the simulation will success, if the density is calculated after 10^-6 s.

[Opaque]

Please keep in mind that certain material properties are not for the faint of the heart. They must satisfy strict thermodynamic requirements; otherwise, the solution may not satisfy the basic principles we are trying to model.

A thermodynamic property can only be a function of a minimum number of independent thermodynamic variables that uniquely define the state. It is common practice to be (T,p and composition) though it can also be other combinations, for example, h(p,s, composition). Time explicitly is not one of them, though it influences them through variation of T and p as the solution progresses.

With that out of the way, perhaps it is best to open another thread explaining your convergence issue when modeling the arc. I assume your density is changing rapidly during the iteration, and the solution becomes unstable or stiff. You may need some kind of under-relaxation to control the decoupling between your electric field, energy equation and properties.

Hope the above helps,