Freecad CfdOF Tutorial #7: Wind Turbine

[@monkeywrench1951]

Came here from the Freecad forum. Thank you for showing the rotatingWallVelocity for cfdOF in such good detail.

[@fernandocastillo3427]

hey man, excellent videos. These are helping me so much! Thanks!

[@V1rusFPS]

Great video!

[@rahmatkermani1522]

Great video, Bernd. How would you predict the torque and pressure rise across the rotor in cfdof?

[@sandinugroho2598]

What is the computer spec are you using in this project?

[@LT72884]

Excellent video:) i have seen many videos on KZbin done by others, which were excellent as well, but this is the first time i have seen someone put a boundary condition on the turbine itself? Whats the purpose of this boundary? Thanks

[@KyleDunnIt]

Thank you for taking the time to make this and share it. I'm curious, can openfoam solve for the rotational velocity and/or torques on the blades or is that outside the scope of what it is for?

[@techbernd5535]

Hi Kyle Dunn, in my video tutorial, the rotational velocity is a boundary condition, the torque is a result. In your question, you asked for the blades. I evaluated the torque on the rotor, but of course you can also evaluate the force and the torque of one single blade. Openfoam solves the equation for the flow of a medium, the rotational speed of the blades can be regarded as a constraint of the flow. The result of this constraint is a force, multiplied with the length of the lever, you obtain the torque. A reversed calculation - torque as a constraint and rotational speed as a result - is not feasible for mathematical and physical reasons.

[@KyleDunnIt]

@@techbernd5535 I appreciate you taking the time to explain that. Makes perfect sense.

[@amir2012ful]

Hi Bernd, Thank you so much for these videos, they are very helpful. I am just having a problem doing this tutorial. For some reason when I run the simulation with a element size of 20 mm for the main meshing part, I get the the following error ''simulation exited with error''. If I reduce the element size to let's say 15mm, the simulation starts and I can see the convergence curves but the residuals seem to reach a constant value without fulfilling the residual threshold. Do you have any idea what might be the problem since you manage to get it to converge with a larger element size.

[@techbernd5535]

Hi amir2012ful, thank you very much for your question. The right place for your question is the FreeCad-Forum: forum.freecad.org/. You will find there a category for CfdOf: forum.freecad.org/viewforum.php?f=37&sid=9fbeec31bfc45c38a50a1400f846b2e0. In the FreeCad-Forum, you can upload your model. Many other experts like thschrader and oliveroxtoby - the programmer of CfdOf - can help you. Another helpful function of the FreeCad-Forum is the possibility to upload pictures (for example screenshots) which is very helpful for questions and answers.

[@vaughanthompson3717]

Many thanks for an excellent tutorial Bernd. It has really helped me with my wind turbine simulations. Do you know how this method with a moving wall blade compares with the more complex method where the blade is embedded in a rotating mesh? Also, it would be great if you could do a tutorial on an actuator disk representation of a turbine.

[@techbernd5535]

Hi vaughanthompson, thank you for your positive feedback 👍. I chose this model because it is the easiest and fastest way to simulate a wind turbine in CFD. The simulation is steady-state, the effect of the small turbulences is represented by RANS. In FreeCad version V0.20 the evaluation of the torques is already supported by FreeCad CfdOf, I hope that in version V0.21 it will also be possible to enter a "RotatingWall" - Boundary Condition (BC). Then there is no need to edit the input files. The simulation with rotating mesh is of course closer to reality, but the model then becomes more complex. If I find time, I will work on this topic and make a tutorial. I'm also planning a tutorial about ReactingFoam. The actuator disk representation of a turbine sounds interesting. As far as I understand this method, this method does not use Cfd / OpenFoam, nor is it compatible with FreeCad. Thus, this method is not in my scope.

