[doodek]

Hi,
I'm modelling flow around a few rows of pipes in 2d (dp) and unfortunately I have a problem. The solution was converged, but I want to check correctness average heat transfer coefficient on surface of pipe in every row with heat transfer coefficient, which was calculated by criterial equation for this problem. The temperature of pipe is higher then temperature of fluid. There is 2 problems:
1. Nusselt number on the wall of pipe is negative. Why?
2. Difference between average heat transfer coefficient from Fluent and heat transfer coefficient calculated by me from criterial equation is very great (upper 200% difference). I changed length reference value in Fluent to number, which I used to calculate Reyonolds number, which was needed to criterial equation, but it has not solved a problem (the next problem is: what should I write to reference temperature in reference values?). It is possible to check this matter with using Fluent?

My settings in Fluent:
solver: pressure based
fluid: water
viscosity model: SST k-omega
boundary conditions: velocity inlet before pipes 0,6 m/s (it gives Reynolds number around 31 000 in narrow line between two pipes and using this number I have calculated turbulence intensity); pressure outlet (0 Pa) and walls on pipes.

I would be very grateful for tips.

doodek

[ztdep]

that means you take the direction of heat flux into account.

[doodek]

Thanks for answer.
I guess that great differences between two surface heat transfer coefficients result from reference temperature value. What should I write as reference temperature?