Unstable Pump Curves

[adula29]

Hello
I have a small question.
Sometimes in pumps (in point of fact in high specyfic speed pumps) the performance curve is unstable (by decrease capacity).
What is the reason of this effect.
Partial cutback of vanes at exit from the impeller in an inclined plane make that the curve get stalbe ...WHY?

 

[adula29]

I would like renew my question!
Thank for answer If somebody know of course

 

[MikeHalloran]

It's a wonder the damn thing works at all.

... by which I mean that a centrifugal pump is a kinetic device, so changing any part of its geometry is likely to change its behavior, and usually does.

Also because it's a kinetic device, it's not possible to test such a pump in isolation; it's always interacting with the system in which it happens to find itself, and the systems are not universally stable in their behaviors.

If memory serves, somewhere within the pile of books that I have read or bought about pumps, filing the vanes changes the shape of the capacity curve, which in turn moves the operating point within a given system. Sometimes the OP moves to a better place, sometimes not.

I don't think it's technically correct to say that a pump's curve can be unstable, but the pump _system_ can exhibit odd behaviors. Even more odd is the range of theories that local 'experts' will synthesize to explain said behaviors.

As for me, I'd clean out the, e.g., barnacles, and personally measure the operating point before filing anything.



Mike Halloran
Pembroke Pines, FL, USA

 

[MartinSr00]

I'm not an expert, but I think it has something to do with an angle mismatch between the relative velocity vector from the impeller vanes and the volute or diffuser vanes. They're designed to match at design flow but are off at different flows. Remember in very high specific speed pumps, flow is high relative to head, so velocity head losses as a fraction of total head are probably higher.

When you angle cut, you give the fluid space to align itself betore entering the diffuser/volute opening.

That's just a guess.

 

[electricpete]

At the risk of relying on my unreliable memory...

I remember Den Hartog's Mechanical Vibrations describes an unstable pump curve in similar terms to what Mike described.

I also seem to vaguely remember somewhere an explanation related to angle of attack reminiscent airplane stall. Has some resemblance to what Martin said.

I'll be interested to hear what other explanations appear. I'd like to know myself. The only plants at our plant that have this characteristic are the large slow speed axial flow circulating water pumps.... not sure if there is something about this style pump that creates that characteristic.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[25362]

Another probable explanation. Sam Yedidiah in his book Centrifugal Pump User's Handbook Problems And Solutions Chapman and Hall, refers to recirculation on performance at low NPSH saying,

...The inlet edges of blades with high specific speed are relatively long and more inclined relative to the axis.. Here the low pressure zone created by prerotation contacts a large protion of the blade inlets.
Consequently, it has a great effect on performance. It increases the NPSH requirements of the pump by reducing the pressure along significant portions of the blade inlets.
Thus, pumps with high specific speed may cause a lot of trouble when operating at partial flow rates with low available NPSH. In extreme cases, the entire pumping system may operate with high noise and vibration, as if subjected to heavy hammer blows.

 

[stanier]

Check out

http://www.mcnallyinstitute.com

for an answer

 

[electricpete]

can you be a little more specific stanier?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.