Jump to content
Jörg

Centrifugal Impeller Blade Design

Recommended Posts

Hi together,

 

Please find attached a parametric model of a centrifugal impeller post-8-0-62386100-1486040489_thumb.png which was built in CAESES. The project file will probably also be included in version 3.0.11.

 

Note that pretty much everything can be customized in this model and used for manual or automated blade design, such as:

 

  • Merdional contour (i.e. hub and shroud contour)
  • Profile shape by means of user-defined thickness distributions and beta-angle distributions
  • Fillet shape at the hub region by using a constant factor (BTW: this can also be varied with a function along the blade)
  • Ellipse factor of leading edge
  • Size of tip gap
  • Rotational direction
  • Number of blades
  • Thickness of casing

 

The splitter blade is completely decoupled from the main blade in terms of the beta-distribution (so this gives more freedom) - but can be linked to the main blade, too.

In addition, the model comes with some support geometries for meshing/CFD ("periodic boundaries"). These are automatically adjusted to the blade shapes.

 

The camber lines of the blades are generated from the trailing edge which makes it easier to vary the beta distributions (e.g. fix the TE ends of the blades). There is also a feature definition included which generates the camber line with the leading edge as starting point.

 

When working with the model, switch off all scopes that are "downstream" of where you currently manipulate things. For instance, if you want to change the hub and shroud contour, then set the scopes 02_main, 03_splitter and 04_cake invisible. If you want to change the beta distributions, only visualize the sections that are given in 00_sections. This allows interactive changes to the model with a very short response time. You can also switch off the cake part since it is only there for visualization purposes and CFD pre-processing.

 

Let me know what you think about the model (of course, I hope you'll like it....).

 

Cheers

Joerg

 

  • UPDATE AUGUST 2015:
    There are new samples directly included in CAESES that you can use for impeller modeling. See my more recent post below. The project files in the attachment do not work correctly for CAESES 4.x versions, we now have much easier ways to design and control hub fillets.
     
  • UPDATE FEBRUARY 2016:
    I removed the old project file and added an updated CAESES model to this post. Should work with versions >= 4.0.3.
     
  • UPDATE FEBRUARY 2017:
    Note that there are now FULL FREE ACADEMIC versions of the pro edition CAESES for students and PhD students as well as trial licenses with variable time frames. There are also special editions for small companies, start-ups and freelancers.

post-8-0-84744500-1378298807_thumb.png

post-8-0-58349800-1378298816_thumb.png

post-8-0-77280200-1378298924_thumb.png

post-8-0-43621700-1378393260_thumb.gif

FS_CENTRIFUGAL_IMPELLER.pdf

turbocharger_easyuse.fdb

  • Upvote 1

Share this post


Link to post
Share on other sites

Hi Ahmed,

 

it doesn't look like it would be a big problem to model the geometry that you posted with CAESES. Do you have any specific question in that context?

 

If this is just a general question, then I would suggest that you get familiar with CAESES by doing the tutorials, if you haven't already done so. Then you can study Joerg's model in order to get an idea about the modeling of you specific geometry. If questions arise in this process, let us know.

 

Cheers,

Mattia

Share this post


Link to post
Share on other sites

Hi together,

 

With version 3.1, we have an updated impeller model that is also part of the free version. In the GUI, you can find it in the samples section of the documentation browser.

 

The updated model uses the new BRep functionality, which helps a lot for the fillet part at the hub region. In addition, the new meridional camber curve speeds up the refresh process when changing parameters.

Share this post


Link to post
Share on other sites

Here is another parametric model, ready for shape optimization:

 

post-8-0-50605200-1439542275_thumb.png

 

This impeller has been modeled in CAESES, and it has a set of efficient controls! You can intuitively change the meridional contours, as well as the camber and thickness function graphs for global and smooth 3D shape changes. The splitter blade is a separate model and it can be controlled with individual functions.

