[hamidreza118]

Hi guys,

The centrifugal pump analysis after fluent's convergence is attached below for clarification. The problem is, why I can't see any further pressure increase in casing due to conservation of mass? (I mean the diffuser is supposed to increase pressure, right?)
Actually I believe the reason is constant velocity in casing zone as is shown in attachment BUT why should the INTERIOR of casing be still?! I set the Casing Frame Zone to be stationary but I have SET Interface between Casing Zone and Impeller zone (which is indeed rotary) at the blade ends.
Isn't it enough to Provide velocity in casing zone?! I'm so confused and would appreciate any HELP Thank you all
And what turbulance model is suited well for a case like this?
Height of the casing : 1.3 cm
Height of the impeller (blades) : 1 cm
Velocity Inlet : 2
OUTLET BC-> OUTFLOW ( I know it's better to define pressure outlet in this situation but I've tried that too)
Angular Velocty : 3000 rpm
Operating Condition : 0 pa
Thanks in Advance

[swtbkim]

Have you directly calculated the pressures at inlet and outlet part of diffuser?

It's not touching even me, but it can be expressed as same color zone in contour because pressure difference in diffuser is too small, compared to extremely large difference in pump.

[hamidreza118]

dear swtbkim, thanks for ur reply,
i had previously checked that but that was not the point.
I didn't notice that that the velocity in casing interior is COMPLETELY zero
as in previous attachments you can see there's no particle moving in any direction in casing and the green color in those pics is between a negative number and positive one and indeed is zero
velocity magnitude is zero and it seems there's no fluid movement in casing but I wonder how the pressure exists
I don't know what's happening here
AND it Seems the mesh in casing is not fine but I check all it's parameters like skewness, AR and .. and they were fine but WHEN fill button is off in showing contours i get this pic which is showing no calculation in this area. any idea?

[ghost82]

I think you have problems with interfaces: check velocity vectors between your rotating frame and stationary frame (you will see if your interface is a wall or a real interface).

You can also go back and merge the 2 interfaces in pre-processing and define 2 fluids (one rotating and one stationary), if you use mrf approach.

Daniele