[thermal energy]

dear friends,

ı have a problem. problem is that

ı simulate melting for horizontal shell in tube (annulor). boundry conditions are that isothermal condition at the inner surface of the annulus and the adiabatic condition at the outer surface of annulus.
ı used simple algorithm (pressure based solver and transient).
As a spatial discretization, ı used least squares cell based, presto, power law and power law for gradient, pressure, momentum and energy respectively.

as an phase change material ı used paraffin (melting range Tm=56-58). paraffin properties are that ı used boussinessq mean (solid and liquid) density value ,constant cp, k, viscosity and thermal expansion coefficient. also, ı used melting point instead of range , Tm=57.

my geometry dimensions are like that; inner diameter is 28mm and outer diameter is 110mm. ı simulate only half of the annulor (symmetry condition).

my melting simulation is mentioned above. but solution is not good. that is my experimental data and simulated data do not have good agreement as a characteristic of figure. the figure is attached (T14 is temperature of pcm which is in lower part of annulor).

why does this difference occur?

any suggestion?
ı am really in need of help