Wind Turbine - Rotation of the Baldes

moonomid · April 12, 2011, 15:06

Dear Friends
I've a problem with the rotation of the blades in Wind turbine. My friend and I have designed a wind turbine in design modeler.
I want to know how I can rotate the blades when the wind is as the input?
for example the speed of the wind is 30 m/s and I want the blades rotating spontaneously by affection of the wind.
I've searched also and I've found this topic:
http://www.cfd-online.com/Forums/cfx...imulation.html
but I'm confused a little about the boundaries, hub, interfaces and ... for the blades and other parts.
I've done the tutorials in ANSYS which are about the turbo mode:
1- Axial Turbine Equilibrium and Non-Equilibrium Steam Predictions
2- Flow in an Axial Rotor/Stator
but as you know the files were ready (meshes) and I couldn't find the ways for learning the domains and boundaries and ...
"actually my first problem is how to define the domains, boundaries and interfaces that are suits for the turbo mode?
and the second problem how can I rotate the blade by influence of wind?"

This clip is showing what exactly I want:
http://www.youtube.com/watch?v=nj_iL8PXOD8

Thanks

This is the shape:

ghorrocks · April 12, 2011, 18:33

Your domain boundaries in the image look pretty good to me.

Do you want to define the rotation speed as a function of inlet speed, or do you want the simulation to find the rotation speed based on the wind?

ghorrocks · April 12, 2011, 18:34

And another point - you may well be able to model this with a single blade as a frozen rotor approach. This means you only need to model a single blade and that will simplify things immensely.

cfd_newbie · April 13, 2011, 00:55

Quote:

Originally Posted by ghorrocks

And another point - you may well be able to model this with a single blade as a frozen rotor approach. This means you only need to model a single blade and that will simplify things immensely.

Glenn is right here.
There are many papers which have proved that you can very well model a single blade of wind turbine without nacelle tower etc. If you want I can give some references for it.

I suggest you should take a look at the following papers for HAWT, which will give you valuable information about wind turbine simulations.

"Predicting 2D Airfoil and 3D Wind Turbine Rotor Performance using a Transition Model for General CFD Codes",
R. Langtry, J. Gola and F. Menter, ANSYS CFX, Otterfing, Germany, AIAA-2006-0395 44th AIAA Aerospace Sciences Meeting and Exhibit

and

http://wwweng.uwyo.edu/mechanical/fa...-2009-1221-908

Caution - CFD analysis of wind turbine is a tough nut to crack according to my experience.
Raashid

moonomid · April 13, 2011, 02:49

Quote:

Originally Posted by ghorrocks

Your domain boundaries in the image look pretty good to me.

Do you want to define the rotation speed as a function of inlet speed, or do you want the simulation to find the rotation speed based on the wind?

I want to define the rotation of the 3 blades based on the wind. I mean that I input the wind as the Inlet (with speed 30 m/s) and I want the 3 blades rotate by the wind.

I've read in one topic that we have to define two types of domain: Stationary and Rotating but when I want to use the turbo mode I don't know how can I define the domains for the wind and the blades. And how can I define that the rotation of the blades depend on the wind.

Thanks

moonomid · April 13, 2011, 03:09

Quote:

Originally Posted by ghorrocks

And another point - you may well be able to model this with a single blade as a frozen rotor approach. This means you only need to model a single blade and that will simplify things immensely.

Yes, I know but the main subject is the tower. I want to calculate the effective Force (pressure) from the blades on the tower (as an input).

jbritton · April 13, 2011, 07:31

It looks like the way to you want to do this, you want to find the speed at which the rotors will rotate at 30m/s first, then impose this on your current model and run an FSI with the rotational load and wind load imposed on the model to see how the tower reacts.

Do you have any data for the blades, i.e. thrust curves, profiles etc. if not you may have to carry out some preliminary simulations/calcs to get an estimate for the rotational speed of the turbine at 30m/s.

But also most HAWTS lock down their blades at higher velocities, such as this, so you may not have to take rotation in to account and can simply apply the wind loading into a simple FSI

moonomid · April 14, 2011, 02:56

Quote:

Originally Posted by jbritton

It looks like the way to you want to do this, you want to find the speed at which the rotors will rotate at 30m/s first, then impose this on your current model and run an FSI with the rotational load and wind load imposed on the model to see how the tower reacts.

yes, exactly!

