CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Wind Turbine Fluent setup. Is it correct? (http://www.cfd-online.com/Forums/fluent/96942-wind-turbine-fluent-setup-correct.html)

federvo.mala February 5, 2012 20:31

Wind Turbine Fluent setup. Is it correct?
 
Dear all,

I am trying to simulate the NREL Phase VI experiment. Here is the setup I have done:

http://i.imgur.com/pC8TF.jpg



1) General

http://i.imgur.com/IfX1I.jpg

2) Cell zone conditions

I am treating the whole mesh as a whole Single Moving Reference Frame

http://i.imgur.com/alu36.jpg

3) Boundary conditions

a) inlet - velocity

The frame is absolute, right? Plus I am not sure what to put in the initial pressure gauge value. I only know the velocity value.

http://i.imgur.com/KGLuT.jpg

b) Blade/rotor

The blade is set a rotating wall http://i.imgur.com/ngbs5.jpg
http://i.imgur.com/ngbs5.jpg
c) Far field

Note the difference in the frame. For the far field I have used absolute frame ( as specified by the user's guide)

http://i.imgur.com/GwIT6.jpg

d) Outlet: pressure outlet with 0 gauge pressure

4) Solution initialization

Is the reference frame correct?
http://i.imgur.com/DQ8i0.jpg


To conclude, the things I am not really sure about are:

With this setup, the far field is affecting the whole flow because is set as wall, how can I set it to not influence the far field, something like open atmosphere or maybe a free-shear wall?

Have I set the Moving refererence frame correctly?

And what about the initial gauge pressure at the inlet bc?


Any help is greatly welcomed as I am a bit stuck and would really need some experienced help.

Thanks so much,
Federico

zhenglun.wei February 6, 2012 17:00

3 Questions for you
 
Dear Federico,

According to your questions,
With this setup, the far field is affecting the whole flow because is set as wall, how can I set it to not influence the far field, something like open atmosphere or maybe a free-shear wall?
A: As I know, outflow is good option. Conceptually, it just let the flow go out.

Have I set the Moving refererence frame correctly?
A: To my limited knowledge, it is correct. I tried it in a turbine simulation in aircraft engine case before.

And what about the initial gauge pressure at the inlet bc?
A: No idea on this one.

After reading your thread, I have 3 questions:
1, I can not see the Fig.1;
2, Where did you get the mesh file for the NREL wind turbine? Is there any open source for that?
3, Why do you want to use wall roughness = 0? Are you assuming that the wind turbine is pure smooth?


best,
Alan

federvo.mala February 7, 2012 08:49

Quote:

Originally Posted by zhenglun.wei (Post 342982)
Dear Federico,

According to your questions,
With this setup, the far field is affecting the whole flow because is set as wall, how can I set it to not influence the far field, something like open atmosphere or maybe a free-shear wall?
A: As I know, outflow is good option. Conceptually, it just let the flow go out.

Have I set the Moving refererence frame correctly?
A: To my limited knowledge, it is correct. I tried it in a turbine simulation in aircraft engine case before.

And what about the initial gauge pressure at the inlet bc?
A: No idea on this one.

After reading your thread, I have 3 questions:
1, I can not see the Fig.1;
2, Where did you get the mesh file for the NREL wind turbine? Is there any open source for that?
3, Why do you want to use wall roughness = 0? Are you assuming that the wind turbine is pure smooth?


best,
Alan


Dear Alan,

First of all thanks for your reply.

1. I will upload fig.1 again although it's not important at all.

2. I built the geometry on solidworks and then created the mesh on ICEM CFD. It is an unstructured mesh with prism layers around the blade. I can give you the geometry if you want.

3. I set the blade as no slip. I though this would be only straightforward option.

Here some other questions for you:

1. What bc shall I use for the outlet? Do you think that pressure outlet at 0 pascal would do the job?

If you need any more info regarding the NREL let me know.

Thanks,
Fred

zhenglun.wei February 7, 2012 12:17

I always try the 'Outflow' first. You may try it to see if there is any problem like 'reversed flow'. If no, you can use 'Outflow' for simplicity. To my experience, I use 'Pressure Outlet' only when I come across unexpected 'reverse flow' problem. Hope this helps. ^_^

Alan

federvo.mala February 7, 2012 16:25

Quote:

Originally Posted by zhenglun.wei (Post 343159)
I always try the 'Outflow' first. You may try it to see if there is any problem like 'reversed flow'. If no, you can use 'Outflow' for simplicity. To my experience, I use 'Pressure Outlet' only when I come across unexpected 'reverse flow' problem. Hope this helps. ^_^

Alan

Hey Alan thanks again,

and you said outflow for the farfield too?

I have tried giving it 'velocity-inlet' with the same value as the free-stream velocity, with the difference that instead to 'normal to the boundaries' I put it parallel to the boundaries. It seems to work too but it may be wrong too :S

Thanks

zhenglun.wei February 7, 2012 18:07

^_^ I tried that to deal with unexpected 'reversed flow' issues before. It helps.

Alan

federvo.mala February 8, 2012 22:33

Quote:

Originally Posted by zhenglun.wei (Post 343228)
^_^ I tried that to deal with unexpected 'reversed flow' issues before. It helps.

