[daysley]

Hi Everyone,

Like so many others I've had the joy of using ANSYS to create a simulation of a vertical axis wind turbine using the sliding mesh theory/ technique. However, I found it extremely annoying at first as the tutorials don't focus on such basic 2D problems and it took me a few weeks to actually get the thing working albeit a hit and miss operation. So due to the great help I received from users on CFDonline I'd first like to say thank you for the help and also I'd like to help others with the same problem.

I recently wrote up a tutorial on the basics of geometry modelling, meshing and simulation set-up and am in the process of creating a free website to provide a basis for others to learn from.

The link is www.student-engineer.moonfruit.com, it’s not fully completed yet and due to the lack of experience building websites isn't much to look at but hopefully it will help.

The tutorial I've wrote may not be the best or easiest way but it's what worked for me. I will be adding bit later to help with results processing and stuff but for now it’s just basic elements

Hope it helps

Cheers

[JTDN]

I totally understand what you're going through, i'm doing almost exactly the same thing at University of Sheffield (UK). My mesh also has a rotating circle around the turbine, however the circle itself is mesh differently. Anyway i'm not very experienced with CFD (i do more wind tunnel) and i was wondering, how did you make the circle mesh (around the turbine) rotate? Do you implement it in the solver or the meshing software (im using Fluent for both) ?

Thanks!

[daysley]

Hi JTDN,

Have you set up the geometry and meshing just as i did in the tutorial or similar to it? and is it 2D?
If so open fluent and once the program has loaded go to cell zone boundary conditions. edit the circle surface and click mesh motion check box and enter the coordinates for the center of rotation and the desired rotational speed. then go to boundaryconditions>>circle edge and change the type to interface do this again with the circle surface so you should have two interfaces. Move to interfaces tab>>and click create interface. give it a name and then select circle edge in the first box and circle surface in the second box. change the type to periodic boundary repeats and uncheck the compute offset tickbox. that should set up the meshes to rotate and to check, in the interfaces tab click preview mesh motion and enter a time step size (depending on the angular velocity of your turbine make it quite small so you can see it rotate properly). hopefully that works if not tell me what happens and ill have another look. Good luck

[JTDN]

Originally Posted by daysley

Hi JTDN,

Have you set up the geometry and meshing just as i did in the tutorial or similar to it? and is it 2D?
If so open fluent and once the program has loaded go to cell zone boundary conditions. edit the circle surface and click mesh motion check box and enter the coordinates for the center of rotation and the desired rotational speed. then go to boundaryconditions>>circle edge and change the type to interface do this again with the circle surface so you should have two interfaces. Move to interfaces tab>>and click create interface. give it a name and then select circle edge in the first box and circle surface in the second box. change the type to periodic boundary repeats and uncheck the compute offset tickbox. that should set up the meshes to rotate and to check, in the interfaces tab click preview mesh motion and enter a time step size (depending on the angular velocity of your turbine make it quite small so you can see it rotate properly). hopefully that works if not tell me what happens and ill have another look. Good luck

Hi Daysley

thanks for the reply! It's 2D and actually very similar to yours, but i have a few differences in the geometry (ie an ellipsis around each aerofoil to have an o-grid that captures the BL correctly). However i have an issue i cant seem to get around: in the solver, the Boundary Conditions zones are set automatically by Fluent, and they're not convenient at all for me. I've got walls where i should have interiors, and BC of different nature are set in the same zone, etc. I tried fiddling with the named selections back in the meshing software, but i still couldn't get it to work like i want it to. Is there a way, in the solver, to directly modify the BC zones? ie to choose a BC zone from the geometry, rather than have the solver pick them up automatically?

Thanks

[daysley]

Yes, I have tried that approach too. How did you model the different zones around the airfoil, did you project lines on to the circular turbine domain and then apply edge sizings/ mapped mesh? Also how did you name the different areas, when I used mesh zones around the airfoil, I projected lines on the area, applied edge sizings, then named all the sections together along with turbine circle domain. Have you tried that or have you named them separately? As for fluent I'm not entirely sure how you change them to a zone which is not on the drop down list and found that is due to the naming selections, but I'm just a beginner so not really sure myself.