Quote:

Do you have any data for the blades, i.e. thrust curves, profiles etc. if not you may have to carry out some preliminary simulations/calcs to get an estimate for the rotational speed of the turbine at 30m/s.

Unfortunately, I don't have too much information about wind turbine. only the size the with turbine and the weight of each part.

Quote:

But also most HAWTS lock down their blades at higher velocities, such as this, so you may not have to take rotation in to account and can simply apply the wind loading into a simple FSI

Could you please explain more?

Regards

jbritton · April 14, 2011, 04:35

At high wind speeds most HAWTS have their rotation lock so the turbine no longer turns, I believe some also turn the blades out of the direction of the wind too. In order to reduce the forces on the blades and tower.

You look to have the geometry of the blades already, you will probably have to run a quick analysis on 1 blade to work out its Cl if you have no profile or other data. From this you can work out the rotation speed at 30m/s

moonomid · April 14, 2011, 04:46

Quote:

Originally Posted by jbritton

At high wind speeds most HAWTS have their rotation lock so the turbine no longer turns, I believe some also turn the blades out of the direction of the wind too. In order to reduce the forces on the blades and tower.

You look to have the geometry of the blades already, you will probably have to run a quick analysis on 1 blade to work out its Cl if you have no profile or other data. From this you can work out the rotation speed at 30m/s

Thanks for the information but at the moment my problem is how I can rotate the blades by wind in CFX not the rotation lock.
This clip shows everything that I need.
http://www.youtube.com/watch?v=nj_iL8PXOD8
as you can see the blades is starting to rotate by the wind.

jbritton · April 14, 2011, 04:56

yes my guess that simulation was done using a cel function to define the rotational velocity of the blades wrt the wind speed as Glenn suggested above.

You cant do this because you don't know how fast your turbine will rotate at any given wind speed, you need to figure this out.

You do this by either looking for further data on this HAWT in particular, or by figuring it out yourself through further analysis/calcs based of the blade geometry you have.

moonomid · April 17, 2011, 05:48

Quote:

Originally Posted by jbritton

yes my guess that simulation was done using a cel function to define the rotational velocity of the blades wrt the wind speed as Glenn suggested above.

Actually, I have a problem with this part. I've read this topic http://www.cfd-online.com/Forums/cfx...imulation.html
and I've found that first I have to obtain the torque on blades and then use a CEL to specify the rotation rate as the function of the torque. But the person used the FLUENT for calculating the torque! The company doesn't have any license for FLUENT, so how can I calculate the torque in CFX?

or how can I define the blades rotation by the wind velocity as the input?

ghorrocks · April 17, 2011, 07:29

Getting CFX to find the operating speed like this is likely to be difficult and have convergence problems. Maybe somebody has done it and can say otherwise, but this sounds like the hard way to me.

Far easier, but a little less elegant is to sweep through a range of rotation speeds in a series of steady state runs. This will give you a torque/power versus speed curve. Then interpolate on this curve your intended torque/power and there you have it. If you like then run this point as well to improve the accuracy of the interpolation.

This approach is much easier and can be automated in Workbench or a simple script file. It will not need much user intervention.

Matjaz · April 17, 2011, 16:18

Hello, it is quite easy thing to do in Ansys CFX v13 using 6dof solver, where the blades are modeled as a rigid body. There you can also specify load of the generator as an expression or a constant value so they won't reach runaway speed. It is all well documented in tutorial nr. 31 (the last one). So just adapt this model in tutorial for your problem.
I'm also dealing with such an issue going the old way, using fortran routine to calculate an Angular velocity for each step, based on torque on the blades from previous step, but i stil haven't found a way to keep a variable that would sum up my angular velocity value - what i get from fortran routine are delta values of it. Some monitoring function, such as Rotation Velocity@ blades is also not allowed for using in Angular velocity for MFR domain.
Regards, Matjaz

