Negative pressure in centrifugal pump
 

A Centifugal PUMP is modeled in Ansys CFX And solved with single phase condition, The problem is that in eye, high negative pressures occurs(Refrence pressure=1 Atm and negative pressure in eye= -10Atm) on small areas of leading edge of blades and hub.
I wanted to know why this happens and how it could be prevented.

Thanks for your answers

How do you define you material?

Assuming the effect you have modelled is real:

It is quite common for incompressible simulations to give negative absolute pressures. Physically what happens is a small cavitation bubble forms and the pressure is relieved. But for the simulation you can either add a cavitation model if this is important, or you can ignore it if the cavitating region is small enough to be ignored. You have to decide based on what you are trying to model.

negative pressure in centrifugal pump
 

hi my frinds
thankyou very much for your answer.

I have another problems:

1) I use of a inducer before impeler, but it cuse to reduce the prssure in end of itself !!!
and there are small area with negative pressure(approximately -30 bar) in end of inducer similar.

2) and my Yplus in cfx post is (betwen 0 to 4750)

3) when i run my simulation, i have a notice that a wall has been situated in a portion of boundry condition and .... i don know why?

4) i dont know how i can use of the CFX cavitation ability AND wich model of cavitation is suitable?

my turbulence model is SST.
my inlet boundary condition is inlet static pressure and outlet massflow rate.
and my simulation modeled in general simulation type.

thanks a lot.

I have already explained what the -30 bar pressure is about.

Whether your y+ is good enough has to be established by a sensitivity analysis.

The wall at the boundary has been asked many times ont he forum, and is explained in the documentation. Do a search on the forum or read the documentation.

It sounds like you are a CFD beginner so don't even consider doing cavitation modelling.

negative pressure in centrifugal pump
 

hi Gelenn
thanks so much for your reply
it was very helpful for me

1- do i have to use of the Multiphase ability for modeling the cavitation?
2- is it true that inducer use to reduce of pressure? isn't it non logical?
in fact we use the inducer before impeller for increased pressure and prevent of cavitation!!!!!

amjad

best regards

Yes, cavitation modelling is multiphase. But as I said before - don't do a cavitation model. Until you have the basics under control a cavitation model will not help.

negative pressure in centrifugal pump
 

hi Glenn

thanks a lot. your reply was very helpful.

now. i have a pipe that an inducer or (helix or spiral) rotate in pipe.
that the gap betwen helix and pipe is 2 milimeter.

what is this that domain ? rotating or stationary domain?

is this true? stationary domain with rotating wall.

best wishes

Sorry, I do not understand your question. Can you explain it again?

negative pressure in centrifugal pump
 

Attachment 6498http://www.cfd-online.com/Forums/L:\inducer

hi glenn

thanks again for your reply
I said i have a pipe in wich there is an inducer rotating.
The axis of rotating of inducer coincides withe the axis of symmetry of the pipe. The fluid in the pipe is driven by the inducer.
I want to model the flow with CFX. in your opinion is it better to assume the fluid domain as a stationary domain and the inducer wall(surfaces) as rotating wall OR the fluid domain as a rotating domain and the inducer wall as a rotating wall too . the gap between inducer and pipe wall(fixed wall) is 2 millimeter.

best regard

Looks like this can be modelled with a single rotating frame of reference.

But be aware I think your inlet and outlet bounadries are too close to your impeller to get good results or convergence.

physical significance of negative pressure
 

Dear respected sir,

I want to talk on this.. In my simulation CFX , of centrifugal pump, i am getting negative press -30000 pascal and that too, not in small region, but substantial region.

What is this.. i mean .. we can not go below 0 bar. isnt it.? even if cavitation has to occure, we should get pressure below 3200 Pa which is vapor pressure of water isnt it.?

Thanks and regards
ARohit

This thread already discusses what to do about the negative pressures. Have you read the other posts on this thread?

reply: thank you.
 

Ohk..

I am sorry.

I am struggling with another probl. Head i am getting in this sim is more than experimental by 300%.. this is too much error.

i am refering work done by another senior person.

i took same mesh, same no of elements, used same scheme (first order) and same converg criteria..

still my results are not good.

can you tell why ?

thank you.

That general question is a FAQ: http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

hi dudes
you know as the matter of fact in reality when you wanna measure the pump pressures there couldn't be any kind of negative results in terms of absolute pressure. there isn't any kind of pressure measurer that shows the absolute pressure in negative results. I think when you're facing with negative pressures in your numerical investigations it's because of your solver that solves the pressure-velocity coupling in a relative manner not solving absolute pressure. so the question is how could you set the options in ansys solver that the navier stokes equation will be solved in absolute pressure terms not relative terms?
I think even you didn't turn the cavitation settings on, you shouldn't have the negative pressures.

Hi,
I have also faced such problems in CFX,. in rotating domains, the values of "pressure" are surprisingly negative , in orders of -5 bar :-(

This problem may occur when using rotating domain for impeller. It would be better to use "absolute pressure" instead of "pressure" in the rotating domains.

The absolute pressure and pressure variables are the same, just with an offset of the reference pressure. So if you are getting a pressure of -5 bar and your reference pressure is 6 bar then there is no problem as the absolute pressure is +1 bar.

But if you are getting negative absolute pressure then you have a problem. Post #3 on this thread already mentions the most common source of negative absolute pressures.