[JTDN]

I drew ellipses around each aerofoil, projected lines from each AF to its ellipse and used edge sizing and mapped face meshing to get an o-grid. The turbine circle itself is unstructured, with edge sizing and a high refinement. I also have a central shaft with an o-grid around it. In the end, my only named selections were an edge for inlet, an edge for outlet, and the circular surface of the turbine (with everything in it included) as a united naming. As a result, in the solver i had a wall boundary zone for the turbine that included the circle and the ellipses inside it, and a wall boundary zone for the "outer surface" that was just the turbine circle. I converted their type to 'interface' (in the BC tab) and I used them as interface zones 1&2 in the Mesh Interfaces tab. According to the preview the mesh rotates correctly, however it created walls as Boundary Zones 1 and 2, meaning that the fluid does not get inside the turbine. It's odd, because now in the BC tab i have two interfaces and two walls that are actually the same geometry. And the solver treats the turbine as a simple cylinder, flow does not get in. Any idea why this happened?

[Bharatt]

Hey Sam,

How are you.

I am very much happy for what you are doing for the upcoming enginners. It is often difficult to find tutorials for such problems and you are doing a great job by providing all the essentials.

I was searching for tutorials over the internet, but was unlucky and finally ended up here(your post). It is exact the same i am searching for. I am doing vertical axis wind turbine(with 2D mesh) simulation using gambit and fluent. on visiting the blog i found information upto meshing and my search was for fluent inputs. As it is exactly the same what am doing. I am ready with mesh, but lagging of fluent inputs. As i see you are done with the simulations and they worked quite well for you, could you please share a fluent tutorial(step by step procedure/screenshots). i will be so thankful to you. my mail id is brt307@gmail.com


Thanks in advance,

Regards,
Bharat

[daysley]

Originally Posted by Bharatt

Hey Sam,

How are you.

I am very much happy for what you are doing for the upcoming enginners. It is often difficult to find tutorials for such problems and you are doing a great job by providing all the essentials.

I was searching for tutorials over the internet, but was unlucky and finally ended up here(your post). It is exact the same i am searching for. I am doing vertical axis wind turbine(with 2D mesh) simulation using gambit and fluent. on visiting the blog i found information upto meshing and my search was for fluent inputs. As it is exactly the same what am doing. I am ready with mesh, but lagging of fluent inputs. As i see you are done with the simulations and they worked quite well for you, could you please share a fluent tutorial(step by step procedure/screenshots). i will be so thankful to you. my mail id is brt307@gmail.com


Thanks in advance,

Regards,
Bharat

Hi Bharatt,

Thanks for the kind feedback I'm really glad it helped you as I was in the same situation last year. I did mean to update the blog to incorporate the fluent set-up, however i have written a script which I was going to upload. I hope this helps, however please feel free to ask me any questions.

Sam

[AvaShahrokhi]

Hey Sam

thank you so much for the links. I have already started using fluent for modelling vertical axis wind turbine and was so worried as I did not even know where to start until I came across your posts here and felt so happy that finally found something that can help me.

I suppose that I will soon need to get back to you to ask questions as I proceed.

Ava.

[amianjan]

Originally Posted by daysley

Hi Everyone,

Like so many others I've had the joy of using ANSYS to create a simulation of a vertical axis wind turbine using the sliding mesh theory/ technique. However, I found it extremely annoying at first as the tutorials don't focus on such basic 2D problems and it took me a few weeks to actually get the thing working albeit a hit and miss operation. So due to the great help I received from users on CFDonline I'd first like to say thank you for the help and also I'd like to help others with the same problem.

I recently wrote up a tutorial on the basics of geometry modelling, meshing and simulation set-up and am in the process of creating a free website to provide a basis for others to learn from.

The link is www.student-engineer.moonfruit.com, it’s not fully completed yet and due to the lack of experience building websites isn't much to look at but hopefully it will help.

The tutorial I've wrote may not be the best or easiest way but it's what worked for me. I will be adding bit later to help with results processing and stuff but for now it’s just basic elements

Hope it helps

Cheers

Hello Sam,

Thanks for your tutorial. It is helping me a lot.

