Centrifugal Pump Impeller design
Hello friends,
I am a Master's student at University of Cincinnati. My thesis is about Study of Two-phase flows in Centrifugal pump impellers and I am new to CFD. Can anyone please help me out with how to design and mesh an impeller in Gridgen, or can you suggest some tutorials which will be useful in doing so? Tutorials in Solidworks or any other software is also fine. I just want to know the procedure of creating an impeller. Thanks. |
There's a script for that.
Rajiv:
You might try the Kreila import script on Pointwise's Glyph Script Exchange. It reads profile curves for hub, shroud, and blade geometries for impellers and creates Gridgen database entities for them. Hope this helps, Rick |
Thanks Rick. I will let you know if I run into any trouble. Thanks once again for helping me. :):):)
|
What software are you going to be using?
|
I want to use Fluent. Another possibility that I am looking at is OpenFOAM. I just started learning OpenFOAM and I am not sure yet as to how it handles Turbomachinery.
|
I see. I am very experienced in CFD and have gotten reasonably up to speed with CFX. I am getting ready to treat a centrifugal pump problem in CFX. I was hoping that we could help each other but unfortunately I am not familiar with FLUENT or OpenFoam. Good luck with your analysis.
|
Well, one of my lab mates is using CFX and he has done some work on inducer of a centrifugal pump. I have not looked at CFX, so I cannot say anything about it.
I will try to look at that and if I find it easy to learn and use, I will use it. Anyways, keep in touch so that we can share some more information and good luck to you too!!!!! |
Thanks!
If you're deciding between CFX and FLUENT, I understand that CFX is far more user-friendly and fully integrated into Workbench. However, I do not know which one is better for turbomachinery. I would think that CFX would be just fine and the user-friendliness would put me over the fence. I actually just recently evaluated the two and made a recommendation to my company as to which software to purchase, and CFX won out. ANSYS has a Capability Chart which compares the CFX and FLUENT which may be of interest to you. |
Hi all,
Ive been working on pump geometry for the past year, and there is a balanced opinion of preference of FLUENT over CFX and vice versa. OpenFOAM mite give some problems with rotating domain (version 11 too) I am rite now starting work on centrifugal impeller. I am using GAMBIT. Can you please tell me if its a suitable software or is there anything better available??? |
please help me
please anyone have videos or pdf theses for this model to learn from it:confused:
|
you can mesh with gambit and gridgen easily. If needed I can show you method on my computer with the help of skype.
|
centrifugal pump impeller
2 Attachment(s)
Hello,
I have developed unstructured mesh to the centrifugal impeller and I am using CFX solver for it. do u recommend structured grid for impeller? is it difficult to develope hexa mesh in icemcfd- I tried blocking the impeller but I found very difficult. please suggest me Attachment 9784Attachment 9785 I tried a trail please see what went wrong in it |
Although it is not necessary to use the full structured grid for the turbomachinery or any other type of flows. But for turbomachinery flows, structured grid is used for the following reasons
1. It is custom in turbomachinery field to use the structured gird 2. due to flow alignment, the number of nodes can be reduced in the particualr direction 3. In turbomachinery flows, boundary layer is very much important, so you may create a layer of 10-20 cells around blade with structured elements (o-grid) and rest of the domain may be meshed with unstructured elemetns. 4. Keeping in view point 3 above, you can control the Yplus with structured mesh in the vicinity of no slip walls I strongly recomend the structured gird and I think it is not very much difficult. Moreover it is worth to spend some time on high quality meshing than having low confidence in the final results. |
pls also attach the pic of convergence plot
|
convergence
1 Attachment(s)
Mr.
You have provided useful information, on structured grid I am using ICEM-cfd for the meshing. to create o-grid for the complete impeller is the only method. or any method that I could not find can you please suggest some tutorial. please. I have attached the convergence plot to it. please give suggestions.. |
Suggestions plz
HELLO:confused:
|
1. which turbulence model you are using.
2. If you want to mesh in ICEM, I think you must go through blade pipe and wing body tutorial first. Creating mesh for centrifugal compressor is bit tricky. Here are some steps you should follow. a) make the geometry (or create part ) for periodic, casing (enclosed by periodic surfaces), hub (same as shroud), inlet and outlet b) then create block and split them in such a way so that we can associate to periodic, casing, hub, blade , inlet and outlet c) dont forget to project mesh on surfaces |
Quote:
|
1) What is your time step setting?
2) What happens if you continue the run for another say 2-3 hundred iterations? |
TurboGrid
I would recommend CFX as the solver with blade modeler for the geometry creation and TurboGrid for the meshing.
CFX is well known to be the leader in turbo machinary. And while ICEM CFD is a good flexible tool that can be used for turbo machinery, TurboGrid was designed for it. It is automated and features excellent smoothing. You will also find more applicable tutorials for TurboGrid. These are all in Workbench which has its own benefits, including the ability to to a parametric and persistent optimization (if you wanted to head that way). Best regards, Simon |
Quote:
|
That is sort of funny. You would think that after all the optimization that has been done on these pumps the design space would be fully explored. But we have thousands of simulation engineers still working furiously at making a better pump... Every now someone pushes the performance up a little bit or tailors the performance more closely to the needs.
The funny thing about this sort of turbo machinery optimization is that when the analyst gets even a tiny improvement (just a percent or two), they are usually very excited. This is also why they need a good hexa mesh. A tetra prism mesh just has too much noise from the changing meshes and you can not be sure that the 1 or 2 percent isn't just due to the different mesh topology. |
The range of applications is just very large. And most customers these days want pumps that run at 90% efficiency from 10 to 100000 m3/hr, free from cavitation. :)
|
yes, multi-point and multi-objective design optimisation is a hot topic for pumps and turbomachinery in general.
I often refer to this paper for such pump design requirements, it illustrates well how very often design requirements are in the form of trade-off between the different objectives: http://dx.doi.org/10.1115/FEDSM2009-78348 |
yes, multi-point and multi-objective design optimisation is a hot topic for pumps and turbomachinery in general.
I often refer to this paper for such pump design requirements, it illustrates well how very often design requirements are in the form of trade-off between the different objectives: http://dx.doi.org/10.1115/FEDSM2009-78348 |
impeller design with CFX 13
Quote:
following of your above information , i would like your help to design an impeller with CFX 13 , actually i am working on the probleme of study of the fall of performence of a centrifugal pump in a system of two-phase flow (diphasique ) , while the pump of studies is a real pump in petrolium field , i have all the neccessary compenents to acheve the reverse engineering of this pump and processing it in CFX , but also i have probleme to draw the scheme of impeller and get the real angle mesurement . please if you have any idea about this work , send us your proposition . thanks in advance . |
For this purpose you need some reverse engineering machines e.g. 3d scanner;)
|
CFD simulation of centrifugal pump impeller
I am working on CFD simulation of flow through the impeller of centrifugal pump. I am not considering the volute. My problem is that I don't have a clear idea about how much pressure rise should occur in the impeller. I mean, it is just giving kinetic energy to the fluid so that means, static pressure should not rise too much. But in my case, the results are:
Inlet BC: 1.99 bar pressure Outlet BC: 332 kg/s flow rate Wall BC: Shroud, hub and blades From simulation, I get Outlet Pressure: About 9 bar My question is that isn't it too high for an impeller of 48 cm diameter running at 1480 rpm .. ? |
All times are GMT -4. The time now is 03:08. |