Learning Forum

[rohant]

I am trying to apply a 2D flux to a pair of curved plates mimicing channel flow. I know what this flux should be in 3D, however I have some doubts about the applicaiton in 2D. I know Fluent takes the depth as 1m - but does this affect what my flux should be? I can change the depth in the reference values section to be what I am expecting - but is this the right way to go about it?

Or should I apply the flux assuming my total heat input is over the curved length * 1m depth? 

Thanks for your help. 

[Nikhil N]

Hello,

I would recommend using the same heat flux value, but changing the depth in the reference values section accordingly so that the total heat input remains the same.

[Rob]

To add you need to use some care here.  If you scale the energy input you risk altering the W/m2 and therefore the system temperature. If you use the same W/m2 the overall flux may be different, but that also depends on how you scaled the flow....  If you want an interesting one to think through, what about mass flow and velocity: how do you scale those? 

Changing the reference also requires care as no one ever remembers to check it when viewing the results. It's also an easy setting to miss in a future case. 

[rohant]

Essentially, as long as the reference depth is scaled appropriately - mass flow rate and heat flux should be the same in both 3D and 2D cases correct? I would agree that for the default depth of 1m mass flow rate would be different in 2D and 3D planar cases. 

In this case we are inducing evaporation-condensation in a closed domain, the only boundary condition input is heat. 

[Rob]

Agreed. It's making sure all of the scaling is correct. 

For the above scenario make sure the gas is modelled as fully compressible (ideal gas) otherwise Fluent will do something fun with the spare mass!