Learning Forum

[Chun-Wei Tsai]

I am doing transient thermal analysis of jetfuel for long haul flight, I have set the boundary condition of walls with heat transfer coefficeint of 15 and temperature of -56.5degree C. And the initial temp. of jetfuel inside being 15degree celcius, the results doesnt really make sense as I expected the temperature would cool from outside to inside gradually. The I have set jetfuel properties with the default piecewise polynomial defalt for jet a 1 from the database and turn the gravity on. Also I have chosen laminar ,energy model for the simulation.

[Chun-Wei Tsai]

 

This is the residuals plot

[Chun-Wei Tsai]

and the meshing

[Rob]

And density varies with temperature in a closed domain? Where is the extra volume coming from or going to? 

[Chun-Wei Tsai]

I didn't think of that at the first place, maybe that's the problem. I have tried using constant density and the result of contour plot showing something like this, I wonder how reasonable is the result.

[Rob]

If it's constant density you'll not see any convective mixing. Just diffusion & conduction.  With changes in density you also need to account for volume: there's a reason for the Bousinesq approximation. 

[Chun-Wei Tsai]

Thank you for your answer!! If I'd like to use density change, what kind of change in my model would you recommend? Will it be much more complex? Or do I just need to leave some empty space

[Rob]

You'd either need Bousinesq (read for why) or add a pressure boundary somewhere to let the solver add/remove a tiny amount of mass to balance the volume. 

Leaving a "space" would require a multiphase model, and that's another layer of complexity that you may not want, or need, at present.