CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Modeling a wind turbine (http://www.cfd-online.com/Forums/main/65147-modeling-wind-turbine.html)

psd June 5, 2009 09:11

Modeling a wind turbine
 
What is the best method to model a Darrieus-type vertical axis wind turbine in two dimensions? Basically there are three potential techniques I'm considering: streamtube models, vortex models and FEM. What are the advantages and disadvantages of each technique? Are there any streamtube or vortex codes around? I couldn't find any.

Thanks.

harishg June 5, 2009 15:53

What do you exactly expect to get out of your modeling ? Depending on your requirement you can do anything between a panel method and a full LES.

psd June 6, 2009 10:41

Ultimately, the goal would be to extract the information how much power does the turbine produce. So quantities of interest are the forces on the blades, torques etc.

What are panel method and LES?

wei June 8, 2009 05:46

You can use CFX. Hope this helps a wee bit.

Dai, Y M and Lam, W, Numerical Study of Straight-bladed Darrieus-type Tidal Turbine; Proceedings of the Institution of Civil Engineers, Energy, 2009, 162, No. 2, 67-76.

gocarts June 8, 2009 10:09

Panel Method
 
Quote:

Originally Posted by psd (Post 218439)
What are panel method...?

For a panel method description try:
http://www.symscape.com/blog/why_use_panel_method

For a vertical-axis wind turbine example (Lenz2 specifically) try:
http://www.symscape.com/node/662

Full disclosure: I present Symscape who develop and sell a panel method.

psd June 10, 2009 01:54

Quote:

Originally Posted by gocarts (Post 218598)
For a panel method description try:
http://www.symscape.com/blog/why_use_panel_method

For a vertical-axis wind turbine example (Lenz2 specifically) try:
http://www.symscape.com/node/662


Panel method certainly seems like a good option for an inexperienced modeler. What are its main limitations for this kind of problem? It seems that lots of assumptions about flow conditions are made. Are there any additional resources for more insight on the subject?

It seems that streamtube and vortex methods are still the most used techniques for Darrieus modeling. I haven't found any codes implementing these though and any tips as where to find them would be greatly appreciated.

gocarts June 10, 2009 09:35

Quote:

Originally Posted by psd (Post 218744)
It seems that lots of assumptions about flow conditions are made. Are there any additional resources for more insight on the subject?

An excellent discussion and description of panel methods is available in the book Low Speed Aerodynamics, from Katz and Plotkin.

psd June 17, 2009 05:43

I have access to CFX, Fluent, Elmer and OpenFOAM. Some are based on finite element method and some on finite volume method. Which method suits the problem best? Are Elmer and OpenFOAM on par with Fluent, being open source and all?

Also, the Symscape package seems interesting. But why would one choose that instead of the abovementioned packages? What are its advantages?

gocarts June 18, 2009 09:34

Why use SymLab?
 
Get the marketing out of the way first :)

SymLab Professional is affordable ($300/yr, 50% discount for academic users), easy to learn and easy to use. It incorporates a panel method flow module.

Now for the meat:

A panel method solver is good for predicting lift and pressure drag (no skin friction) for fully attached airfoil flows away from stall (as you'll find for a Darrieus under steady running conditions). Panel methods only require a surface grid unlike finite-element/volume methods which require a volume grid - making panel methods easier to setup and faster in obtaining a solution.

Arbitrary rigid body motion (such as you'll find with a rotating wind turbine) is relatively easy to model in a panel method.

I see panel methods as complimentary to finite-element/volume methods. Panel methods are useful in the concept design phase of streamlined bodies as a precursor to committing to a more thorough and time consuming simulation accounting for viscosity.

Again, for more on the pros/cons of panel methods try http://www.symscape.com/blog/why_use_panel_method

OpenFOAM is on a par with FLUENT in terms of capabilities, at least for your application. However, OpenFOAM lacks an integrated GUI - this is about to change so watch this space: http://www.symscape.com/fourth-openf...-2009-abstract

Full Disclosure: I represet Symscape the developer of SymLab


All times are GMT -4. The time now is 20:18.