Share this post


Link to post
Share on other sites

Hi Joerg,

 

Thank you very much for putting together this model. I tried to open the attachment from your original post with my CAESES Free 4.0.3, and the program is struggling in resolving some of the fillet features (see attachment). Do you by any chance have an updated project file that will work with my version of CAESES? I am most interested in the splitter blade implementation, and possibly some demonstration of the asymmetrical thickness distribution. 

 

Best,

Ruhou

post-589-0-06700200-1455742416_thumb.png

post-589-0-59428600-1455742416_thumb.png

Share this post


Link to post
Share on other sites

Hi Ruhuo,

 

Nice artwork, isn't it?

 

Sorry for the issues. Yes, the model from above is a rather old one, and it doesn't work with the newer versions of CAESES. The main reason for this is that we completely replaced our old geometry kernel some time ago, and we couldn't maintain the full backward compatibility. Please find attached an updated model which should give you the basics, for generating your custom design. There might be still some issues with the fillets for certain parameter combinations, but with the upcoming version 4.0.4 this model is stable.

 

Usually, we give specialized blade and impeller training on these kinds of geometries. We've noticed that each company or designer has different requests and features for such a model. The thickness definition as well as the camber definition of this model have a specific design and specific parameters, but the user is free to come up with own functions and definitions. I think you can model almost any kind of e.g. turbocharger impeller with all the options that are available in CAESES.

 

It would be great if you can give me some feedback on this model within the next weeks.

 

Cheers

Joerg

 

post-8-0-65916000-1455782608_thumb.png

turbocharger_easyuse.fdb

Share this post


Link to post
Share on other sites

Hi,

 

I'm just trying to understand the theory behind the beta and theta definitions for a meridional camber curve. I've searched through the documentation and online and haven't seemed to find anything. For example, if I set the beta values, can I calculate in Excel or something what the theta values will be? Basically I want to be able to know what the wrap angle of the blade is. I know the beta values are essentially a slope of what the theta values are and that the theta values simply refer directly to wrap. However, just integrating from a curve fit of the beta values and then attempting to calculate theta doesn't seem to come out the same.

 

This is for performing optimization with ANSYS. I'm using an external optimizer that will send parameters for a blade configuration. So it would be nice to know before even opening CAESES what the wrap angle will be so I can set a constraint that will prevent it from being too large.

 

Thanks!

 

Tim

Share this post


Link to post
Share on other sites

Hi Tim,

 

There is some documentation for the angle definitions, you'll find it at the stream section curve type post-8-0-76315300-1475585587_thumb.png. Basically, when you look from the front of the device (into z.-direction) than theta gives you the rotational angle at a location on the camber curve in the projected XY-plane. Furthermore, the entire theta angle distribution is normalized (for the stream section curve type) i.e. it runs in the interval [0,1] which basically corresponds to the normalized length of the meridional contour. I would recommend to use the stream section curve type for blade design instead of the meridional camber curve. It is faster, and comes with almost the same functionality. The meridional camber curve was developed for special pump design workflows.

 

Check out the newer demo model of the impeller (comes with the installation). We had reviewed it just recently for the version 4.1.2, to make it easier to understand.post-8-0-37510000-1492084724_thumb.png

 

Cheers

Joerg

Share this post


Link to post
Share on other sites

Once the model has been set up in the user interface, the theta or beta distributions as well as any other quantity (thickness parameters, leading edge control, meridional contours) can also be controlled in batch mode. All you need is an *.fsc script (ASCII) that opens the CAESES project, changes the values of the design variables, and exports the geometry using IGES/STEP/STL etc.

 

Here is an example of a *.fsc script: post-8-0-59124000-1492085571_thumb.png

 

For more information about the batch mode of CAESES, see this post. With this, you can use CAESES as a robust geometry engine for your impeller geometries, i.e. to create new design candidates with just a few clicks. The fsc-format is simple to write and parameterize (e.g. from other optimization tools), and running CAESES in the background can be integrated into pretty much any work flow.

 

If you have constraints to fulfill, you can directly embed them into the model (e.g. run internal optimizations to match a quantity such a wrap angle or cross section areas etc). This will give you only feasible designs.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...