CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Bulk temperature for HTC extimation (

Italy July 9, 2007 13:41

Bulk temperature for HTC extimation

i need to calculate wall heat transfer coefficient and i know its values depends on the distance of the first node from the boundary wall.

I should set a bulk temperature, but it is very difficult to estimate it, beacuse i have a complex domain with three inlets and three different fluids (with three different temperatures and mass flow) on each inlet, mixing processes and variable velocity inside the domain

How can i estimate the value of bulk temperature for htc calculation? Is CFX able to calculate itself a bulk temperature localy or should i have to set a unique value for the entire domain? In this last case i think that big mistakes can occur in heat transfer coefficient calculation.

Thank you.

opaque July 9, 2007 15:46

Re: Bulk temperature for HTC extimation
Dear Italy,

What is your definition of Bulk Temperature? For CFX (though called) Bulk Temperature, it works like a reference temperature. That is,

Q_wall = HTC * (T_wall - T_?)


T_? is a temperature that is dependent of the definition of HTC. Near wall (first node) temperature is one choice. Any choice is valid as long as you use the HTC consistent with this definition.

The term bulk temperature creates some confusion because is not a mean temperature, but (for flow in pipes) is defined as

T_bulk = Integral over the cross section of (density * enthalpy * velocity dot dA) / Integral over cross section of (density * Cp * velocity dot dA)

Another example, heat transfer in a differentially heat cavity where

Q_wall = HTC * (T_hot - T_cold)

There is no need for a bulk, mean or any other reference temperature.

What is your definition for HTC? That will tell you how to compute T_?

Hope this helps,


Italy July 9, 2007 17:39

Re: Bulk temperature for HTC extimation
Dear Opaque,

In my opinion the bulk temperature near a wall is the temperature of the fluid just outside the boundary layer. When CFX computes HTC and the user don't specify a bulk temperature (expert parameter), CFX uses near wall (first node) temperature as T_?, where Q_wall = HTC * (T_wall - T_?. Doing that, if i make a plot of HTC on a fluid_solid interface, i obtain values that are very high and dependent from the first layer thickness. If I am using SST, my first nodes is very closed to surface and i obtain values of the order of 5e+5 W/m^2 K.

Is it possible to fix bulk temperature as the fluid temperature just outside the boundary layer? Obviously it should vary locally.

Thank you very much Opaque.


Helge July 10, 2007 05:35

Re: Bulk temperature for HTC extimation
We have seen that it is not easy to define a proper bulk temperature. Is it a local mean temperature or a global mean temperature. It is the mean value you would like to apply based on criteria like: What is the value I would like to compare with? How was this value been calculated? This is always the problem with relative values. The only absolute value in a conjugate heat transfer problem is the heat flux through the wall

prateek jain July 12, 2007 00:05

Re: Bulk temperature for HTC extimation
For avg valu of bulk temperature you can try manuly the Grashaf No. equation in which beta is a term which gives u avg. film temp.Might it will help you?

All times are GMT -4. The time now is 18:10.