[@vaughanthompson3717]

@@techbernd5535 Dear Bernd Thanks Bernd. Since my last comment, I have done some "actuator disk" simulations in CfdOf using the porous zone facility. By varying the zone Darcy parameters, it is possible to represent the change in axial momentum of a wind turbine through the pressure drop accross the zone. Hence it is possible to estimate the variation of power output with axial induction (essentially the change in air speed through the disk) and this gives good agreement with the analytical Froude method and predicts the Betz peak efficiency quite well -- but both methods only apply to ideal turbines. Both the actuator disk and Froud approach obtain the turbine power output from axial momentum changes only and do not represent the angular momentum imparted to the wake. For my studies, I am interested in the performance of ducted wind turbines with contra-rotating blades and this limitation is not serious. I have also been looking at the analytical BEM (Blade Element Momentum) method which applies to non-ideal turbines and does take into account the angular momentum. It requires a knowledge of the lift and drag coefficients of the turbine blades. I was able to get good agreement between this approach and experimental results for a 1.75 m diameter commercial wind turbine. I also compared these results with the rotating wall method in your tutorial - by modelling the three-bladed commercial turbine in CfdOf over a range of rotor speeds. There is good agreements between the rotating wall and BEM results at low RPMs but the rotating wall approach appears to under-predict torques at higher RPMs. I believe that this is due to the fact that the method cannot capture the change in angle of attack as the rotor speed increases (the positive blade torque due to lift is independent of angular speed whilst the negative torque due to drag increases). The method does however impart angular momentum to the wake and may be used to examine the effect of turbine wake on other turbines located downstream over a wider range of RPMs. I hope this helps, and once again thanks very much for your illuminating tutorials.

[@poulmack]

Great tutorials, I wonder if you can do a tutorial for HVAC simulations

[@techbernd5535]

Hi @poulmack, thank you very much for your question. I am a mechanical engineer and I have never performed HVAC simulations. I even had to google the abbreviation "HVAC". My advise to you is to ask this question in the FreeCad Forum, sub-Forum "CfdOF. In ths forum you will please find a lot of experts who can give you advice for HVAC and who do not need to google this abbreviation. If you ask your question in the FreeCad-Forum, I would advise you to detail your questions more. Are there any standarts you are referring to? Do you have pictures to graphically support your request? Do you plan to calculate a special purpose? The address of the FreeCad-Forum is: forum.freecad.org/viewforum.php?f=37

[@GrizleTheStick]

Do you have a location to obtain the PDF tutorial you reference?

[@techbernd5535]

Hi GrizelTheStick! You can find the PDF tutorial here: kzbin.info?event=comments&redir_token=QUFFLUhqbTQyQ0MtUEt4M3pyTndwMmZaalZ3SlpacnRZd3xBQ3Jtc0ttNVY1bEkyTnFtWFBLWDF2cURqNzBKRE1UalpGWTN1eTU3SFNfdGpfeEVrazZqZ0xZQ3JqQTR6NU9iV0NmQWEzbEVPZEZUTnZ1T3diRE1CWjVraWl4dUhuQXJhcHhGSUxUR0w5Q25OQkl3X2N1LVNyYw&q=https%3A%2F%2Fgithub.com%2Fopensimsa%2Fopensim%2Fraw%2Fmaster%2FDocumentation%2FCFD%2FUAV2%2FCFD%2520Tutorial%25204%2520-%2520External%2520Aerodynamics%2520of%2520a%2520UAV.pdf&stzid=Ugy0R7D8t1h0CFf64yV4AaABAg.9bcTQqWOnbe9bcZcy4exWp If you use the version V0.20 of FreeCad, this workaround is not necessary any more, the frontend of FreeCad offers you a "reporting function". With this function, FreeCad CfdOf creates a "controlDict" - File with all add-ons necessary.

[@hafizuddinmohdlowhim8426]

Hi how do you know all these tutorials? It seems you have a textbook for this.

[@techbernd5535]

@hafizuddinmohdlowhim8426: Thanks for the compliment. Normally, I practice the simulation 1x or 2x, and then I can record. If I make too many errors, I can delete the sequence afterwards and re-record it.

[@hafizuddinmohdlowhim8426]

@@techbernd5535 so you do it by trial and error?

[@weiquankhoo2345]

what's the purpose of the refinement mesh?

[@techbernd5535]

It would be very advisable if you asked questions of this kind in the FreeCad forum. The structure of the FreeCad forum makes it easier to answer your questions in a structured manner. In the FreeCad forum, not only I, but also the other experts can give you a qualified answer. You can find the FreeCad forum here: forum.freecad.org/viewforum.php?f=37