I am facing a a problem in Fluent, When i check the boundary conditions menu.. I can't find two surface bodies.As a result i cant set up the sliding mesh technique.
I can only find one surface body named "surface_body- rectangle surface".. Its is of "wall"type and when i display its corresponding mesh two vertical lines and the circle is shown in fluent.

What should be the solution to this problem?

[daysley]

Originally Posted by amianjan

Hello Sam,

Thanks for your tutorial. It is helping me a lot.

I am facing a a problem in Fluent, When i check the boundary conditions menu.. I can't find two surface bodies.As a result i cant set up the sliding mesh technique.
I can only find one surface body named "surface_body- rectangle surface".. Its is of "wall"type and when i display its corresponding mesh two vertical lines and the circle is shown in fluent.

What should be the solution to this problem?

Hi amianjan,

I think the problem may be the naming of the selection in the mesh program. sometimes if they are named incorrectly fluent may assume they are different zones than intended (something i'm not entirely sure of myself). If this doesn't work try looking at how you select the bodies for naming i.e. use the body select tool instead of the face select tool?

let me know how you get on, as this was a problem i had that either disappeared on its own or i had to make the model again?

[luilly3d]

Hi! just want to thank for the information you are providing for the VAWT simulation. I'm currently working on a VAWT project and your tutorial gave me the information to aim my work.

[shahizare]

Originally Posted by daysley

Hi Bharatt,

Thanks for the kind feedback I'm really glad it helped you as I was in the same situation last year. I did mean to update the blog to incorporate the fluent set-up, however i have written a script which I was going to upload. I hope this helps, however please feel free to ask me any questions.

Sam

dear Mr sam

thank you for kindly helping . Im also working on the VAWT i faced with a tiny problem . i dont know the swept area in VAWT in 2d simulation and what is the value of the length and area in fluent for refvalues?
finally how can i calculate the Cp after generate the momentum from fluent
thanks in advanced

[shafqat]

i just want to know why we have used two circles in geometry.thank u

[daysley]

Originally Posted by shafqat

i just want to know why we have used two circles in geometry.thank u

Hi Shafqat, the reason is one geometry becomes a rectangle with a circle missing and the other zone is the rotating circular zone. Without the two circles in the geometry you would have one rectangular zone with no "edge" to act as an interface.

[shafqat]

so when i try to make surfaces from sketch i would make surfaces to both circle?or just one?

[shahizare]

Originally Posted by daysley

Hi Everyone,

Like so many others I've had the joy of using ANSYS to create a simulation of a vertical axis wind turbine using the sliding mesh theory/ technique. However, I found it extremely annoying at first as the tutorials don't focus on such basic 2D problems and it took me a few weeks to actually get the thing working albeit a hit and miss operation. So due to the great help I received from users on CFDonline I'd first like to say thank you for the help and also I'd like to help others with the same problem.

I recently wrote up a tutorial on the basics of geometry modelling, meshing and simulation set-up and am in the process of creating a free website to provide a basis for others to learn from.

The link is www.student-engineer.moonfruit.com, it’s not fully completed yet and due to the lack of experience building websites isn't much to look at but hopefully it will help.

The tutorial I've wrote may not be the best or easiest way but it's what worked for me. I will be adding bit later to help with results processing and stuff but for now it’s just basic elements

Hope it helps

Cheers

Hi My firend
I'm really faced with terrible problem on the validate my simulation , i dont know what is the problem . i try so many meshes i think the problem is in the reference values . could you please help me ? can you give me your mail address ill send my case file ,
regards

[AnvGang]

Hello Sam,

I thought I'd thank you. Your tutorial helped me in my project, but i'm still facing a problem. The turbine is rotating off-centre and I have no clue as to how to fix it. It'd help me alot if you could advise me on how to fix this

[daysley]

Hi AnvGang,

I'm glad my tutorial has helped you I struggled with the same problem myself. Make sure the origin of the rotating zone/ angular velocity is set at the Centre of the turbine blades. How did you make the turbine is the turbine center at 0,0 or at an arbitrary point within the modelling domain? Whichever, make sure this is where the domain is set to rotate about :-) good luck

Regards
Sam

[kartikeya]

how to do post processing in cfd on wind turbine