Hello

I am solving heat flow problem between two rotating cylindrical surfaces (part of my master thesis). The gap between them is more less 0.075 (depends on case) .

As far as i know this heat flow can be modeled in analitic way by miechiejew equation (Nusslet number is function of Reynolds; Prandtl_fluid; Prandtl_wall). The Heat transfer coefficient obtained on paper is arouns 850 W/m^2/K.

In Fluent i obtain Surface heat transfer coefficient around 68 W/m^2/K. But the Wall Func. Heat transfer coefficient is 1530 W/m^2/K.

My sense tells me that is the result of different aproach of calculating heat transfer coefficient by Fluent. I tried to find anwser in help but i did'n manage to find it. Does anybody can redirect me to source where is the description how heat transfer coefficient is calculated and what is Wall Func. Heat transfer coefficient . Thank you in advice.

PS: details about solver settings: Standart k-epsilon model with viscous heating with standard wall functions.

Regards Stan

Fluent directly calculates heat flux on the boundaries. In Fluent, heat transfer coefficient (h) is a derived quantity using reference temperature, adjacent fluid temperature and heat flux. In the actual scenario, the reference temperature should be the wall temperature. Therefore, the values reported by fluent will be an indicative values. To get the actual h values you need to write a UDF according to your need.

Thank you very much. Now i understand how h is calculated. According tp help this is the way of calculating it in laminar flow. For turbulent flow it uses wall functions which are beyond my borders of understanding, at this momemnt.

I do not understand at all second part of your post. Anyway my question was touching the difference between two values available for postprocessing: Wall Fluxes\Surface heat transfer coefficient Wall Fluxes\Wall Func. heat transfer coefficient

Additionaly I wolud like to know how they are calculated (my thesis supervisor wishes that because difference between analitical model and numerical are to big and he do not buys that this is caused by material data).

[doodek]

Hi,
I have identical problem. How to find actual h values by UDF?

Best regards,
doodek

[cdf_user]

The h values reported by the fluent are wrong. It has something to do with the reference temperature which can be different for different models. Ask fluent engineer. To get the right h value you need to write a udf.

h = heat_flux / (T(z) - Tbulk(z))

Tbulk = SUM mu*cp*rho*Tf dA / SUM mu*cp*dA

[fierceyeo]

Hi, following the above discussion, therefore is Fluent overestimating the h value, or underestimating? Because I have h value (surface heat transfer coefficient) of about 400, and it is natural convection

[cdf_user]

Can you describe what you are modeling, such as geometry details? Also the 400 value, is that a constant h value across your fluid domain?

[fierceyeo]

It is a natural convection in an enclosure (with heating element at the center). The heating element is a square, with heat transfer coefficients at the top, left, and right surface are about 150-400 W/m^2.K.
My question is...normally surface heat transfer coefficient in Fluent will be difference by how much from reality (as in using UDF)?

[cdf_user]

Your natural convection is very high. Such high convection is only found in micro and nano channels. As far as the difference between fluent h and real h is concerned, there is no specific number. If you raise or lower your reference temperature, the difference that you are asking will change. Thats why I suggest you make planes across your fluid domain and use the formula above to get Tbulk and then h = q/T(z) - Tbulk.

[fierceyeo]

Hi, I am new to UDF. So, in order to write these functions..
"h = heat_flux / (T(z) - Tbulk(z))

Tbulk = SUM mu*cp*rho*Tf dA / SUM mu*cp*dA"

how to I define these variables "mu, cp, rho"? and is this equation applicable for 2D problem?

Thank you.

[oky]

Hi, everyone.

I need help,
How to get the value of convection coefficient [h] from Fluent directly ?

Thank you,