Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Centrifugal Pump Analysis 2D Transient Turbulent Flow

Please login with a confirmed email address before reporting spam

Hello
I'm trying to analyse the total transient force acting on impellers of a centrifugal pump. But unfortunately comsol couldnt find an intial solution. I'm getting an error as
"Failed to find consistent initial values." I used two methods described in various forums like "using a ramp function for rpm" and "finer mesh".
I set boundary conditions as in the attached figure.
Is there any suggestions?


6 Replies Last Post 5 janv. 2012, 15:44 UTC−5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 31 déc. 2011, 10:29 UTC−5
Hi,

The very first question that comes to my mind is how did you define the inlet velocity? Why is there an arrow in the middle of your domain?

Cheers
Hi, The very first question that comes to my mind is how did you define the inlet velocity? Why is there an arrow in the middle of your domain? Cheers

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 31 déc. 2011, 10:35 UTC−5
Hi

And are you sure your blade curatures shoud not be inversed, the slope seems odd, for a centrifugal pump.
But you can also mirror the outlet to have it going "down", without changing other things in your

--
Good luck
Ivar
Hi And are you sure your blade curatures shoud not be inversed, the slope seems odd, for a centrifugal pump. But you can also mirror the outlet to have it going "down", without changing other things in your -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 5 janv. 2012, 09:11 UTC−5
Dear Amir,
I defined the inlet velocity as a normal inflow option on the boundary that is shown by the arrow. Arrows are only for showing the boundaries. They are not in the model.



Hi,

The very first question that comes to my mind is how did you define the inlet velocity? Why is there an arrow in the middle of your domain?

Cheers


Dear Amir, I defined the inlet velocity as a normal inflow option on the boundary that is shown by the arrow. Arrows are only for showing the boundaries. They are not in the model. [QUOTE] Hi, The very first question that comes to my mind is how did you define the inlet velocity? Why is there an arrow in the middle of your domain? Cheers [/QUOTE]

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 5 janv. 2012, 09:12 UTC−5
Dear Ivar,
I ran the analyse as you suggested but nothing changed. I think the curvatures of blades are correct. They are as in the literature.



Hi

And are you sure your blade curatures shoud not be inversed, the slope seems odd, for a centrifugal pump.
But you can also mirror the outlet to have it going "down", without changing other things in your

--
Good luck
Ivar


Dear Ivar, I ran the analyse as you suggested but nothing changed. I think the curvatures of blades are correct. They are as in the literature. [QUOTE] Hi And are you sure your blade curatures shoud not be inversed, the slope seems odd, for a centrifugal pump. But you can also mirror the outlet to have it going "down", without changing other things in your -- Good luck Ivar [/QUOTE]

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 5 janv. 2012, 10:20 UTC−5
Hi

for the curvature it might be OK, it just puzzled me at first.
To get it solving, I'm not sure what to propose, I got something like that running once in 3.5 a few years ago, not sure where I have it, and I have seen a v4 model pass by, again not sure where to look for that one, normally v4 should be better than V3.5 anyhow ;)

Sorry cannot do much more

--
Good luck
Ivar
Hi for the curvature it might be OK, it just puzzled me at first. To get it solving, I'm not sure what to propose, I got something like that running once in 3.5 a few years ago, not sure where I have it, and I have seen a v4 model pass by, again not sure where to look for that one, normally v4 should be better than V3.5 anyhow ;) Sorry cannot do much more -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 5 janv. 2012, 15:44 UTC−5
Hi,

I am sorry I have not been clearer before. How did you define the inlet velocity? Has it got just a radial component? And if so did you define a vector field or just the overall scalar value? How did you define the rotating domain? Does it just rotate on its own or is it related to the inlet velocity?

There is no boundary mesh, I would recommend one. The overall mesh seems a bit coarse given the high value of rounds per minute.

Finally, in 4.2 you can't virtually get convergence for high rpms, i.e. turbulent flow, I don't remember about 3.5.

Cheers
Hi, I am sorry I have not been clearer before. How did you define the inlet velocity? Has it got just a radial component? And if so did you define a vector field or just the overall scalar value? How did you define the rotating domain? Does it just rotate on its own or is it related to the inlet velocity? There is no boundary mesh, I would recommend one. The overall mesh seems a bit coarse given the high value of rounds per minute. Finally, in 4.2 you can't virtually get convergence for high rpms, i.e. turbulent flow, I don't remember about 3.5. Cheers

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.