CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Variable T_inf for convection problem in a pipe using groovyBC (http://www.cfd-online.com/Forums/openfoam-solving/123115-variable-t_inf-convection-problem-pipe-using-groovybc.html)

Mojtaba.a September 4, 2013 13:50

Variable T_inf for convection problem in a pipe using groovyBC
 
1 Attachment(s)
Hi all,
I am trying to simulate a simple pipe with constant heat flux walls, which are cooled by fluid flow going through the pipe.

Here is an image for better illustration:

http://www.cfd-online.com/Forums/att...1&d=1378316658

Here is what I am gonna do. For wall BC I want to use:

Code:

dT/dn= (q-h*(T-T_inf))/(k_f)
In which T_inf is the average bulk temperature of the fluid, which varies across the pipe.

To calculate bulk temperature I want to use the average value of T as a variable of x (Length from the inlet).

How can I calculate this average in each x and use it as T_inf?

Is it possible to do this using groovyBC?

gschaider September 4, 2013 15:08

Quote:

Originally Posted by Mojtaba.a (Post 449901)
Hi all,
I am trying to simulate a simple pipe with constant heat flux walls, which are cooled by fluid flow going through the pipe.

Here is an image for better illustration:

http://www.cfd-online.com/Forums/att...1&d=1378316658

Here is what I am gonna do. For wall BC I want to use:

Code:

dT/dn= (q-h*(T-T_inf))/(k_f)
In which T_inf is the average bulk temperature of the fluid, which varies across the pipe.

To calculate bulk temperature I want to use the average value of T as a variable of x (Length from the inlet).

How can I calculate this average in each x and use it as T_inf?

Is it possible to do this using groovyBC?

Not the way you'd want to.

The first thing is that T_avg(s) is not really well defined after the pipe bend (BTW: convention is that x is a component of cartesian coordinates. The string length which you mean is s. At least the way I learned it). So the first thing would be to specify/calculate it for the whole geometry. Then you can calculate your T_avg(s) (for this there is actually a functionObject swakExpressionAverageDistribution in the development verson. But currently there is no way to feed that data back to your BC)

Mojtaba.a September 4, 2013 15:50

Quote:

Originally Posted by gschaider (Post 449915)
Not the way you'd want to.

The first thing is that T_avg(s) is not really well defined after the pipe bend (BTW: convention is that x is a component of cartesian coordinates. The string length which you mean is s. At least the way I learned it). So the first thing would be to specify/calculate it for the whole geometry. Then you can calculate your T_avg(s) (for this there is actually a functionObject swakExpressionAverageDistribution in the development verson. But currently there is no way to feed that data back to your BC)

Thank you Bernhard.

jherb September 4, 2013 18:21

Why don't you use the turbulentHeatFluxTemperature boundary condition?

See also http://www.cfd-online.com/Forums/ope...mperature.html

gschaider September 4, 2013 19:10

Quote:

Originally Posted by jherb (Post 449959)
Why don't you use the turbulentHeatFluxTemperature boundary condition?

See also http://www.cfd-online.com/Forums/ope...mperature.html

In principle you're right. The problem is the T_inf he wants: an average of the T in the fluid. I don't think turbulentHeatFluxTemperature does that (last time I looked you specified the T_inf. It didn't get calculated from the solution)


All times are GMT -4. The time now is 12:48.