moonomid · May 11, 2011, 14:42

Dear friends
I have a problem with the rotor rotation in Structural part.
when the rotor is rotating (I've obtained from CFX) in the structural part, the Nacelle is rotating,too. how can I define that the Nacelle is fixed to the tower and the rotor is allowed to rotate. I couldn't find any contacts or constraints for rotation in the structural part.
my question maybe a little confusing. consider that I have a rotor and a Nacelle and I want to define that the rotor is rotating 360 degree and the Nacelle is fixed. So, how can I do it in Structural part?
Best regards

moonomid · May 19, 2011, 14:28

Dear friends
In one way FSI (Transient) , I define the rotor as a wall in CFX then I import the pressure on this wall (Rotor) to the solid rotor but I don't know these pressures are correct or not because my solid rotor is not rotating in the structural . Therefore, Do I need to rotate the rotor in structural for getting the correct result?

saleh alsubari · August 9, 2012, 21:50

Dear Friends
I've a problem with the rotation of the blades in water turbine.
I want to know how I can rotate the blades when the mass flow rate is as the input? and to find the rpm and the torque associated with it ,using star-ccm+
? please help

ghorrocks · August 10, 2012, 06:39

This is the CFX forum. Try the Star-CD forum for star.

But you will find things much easier if you assume a speed and work out the torque. Sweep over a range of speeds and find the point where torque equals the output load. That is your operating point.

saleh alsubari · August 11, 2012, 02:38

Quote:

Originally Posted by ghorrocks

This is the CFX forum. Try the Star-CD forum for star.

But you will find things much easier if you assume a speed and work out the torque. Sweep over a range of speeds and find the point where torque equals the output load. That is your operating point.

thanx alot ...

aghsin · July 4, 2013, 19:50

Hello everybody,
did any body get success in simulating wind turbine from start up? I m trying to rotate blades regarding the torques exerted by wind. but I havent been successful yet. I inserted the rotationa speed of the rotating domain as a function of Torque, Blade mass Inertia ans time, (omega=(torque/Iz)*dt). but my simulation crashes after a couple iteration with return code 252,
Any comment?

April 12, 2011, 18:33		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,701 Rep Power: 143	Your domain boundaries in the image look pretty good to me. Do you want to define the rotation speed as a function of inlet speed, or do you want the simulation to find the rotation speed based on the wind? renzo.ml likes this.

April 13, 2011, 07:31		#7
jbritton Member james britton Join Date: Jan 2010 Posts: 38 Rep Power: 16	It looks like the way to you want to do this, you want to find the speed at which the rotors will rotate at 30m/s first, then impose this on your current model and run an FSI with the rotational load and wind load imposed on the model to see how the tower reacts. Do you have any data for the blades, i.e. thrust curves, profiles etc. if not you may have to carry out some preliminary simulations/calcs to get an estimate for the rotational speed of the turbine at 30m/s. But also most HAWTS lock down their blades at higher velocities, such as this, so you may not have to take rotation in to account and can simply apply the wind loading into a simple FSI Last edited by jbritton; April 13, 2011 at 11:06. Reason: so it made sense

May 11, 2011, 14:42		#15
moonomid New Member omid Join Date: Apr 2011 Posts: 9 Rep Power: 15	Dear friends I have a problem with the rotor rotation in Structural part. when the rotor is rotating (I've obtained from CFX) in the structural part, the Nacelle is rotating,too. how can I define that the Nacelle is fixed to the tower and the rotor is allowed to rotate. I couldn't find any contacts or constraints for rotation in the structural part. my question maybe a little confusing. consider that I have a rotor and a Nacelle and I want to define that the rotor is rotating 360 degree and the Nacelle is fixed. So, how can I do it in Structural part? Best regards Last edited by moonomid; May 12, 2011 at 02:13.

