[BigPapi34]

Hi there,

New to the CFD scene and looking for some guidance.

I've got the geometry of a centrifugal pump (not just an impeller) and am looking to generate a head-flow curve in Fluent and compare against the manufacturer's curve I have.

My problem arises when deciding what boundary conditions to use for the inlet and outlet.

Any ideas/suggestions would be great!

Cheers

[swiftaircraft]

Are you planning on a SRF or an MRF simulation? I would suggest you use a velocity inlet as then you know exactly how much flow you are feeding the pump and for the outlet use "OUtflow". If the outlet of the model is close to the outlet of the impeller a pressure outlet can adversely affect the flow characteristics you see. If you have reverse flow in the outflow then move the outlet of the model further from the outlet of the pump.

[BigPapi34]

Thanks for responding!

I'm using a MRF.

I've seen papers and other places suggesting using mass flow (inlet) and static pressure (outlet) or total pressure (inlet) and static pressure (outlet) but I'm having problems defining what these are.

If I choose to use the velocity inlet then what do you suggest I should be using as the gauge pressure in the inlet boundary screen?

Also, if I were to go for a pressure (outlet) BC, what would be the pressure utilised here?

Finally (apologies) - should this be conducted as a transient analysis?

Thanks in advance for any more info!

[Sixkillers]

MRF is more suitable for a steady state simulation. On the other hand sliding mesh technique is designed for a transient simulation. Needless to say that transient method is more time consuming. So you might want to start with steady state and see how it goes.

[BigPapi34]

Thanks for the info.

I'm having a look at doing this a few different ways, currently using pressures for the inlet and outlet.

Most literature is states the following:

Inlet - 'total' pressure
Outlet - 'static' pressure

Can anyone shed any light on these?

The pump is single stage and, as mentioned, I have the head-flow curve.

Cheers

[sk692]

Hi there

I am too simulating the flow in a centrifugal pump using MRF in fluent. But I am getting almost constant or rather increasing total head with increase in flow rate at constant RPM. However with increase in speed the total head increases i.e. I am getting a almost constant value at a particular speed irrespective of the flow rate.
I have tried all the BC combinations, I can think off. I have tried
- mass flow inlet and pressure outlet
- mass flow inlet and outflow
- pressure inlet ad pressure outlet

But the results are coming the same.

Can anyone shed some light on this problem.

Thanks a lot in advance.

[chingyinh]

Originally Posted by BigPapi34

Thanks for the info.

I'm having a look at doing this a few different ways, currently using pressures for the inlet and outlet.

Most literature is states the following:

Inlet - 'total' pressure
Outlet - 'static' pressure

Can anyone shed any light on these?

The pump is single stage and, as mentioned, I have the head-flow curve.

Cheers

It seems most correct because the flow rate is subject to the system friction so velocity inlet / mass flow inlet should be wrong where the flow rate should not fix in the pump.