[stage81]

Hi all,

I'm trying to simulate a 2D savonius turbine (2 semi-cylinder blades) using FLUENT, the TSR = 0.8 and air velocity at the inlet is 6 m/s. I'm looking for the torque acting on the blades. By using REPORT--> FORCE RESULTS --> MOMENT (N-m) so this is the toruqe but the value of this torque is negative at each time step? i need to find a positive one ! please do you have any idea about that or advise?

Thanks a lot

[Hamidzoka]

Hi;
first of all check the physics of the flow inside the domain. Since there is a turbine the total energy content of the fluid in terms of total pressure ans total temperature should go down form inlet to outlet. If everything is ok and the distance of blade from the rotation axis is right the reported torque is true and the sign is not be a problem at all. This is maybe due to coordinate frame direction you have selected.

regards

[stage81]

thank you Hamidzoka for your replay. what do you mean when you said 'the distance of blade from the rotation axis is right'?

I have also another question, do you have an idea about how to plot the torque versus rotating angle? how can I found the rotation angle in fluent

Originally Posted by Hamidzoka

Hi;
first of all check the physics of the flow inside the domain. Since there is a turbine the total energy content of the fluid in terms of total pressure ans total temperature should go down form inlet to outlet. If everything is ok and the distance of blade from the rotation axis is right the reported torque is true and the sign is not be a problem at all. This is maybe due to coordinate frame direction you have selected.

regards

[flotus1]

When defining the torque report, did you make sure to pick the correct rotation axis?
A sketch of your setup could clear things up.

[stage81]

Hi Alex,

Thank you for your help. find attached a sketch of my rotating zone including (semi circular blades) (see "savonius test.jpg "). For information the diameter of blade (or length) is 117 mm and the diameter of the shaft is 24 mm (the origin of the shaft is x=0, y =0).

So to display the torque I chose as moment center x =0, y =0 (see picture attached). does these coordinates are correct ?

When I run the simulation by using MRF method in steady case I found a positive torque but its value isvery low compared to literature (see picture "Moving Reference Frame Method-steady.jpg").

When I run the simulation by using "mesh motion" method in transient case I found negative torque (see an example attached "Transient-Mesh motion technique.jpg ")

Thanks for your help

Originally Posted by flotus1

When defining the torque report, did you make sure to pick the correct rotation axis?
A sketch of your setup could clear things up.

[flotus1]

With the axis of rotation in your report being "+z", a negative torque value seems to be correct.
Make a fist with your right hand, stick out the thumb and point it in +z direction. Now your fingers point in positive rotation direction around this axis. This should be counterclockwise in your pictures. Since your rotor is supposed to spin clockwise, a negative torque value should be correct.

[stage81]

Thanks a lot for your interesting answer. Now in your opinion if I multiply the negative torque by (-1) the result will be correct? or I need to change the direction of rotation?

Other question this torque given by Fluent corresponds to the torque acting on one blade or two blades? If I need the torque of the complet rotor I need to multiply by number of blades or not?

Thanks again

Originally Posted by flotus1

With the axis of rotation in your report being "+z", a negative torque value seems to be correct.
Make a fist with your right hand, stick out the thumb and point it in +z direction. Now your fingers point in positive rotation direction around this axis. This should be counterclockwise in your pictures. Since your rotor is supposed to spin clockwise, a negative torque value should be correct.

[flotus1]

You can either multiply the value by -1 or choose -z as the axis of rotation if you would rather have a positive value. Doesn't matter which one you choose, this is only post-processing and does not affect the simulation.
If your boundary "blades" encompasses both blades, you get the total torque for both blades as a result.

[stage81]

Thank you very much. Really your answers were very useful for me and surely for others who will have the same problem !

Please do you have an idea about how can I plot the torque versus rotor angle (how can I access to rotor angle in FLUENT: 20 degré, 40, 90, ....360 for example) ?

Originally Posted by flotus1

You can either multiply the value by -1 or choose -z as the axis of rotation if you would rather have a positive value. Doesn't matter which one you choose, this is only post-processing and does not affect the simulation.
If your boundary "blades" encompasses both blades, you get the total torque for both blades as a result.

[flotus1]

Maybe it can be done in fluent directly, but my first idea would be to do this in an external program.
You could export the torque report values to a file and just calculate the angle from physical time.

[stage81]

I will try that, thanks

Originally Posted by flotus1

Maybe it can be done in fluent directly, but my first idea would be to do this in an external program.
You could export the torque report values to a file and just calculate the angle from physical time.

[stage81]

Thank you

Originally Posted by flotus1

Maybe it can be done in fluent directly, but my first idea would be to do this in an external program.
You could export the torque report values to a file and just calculate the angle from physical time.