topic Re: Defining the convective coefficient in CFD Forum

Defining the convective coefficient

Kousnetsov — Wed, 13 Jan 2021 14:09:11 GMT

Hello! I am exploring the design of a heat exchanger

While it is hard to determine the overall heat transfer coefficient because of the complicated geometry (and impossibility to approve the results with a simple calculation by hand) i decided to define the convection factors separately and started with a simple example: tube, surrounded by air flow (the study is attached)

First of all, i would like to introduce terms and equations to avoid misunderstanding during the discussion

i started with applying materials (air and copper) and made their properies variable, then applied boundary conditions:

airflow (648 m3/hr) and temperature (20 C) at the inlet
zero pressure drop at the outlet
temperature (100 C) on the surface of the tube

I also refined the mesh (at least i suppose that it is fine, because the pressure drop and velocity profile did not change significantly)

after solving i expected that equations (1) and (2) will give the same amount of heat, but the differ very much.
According to the wall calculator:
Airflow average temperature at the outlet is 20.4526 C
Average film coefficient on the outer tube surface is 810.188 W/(m2*K)
Total heat transfer area of the tube outer side 9374.09 mm2
So here are the calculations:

i have also calculated the convection coefficient by hand and got another different number. I trust this manual calculation more, because it is described in many books (my sources are in russian, so i don't give the links)

And by the way,i have noticed, that film coefficient depends on the temperture in CFD, though it mostly depends on the flow velocity, its viscosity and conductivity. The surface temperature must not affect the convective coefficient (at least while it doesn't affect the flow near the tube).

What am i wrong with? Isn't it correct to use film coefficient as a convective coefficient? (ред.)

Re: Defining the convective coefficient

marwan_azzam — Fri, 15 Jan 2021 15:38:23 GMT

Hello,

Quick question:

Did you want the same volume of air to also flow inside the tube?

This is what you have now.

Marwan

Re: Defining the convective coefficient

Kousnetsov — Fri, 15 Jan 2021 19:06:57 GMT

Hello, Marwan!

Thanks a lot for attention to my problem.

As for your question, i would like to assign water flow inside the tube in the future, but now i don't have any BC's inside the tube (actually i also considered it as a solid cylinder). Now i consider the tube as just a heated object inside the air flow to achieve the same value of convective coefficient as in manual calculation.

As soon as i understand how to set the simulation correctly, i will apply the settings to a required geometry

Re: Defining the convective coefficient

marwan_azzam — Fri, 15 Jan 2021 19:28:31 GMT

Hello,

It really has to do with the mesh. In the attached model, I made the domain very thin so it is basically one element through the thickness. The mesh size is manual at 2mm for the air and 1mm for copper.

Running the wall calculator with 20 as the reference temperature gives me a Film Coefficient of 71 W/m^2K

Marwan

Re: Defining the convective coefficient

Kousnetsov — Sat, 30 Jan 2021 14:04:01 GMT

Hello, @marwan_azzam !

Sorry for long reply. I have just checked your simulation and noticed, that you have applied temperature BC to the inside surface of the tube. When i apply it to the outer surface, i get film coefficient 272 W/m2*K and larger heat flux.

Here are the wall calculator results from your simulation. Surface 9 is inner surface, 7 is outer. Please, explain, how to interprete different values of heat flux on both sides of the tube?

Re: Defining the convective coefficient

Kousnetsov — Sat, 30 Jan 2021 19:08:49 GMT

Here is the simulation of conjugate heat transfer. Airflow is outside, waterflow inside. Unfortunately, manual calculations also give different values of heat flux. Heat balance equations for both sides are listed below