Alan

OK, if then you use outflow for both outlet and far field, how do you set the 'mass flow weighting' ? 1 for both?

zhenglun.wei February 8, 2012 23:07

Quote:

Originally Posted by federvo.mala (Post 343490)
OK, if then you use outflow for both outlet and far field, how do you set the 'mass flow weighting' ? 1 for both?

Yes, I just assume the flow is equally devided for all out flow boudaries.

federvo.mala February 8, 2012 23:29

Quote:

Originally Posted by zhenglun.wei (Post 343494)
Yes, I just assume the flow is equally devided for all out flow boudaries.

Makes sense. Cheers

zaheerd21 March 24, 2012 03:40

Dear Federvo,

I am trying to do this same simulation in Fluent but am also stuck. Did you get yours to work? If so, then could you please advise me on how you changed yours from the pictures you have shown above? Also could you please advise me on which turbulence model you used and what discretisation schemes you used?

Thank you.

m5edr March 25, 2012 21:22

Blade B.c
 
Hi All
I also asking about the Blade B.c

why federvo set the Blade >> rotating wall (relative to adj.) instead of setting it "stationary"

(the default in rotating reference frame, that walls rotate with
the grid, and hence are moving.)

thanks

aqstax March 27, 2012 21:36

Quote:

Originally Posted by zaheerd21 (Post 351214)
Dear Federvo,

I am trying to do this same simulation in Fluent but am also stuck. Did you get yours to work? If so, then could you please advise me on how you changed yours from the pictures you have shown above? Also could you please advise me on which turbulence model you used and what discretisation schemes you used?

Thank you.

The following thread may be of use. http://www.cfd-online.com/Forums/flu...i-turbine.html

monaya flower June 2, 2012 06:31

Dear Federvo,
1- i think that the far wall must be symmetry in the bc
2-in MRF you should use relative velocity formulation
3- for the blade condition i think it is right

i want to ask about why you use intemittency , k and omiga rather than intenesity and viscosity ratio

and if you get accurate result how did you calculate the power extracted from the wind turbine
thank you

federvo.mala June 9, 2012 19:02

Quote:

Originally Posted by monaya flower (Post 364404)
Dear Federvo,
1- i think that the far wall must be symmetry in the bc
2-in MRF you should use relative velocity formulation
3- for the blade condition i think it is right

i want to ask about why you use intemittency , k and omiga rather than intenesity and viscosity ratio

and if you get accurate result how did you calculate the power extracted from the wind turbine
thank you

HI

1 - yes you are right. Changed to symmetry in my final model.

2 - probably right, but I always used SRF

3 - :)

I indeed used k-w with intensity and viscosity ratio specs.

I first extracted the torque from Fluent and then calculated power (available) via a simple formula....

What are you working on?

Best,
Fred

monaya flower June 11, 2012 17:27

Hi

I 'm doing numerical simulation of shrouded wind turbine . and i can't calculate the torque .
I know that torque would be calculated from report - moment , but i can't specify x,y,z coordinates correctly. can you help me ?
what moment i should take if the rotating axis is x ?
why did you use SRF rather than MRF .

federvo.mala June 15, 2012 06:30

You should put '1' on the x axis and leave '0' for the other axis.

For a horizontal axis SRF is fine (based on literature) and suggestions. Although for a vertical one things are different.

monaya flower June 15, 2012 11:13

Hi Federvo
thank you for your reply :).
first Are you sure about using SRF ? cause i read the literature about HAWT and they were used MRF .
if you used SRF, what is in your opinion bc should i set to get good results?

federvo.mala June 15, 2012 13:10

Quote:

Originally Posted by monaya flower (Post 366690)
Hi Federvo
thank you for your reply :).
first Are you sure about using SRF ? cause i read the literature about HAWT and they were used MRF .
if you used SRF, what is in your opinion bc should i set to get good results?

100 % sure you can use SRF as well as MRF. If you like people also used sliding or dynamic mesh. Up to you :D

Velocity-inlet

Pressure-outlet

Far-field (semi circular face) = Symmetry

Plane faces of the domain = periodic

Blade = as shown in the pics above

Fred

cuteapathy June 17, 2012 23:59

Hi,

I have a similar question here. I had the mash which made in OpenFOAM and converted it to the mesh type for Fluent. In OpenFOAM, there are two kinds of mesh, one part, the outside cylinder, is static mesh and the other part, the wind turbine inside the smaller cylinder, is dynamic mesh, which is rotating. When I tried to run this case, it did not work. The case only runs in the initial second and can not run for more seconds. The wind turbine can not rotate at all. I am wondering whether I need to redefine or reset up the dynamic mesh. I have no idea how to fix this. Could anyone kindly give me some advice? Any advice would be helpful. Thank you.

sarah4 June 30, 2012 19:44

direction
 
Hello Federvo


I am wondering why your flow direction is -1 for "Y" direction, while your rotational direction is +1 for "Y".

I can see from your picture your turbine rotates +Y, then flow direction (from inlet to outlet) should be +1.

Am i right? Or i missed something?

Thanks.


All times are GMT -4. The time now is 16:58.