wrong velocity value from solution initialization
 

Hello everybody,
I'm trying to make a 2D simulation, with incopressible flow (water) with energy equation solving too, on Fluent. I use mass flow inlet and outflow as boundary conditions at the inlet at the outlet.
I don't know what's happening, but when I initialize the solution, from the inlet, the axial speed value calculated is completely wrong. It returns approximately 390m/s instead of 0.5m/s :confused:. I have checked the values settings ​​for the cross-section , the density and the flow and they are all correct.
Could anyone help me please

Thanks in advance

Can you show the calculation why you think it should be 0.5 m/s?

Hello,
here are my calculations:

mass flow=0.15 kg/s
Diameter=0.02m -> Section=3,14*10^-4 m^2
density=1000kg/m^3

velocity=mass flow/(section*density)=0,4777

thanks a lot

It could be that your axis of rotation is wrong...

Can you let Fluent calculate the mass flow rate through the inlet after initialization? (Reports->Flux reports)

I tried to make this check after initializing and it returns the correct mass flow rate: 0.15 kg/s at the inlet.
After the calculation and convergence it returns the correct mass flow rate: 0.15Kg/s at the inlet and -0.15kg/s at the outlet. So the flow field is completly wrong.
I'm desperate!

Did you check if you need to scale the mesh?

Your model is axisymmetrical. How sure are you that Fluent uses the axis of rotation that you intended?

What is the axis of rotation in your model?
And which axis does Fluent use?

I'm sure Fluent uses the correct pipe's axis, because it's the once in the boundary conditions. I checked the cross section and it's correct. In addiction I tried to rotate my model around the axis...

For ghost82: what do you mean for scaling mesh? I'm sorry

Edit: sorry, wrong calculation.

0,15 kg/s at the inlet are 0,47 m/s in velocity, so what's wrong?
The -0,15 kg/s means fluid is exiting your domain.
Can you attach a picture of the velocity contour?

Don't be so sure. By default, Fluent will assume that the axis boundary is on the y=0 line. (Fluent User's Guide, 6.3.17.) Is that true in your case?

Hi pakk,
sorry for ot, I read what fluent's user guide reports about axisymmetric problems and y axis; however I always set the x axis in the preprocessor as the axis of rotation (and domain in +y, +x region).
Any other axis, y or z, gives divergence at first iteration.
Do you know more about this?

http://www.cfd-online.com/Forums/flu...-momentum.html

Error in the user guide?Or different axis for different fluent versions?
http://aerojet.engr.ucdavis.edu/flue...g/node1217.htm
"For axisymmetric problems, in which the rotation axis must be the x axis,..."

EDIT: ok, y=0 means x is the axis of rotation :)

SOLVED!
I did not know that , by selecting a different material than the default , would create a mix ! I noticed when I created the new geometry to the attachment. Thank you very much.

Sorry if I wasted your time !

:o:o:o

If it is a simply pipe, you should only model half of a cross-section.

Sure, I usually model only a half of the cross section and after the convergence I rotate the model around it's axis.
Thanks a lot everybody for your availability.