CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Velocity in the Loop Pipe (http://www.cfd-online.com/Forums/fluent/101415-velocity-loop-pipe.html)

Blader May 3, 2012 09:26

Velocity in the Loop Pipe
 
Hello,

I am new to this forum and I'm starting to simulate the tool ANSYS / FLUENT. I am developing a system of a pipe closed, and I have problem to set the velocity of the fluid inside.

Can anyone tell me how to do?

The picture attached is the geometry of my system!
https://lh5.googleusercontent.com/-e...PipeFigura.png

Thanks in advance :)

LuckyTran May 3, 2012 10:45

Quote:

Originally Posted by Blader (Post 359086)
Hello,

I am new to this forum and I'm starting to simulate the tool ANSYS / FLUENT. I am developing a system of a pipe closed, and I have problem to set the velocity of the fluid inside.

Can anyone tell me how to do?

The picture attached is the geometry of my system!

Thanks in advance :)

There's nothing you can do, as there are no boundary conditions to apply. Think of it this way, if you had only a closed pipe with fluid in it, could you alter the flow in any way? The answer is no, because you do not have access to the fluid inside the pipe. You need to come up with a more physical/realistic geometry if you want to model anything. I already know the solution to the flow for your geometry, there will be no bulk flow, maybe a little natural convection.

Blader May 3, 2012 12:00

Thanks for the input LuckyTran,

For better understanding, my goal is to heat a wall to carry the heat through the fluid to the other side of the pipe. The effect of the closed loop is very important to keep in the simulation

https://lh4.googleusercontent.com/-h...1-n-k/Pipe.PNG

There is no way to do considering that there is no fluid can enter the loop because the temperature is changing over time?

One idea could be a pump, but I want to avoid this solution.

Thanks for help;)

taekyu8 May 3, 2012 12:07

I also thought similar loop case.
Is there any way to set the flow inside in this loop with this geometry? I heard I can set the volume flow byUDF, however, I can not find similare tutorial. Anyone konws how to set the flow in this case?

LuckyTran May 3, 2012 12:40

Quote:

Originally Posted by Blader (Post 359137)
Thanks for the input LuckyTran,
There is no way to do considering that there is no fluid can enter the loop because the temperature is changing over time?

How is this going to generate any flow? For incompressible flows, it doesn't. The flow would need to be highly compressible or phase changes would need to be involved.

You need to break open the pipe and install a pipe or something similar. So create try to make an open loop version or an open loop that is effectively a closed loop by linking the inlet and exit conditions.

Most importantly, even if you could generate a steady flow, a closed loop simulation with heat transfer can easily diverge from accumulated numerical errors. Because the energy balance must be exactly correct, (heat input, output, and heat transported by the working fluid), any small residuals can easily accumulate over iterations and result in a runaway/divergent simulation.

taekyu8 May 3, 2012 13:32

Thanks for the note.

My case is fixed temperature heater. I will heat up the inside gas up to certain temperature and want to see the temperature change over time. The initial temperature is room temperature. It will be transient calculation.

How do I link the inlet and exit?


Quote:

Originally Posted by LuckyTran (Post 359144)
How is this going to generate any flow? For incompressible flows, it doesn't. The flow would need to be highly compressible or phase changes would need to be involved.

You need to break open the pipe and install a pipe or something similar. So create try to make an open loop version or an open loop that is effectively a closed loop by linking the inlet and exit conditions.

Most importantly, even if you could generate a steady flow, a closed loop simulation with heat transfer can easily diverge from accumulated numerical errors. Because the energy balance must be exactly correct, (heat input, output, and heat transported by the working fluid), any small residuals can easily accumulate over iterations and result in a runaway/divergent simulation.


Blader May 3, 2012 13:43

I have the same question, how do I link the inlet and exit ?

Again, thanks luckyTrain and taekyu8 for help :) !!

LuckyTran May 3, 2012 14:02

Quote:

Originally Posted by taekyu8 (Post 359154)
Thanks for the note.

My case is fixed temperature heater. I will heat up the inside gas up to certain temperature and want to see the temperature change over time. The initial temperature is room temperature. It will be transient calculation.

How do I link the inlet and exit?

So there is no flow? This isn't even a fluid flow problem? If there is no flow then the solution to the energy equation is straightforward (probably even calculator using algebra).

You can force Fluent to model many things (might not be a good idea) by using UDFs. There's an entire forum for it.


All times are GMT -4. The time now is 09:58.