Forced and natural convection using ABAQUS

[curlyq]

I've done a bit of work in ABAQUS in the past, but all has been some form of mechanical loading. I have now been asked to characterize the differences between two proposed heating bath systems. Essentially I have this:

Heated cylindrical bath (either molten salt or sand).
Heat flux into the bottom to maintain bath at 500 C.
Ambient room temp 22 C.
Bath insulated on all sides, open on top.
Natural convection in the air above the bath.
Possible forced convection in the bath media (eventually would like to add this in).
K(salt)=0.42 W/m*K
K(sand)=1.38 W/m*K
H(air)=40W/m^2*K

I am trying to evaluate the two different materials, and assuming a constant 500 degrees C input on the bottom, fully insulated sides, natural convection out through the top - what is the difference in surface temperature, and the temperature gradient from using these two different bath media?

If this were all conduction, I wouldn't have a problem, but I can't seem to figure out how to make convection in an ABAQUS simulation.

 

[curlyq]

I just thought of a good analogy - Think of this like a giant coffee cup, but with a heater on the bottom.

 

[corus]

You'll have natural convection plus radiation from the top surface. If the surface isn't flush with the sides then you'll also have to work out a view factor from the surface to the ambient air, together with the emissivty of the material.
Natural convection usually varies with the temperature difference to the power 1/3. There's no way of inputting this into Abaqus so you'll have to prepare a table of surface temperature against coefficient and input the values that way. You can also add the radiation term to that table or just apply it via surface radiation in the interaction module.


Tata but not yet tara

 

[curlyq]

While physically you are correct that the surface of the bath isn't flush with the sides, I am starting with the assumption that it is. Right now, I'm more concerned with a semi-accurate comparison between these two situations rather than a realistic accurate answer. The later isn't likely possible without CFD software because both bath media are actually under forced convection also.

However, I'll work on the table you mention and see if I can get something to work that way. Thanks for the tip.