Coordinate System in FAST

Dear all,

I am using FAST local flapwise/edgewise/axial forces for different spans in NREL 3-bladed turbine. The ‘OutListParameter.xls’ from FAST archive folder mentions something like this, where I presume the axes of edgewise and flapwise forces are swapped? Additionally, I as a user, would, appreciate if the flap-wise and edgewise forces are mentioned as ‘shear’ forces?

Also, I just want to confirm, if the coordinate axes mentioned in FAST user guide for 2-bladed downwind turbine are same for 3-bladed upwind turbine (blade coordinate system)?

Secondly, I am using these three component forces to apply point loads at different location in my Finite Element Model (FEM). This is basically done in order to replicate the same bending moment and shear force nature along the blade as derived from FAST (for maximum conservative value of flapwise moment in time series). I am using different ‘shear centers’ (point about which twist moment is zero: an assumption considering its small compared to flapwise and edgewise moment) to apply these loads. Is it the correct way to go about it, considering am not considering aero centre and centre of mass?

Honestly, I am not an aerodynamics engineer and hence don’t know much about these terms. Sorry for flooding this topic with so many questions, but I am in the middle of it and its hard to comprehend. Any kind of insight to this, would be appreciated!

Thanks a lot,
Neelabh

Dear Neelabh,

Yes, I agree that there is an error in the OutListParameter.xls file supplied with the present version of FAST. The correct definitions of Spn1FLxb1 and Spn1FLyb1 are as follows:

Spn1FLxb1 = Blade 1 local flapwise shear force at span station 1; Directed along the local xb1-axis; (kN)
Spn1FLyb1 = Blade 1 local edgewise shear force at span station 1; Directed along the local yb1-axis; (kN)

(Likewise for span stations 2-9 and blades 2-3.)

We will make sure to fix this in the next release of FAST. Thanks for reporting this error.

Yes, the blade coordinate systems in FAST are identical for a 2-bladed downwind turbine and a 3-bladed upwind turbine.

Here are the definitions I use:

Shear Center: Location in the cross section where if you apply a shear load, there will be no torsion.
Tension Center: Location in the cross section where the bending axes cross; a normal force applied at the tension center will not cause any bending.
Mass Center: Center of mass location of the cross section.
Aerodynamic Center: Location in the cross section where the specified airfoil lift, drag, and pitching moment coefficients are defined.
Elastic Axis: Loci of shear centers.

The present version of FAST assumes an initially straight blade made up of an isotropic material with no offsets of mass, shear, or tension center from the pitch axis. The local blade loads output from FAST are reaction loads at nodes on this axis. Typically, the shear center is of greater importance to large utility-scale wind turbines than the tension center due to the low rotational speed. So, yes, I would apply the loads output from FAST at the shear center of the FE model. However, if your blade has large deviations of the mass, shear, or tension center from the pitch axis, than the present version of FAST may not be accurate for computing the system’s response.

Best regards,

Thanks a lot Jason for the response. And your logic of using shear center helps too.But why do you say its more valid in case of low rotational speed wind turbine?

Additionally, I would like to know if flagging the direction of rotation of turbine as ‘CLOCKWISE’ in ‘Turbsim’ would effect my results in FE analysis, as I observe the blade there has an Anti-clockwise orientation (I didn’t realize this while I was generating wind files). I am using Bladed style FF files and linking FAST to ‘Aerodyn’ to calculate the loads.

I read this in Turbsim manual about rotation of turbine flag: “Because AeroDyn also reads the Bladed-style FF files based on the direction of rotation, this flag does not affect the results when used with AeroDyn.” I am not really sure what it means.

Thanks a lot,
Neelabh

Neelabh,

I’ll let Jason respond to the other part of your question, but if you are reading TurbSim-generated files with AeroDyn, TurbSim’s “clockwise” flag doesn’t do anything. i.e., you will get exactly the same wind field in AeroDyn regardless of what the value of “Clockwise” is when you run TurbSim.