August 10, 2012, 06:39		#18
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,701 Rep Power: 143	This is the CFX forum. Try the Star-CD forum for star. But you will find things much easier if you assume a speed and work out the torque. Sweep over a range of speeds and find the point where torque equals the output load. That is your operating point. saleh alsubari likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
please help in simulating a wind turbine blade in 6 DOF ????....	usman naseer	Main CFD Forum	2	March 10, 2016 02:53
Ducted wind turbine (BC for the shroud)	Pepita	CFX	4	June 29, 2013 07:09
3D wind turbine modelling/boundary condition defn in rotation	bakay	FLUENT	9	September 14, 2011 05:32
Wind turbine simulation	Saturn	FLUENT	1	June 16, 2006 02:12
Wind turbine simulation	Saturn	Main CFD Forum	1	June 12, 2006 03:57

April 12, 2011, 18:34		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,701 Rep Power: 143	And another point - you may well be able to model this with a single blade as a frozen rotor approach. This means you only need to model a single blade and that will simplify things immensely.

April 14, 2011, 04:35		#9
jbritton Member james britton Join Date: Jan 2010 Posts: 38 Rep Power: 16	At high wind speeds most HAWTS have their rotation lock so the turbine no longer turns, I believe some also turn the blades out of the direction of the wind too. In order to reduce the forces on the blades and tower. You look to have the geometry of the blades already, you will probably have to run a quick analysis on 1 blade to work out its Cl if you have no profile or other data. From this you can work out the rotation speed at 30m/s

April 14, 2011, 04:56		#11
jbritton Member james britton Join Date: Jan 2010 Posts: 38 Rep Power: 16	yes my guess that simulation was done using a cel function to define the rotational velocity of the blades wrt the wind speed as Glenn suggested above. You cant do this because you don't know how fast your turbine will rotate at any given wind speed, you need to figure this out. You do this by either looking for further data on this HAWT in particular, or by figuring it out yourself through further analysis/calcs based of the blade geometry you have.

April 17, 2011, 07:29		#13
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,701 Rep Power: 143	Getting CFX to find the operating speed like this is likely to be difficult and have convergence problems. Maybe somebody has done it and can say otherwise, but this sounds like the hard way to me. Far easier, but a little less elegant is to sweep through a range of rotation speeds in a series of steady state runs. This will give you a torque/power versus speed curve. Then interpolate on this curve your intended torque/power and there you have it. If you like then run this point as well to improve the accuracy of the interpolation. This approach is much easier and can be automated in Workbench or a simple script file. It will not need much user intervention.

April 17, 2011, 16:18		#14
Matjaz New Member Matjaz Fleisinger Join Date: Mar 2011 Posts: 17 Rep Power: 15	Hello, it is quite easy thing to do in Ansys CFX v13 using 6dof solver, where the blades are modeled as a rigid body. There you can also specify load of the generator as an expression or a constant value so they won't reach runaway speed. It is all well documented in tutorial nr. 31 (the last one). So just adapt this model in tutorial for your problem. I'm also dealing with such an issue going the old way, using fortran routine to calculate an Angular velocity for each step, based on torque on the blades from previous step, but i stil haven't found a way to keep a variable that would sum up my angular velocity value - what i get from fortran routine are delta values of it. Some monitoring function, such as Rotation Velocity@ blades is also not allowed for using in Angular velocity for MFR domain. Regards, Matjaz

May 19, 2011, 14:28		#16
moonomid New Member omid Join Date: Apr 2011 Posts: 9 Rep Power: 15	Dear friends In one way FSI (Transient) , I define the rotor as a wall in CFX then I import the pressure on this wall (Rotor) to the solid rotor but I don't know these pressures are correct or not because my solid rotor is not rotating in the structural . Therefore, Do I need to rotate the rotor in structural for getting the correct result?

August 9, 2012, 21:50		#17
saleh alsubari New Member saleh Join Date: Aug 2012 Posts: 25 Rep Power: 13	Dear Friends I've a problem with the rotation of the blades in water turbine. I want to know how I can rotate the blades when the mass flow rate is as the input? and to find the rpm and the torque associated with it ,using star-ccm+ ? please help

July 4, 2013, 19:50		#20
aghsin New Member sina Join Date: Jul 2013 Posts: 21 Rep Power: 12	Hello everybody, did any body get success in simulating wind turbine from start up? I m trying to rotate blades regarding the torques exerted by wind. but I havent been successful yet. I inserted the rotationa speed of the rotating domain as a function of Torque, Blade mass Inertia ans time, (omega=(torque/Iz)*dt). but my simulation crashes after a couple iteration with return code 252, Any comment?