[ryanmoser]

I'm trying to simulate a heat accumulator with FLUENT14. The setup involves a inner tube(φ14×10） through which water flows, a storage surrounding the tube which is filled by composite phase change materials(it can be treated as a conduction problem within the phase change materials).
The charging process and discharging process are simulated. The solidification/melting model is used. Since the Re of water flowing in the tube is about 10000, k-ε turbulent model is also included.
But after the calculation, there's no obvious temperature difference along the axial direction at the same radius, which doesn't obey the physical law and doesn't accord with the experimental results.
I've no idea where to improve my numerical model. the mesh quality/quantity? or any additional boundary conditions?
Can anyone help me? Thanks so much!
Since the cas file is too larger, I cannot upload it as an attachment. If you are so generous to help me, I can send the cas file to you.

[Kwiaci]

What is your outer wall boundary condition? Do you have there constant temperature, convectional heat transfer or none? What about heat fluxes along internal wall - are they constant?
Try to improve these boundary conditions.

[ryanmoser]

Originally Posted by Kwiaci

What is your outer wall boundary condition? Do you have there constant temperature, convectional heat transfer or none? What about heat fluxes along internal wall - are they constant?
Try to improve these boundary conditions.

the outer wall is adiabatic

[Kwiaci]

Do you have any temperature distribution in the solid?

An internally heated solid that is experiencing no external cooling will undergo a constant rise in temperature and you will not see any axial temperature difference.
I think that you need the external cooling to see temperature difference.

First check do you have coupled condition on internal wall.

[ryanmoser]

Originally Posted by Kwiaci

Do you have any temperature distribution in the solid?

An internally heated solid that is experiencing no external cooling will undergo a constant rise in temperature and you will not see any axial temperature difference.
I think that you need the external cooling to see temperature difference.

First check do you have coupled condition on internal wall.

There's temperature difference along the solid tube, but not obvious. The differecne between the maximun and minimun is about 4℃.
Does it relate to the time steps? I have set relativly large time step like 0.06s, and it converged well. Should I try a smaller time step?