[size=85]If you want the details: The binary file is written with the y values going one direction when clockwise is true and the y values going the opposite direction when clockwise is false. However, AeroDyn reads the TurbSim summary file and then reads the binary file in the appropriate direction based on the “clockwise” flag. So, you’ll get exactly the same wind speeds either way. This flag was put in one of TurbSim’s predecessors (probably SNWind) to help someone compare FAST and Bladed. It may not even be necessary with Bladed anymore.[/size]

Dear Neelabh,

With low rotational speed, the centrifugal forces will be small. When the centrifugal forces are small and the aerodynamic forces are the large, the bending moment caused by centrifugal forces and the lateral offset between the tension center and mass center will be less important than the bending and torsion moments caused by the aerodynamic forces and the lateral offset between the aerodynamic center and shear center.

Best regards,

Thanks a lot Jason and Bonnie for your responses.I really appreciate these elaborate and clear answers.

To just sum up my understanding on this whole topic:

“Centrifugal and all other inertial forces (gravitational, gyroscopic, Coriolis) are calculated at the center of mass and when there is an offset between mass center and tension center, there is a bending moment associated with it (otherwise not). Similarly all the aerodynamic forces are calculated at aerodynamic center and when there is an offset between aerodynamic and shear center there is a torsion (and a delta bending moment which changes every time with angle of attack) associated with it (considering bending moment: flap wise and edge wise are constant, about the aerodynamic center).
Hence when I apply my loads in FE model at shear center I apparently, neglect the torsion moment and the delta bending moment associated with the offset between aerodynamic center and shear center!”

Please let me know, if my understanding is flawed somewhere? Also, did I miss the link between pitch axis with all these centers?

Thanks as always,
Neelabh

Dear Neelabh,

I’m not sure I understand what you are saying when you say, “(considering bending moment: flap wise and edge wise are constant, about the aerodynamic center)”, but I agree with everything else in your post.

The pitch axis is the axis of the undeflected blade about which the blade rotates due to changes in the blade-pitch angle. Of course, in general, all of the “centers” I mentioned in my first post above can all be offset relative to the pitch axis.

Best regards,

Dear Jason,

I was mistaken, I wanted to mean ‘Aerodynamic moment’ is constant about the the aerodynamic center (irrespective of change in angle of attacks) which I feel includes torsion and flapwise moment (and not edge wise which is due to gravity)?

Sorry for this mistake!

Best Regards,
Neelabh

Dear All,

I have a query regarding the bending moment (flap-wise and edgewise) conventions used in FAST for a ‘3 Bladee Upwind Anticlockwise’ rotating turbine. Based on OutListParameters.xls and FAST user manual, I drew the following sketches for getting the positive moments:

Coordinate_system.pdf (326 KB)

As I see from this sketch (2nd one), the flap wise moment comes to be positive about “upwind” direction when rather according to standards it should be positive along downwind direction (also makes sense). I am not sure, if blade local flapwise moment ‘along’ local yb axis means +ve or -ve local yb axis, in OutListParameters.xls? I interpreted its +ve axis, and then results come opposite to the standard. Similarly what would be the right convention for edgewise moments for this configuration of turbine?
These things make sense for clockwise rotating turbine but its confusing for anticlockwise configuration.

Can someone please, shed some light on this?

Thanks and Best Regards,
Neelabh

Dear Neelabh,

FAST can only model wind turbines rotating in the clockwise direction (when looking downwind), so, I’m not sure how you are going about modeling a wind turbine rotating in the anticlockwise (counterclockwise) direction. For a rotor spinning in the clockwise direction in FAST, what you identify as the x’ axis in your sketches should point toward the suction surface (downwind direction), the y’ axis should point towards the trailing edge, and the z’ axis should point towards the tip.

The user’s guide for the old YawDyn code (attached) had a very nice section that described how to adapt a model that only works with clockwise-spinning rotors to model counterclockwise-spinning rotors. See the section, “Nomenclature and Sign Conventions.” This section, which was written for YawDyn (of course), but applies equally well to FAST.

Best regards,
YawDyn.pdf (274 KB)

Dear Jason,

How I was thinking of modeling it, is to interpret the axes of FAST for the anticlockwise orientation as follows:

Coordinate_system.pdf (351 KB).

So here, x’ is in downwind directions, y’:points towards LEADING edge direction, z’ points towards tip. Also the forces have been represented accordingly for getting positive moment conventions. I think this seems reasonable as it doesn’t violate the standard in terms of x’ and y’ orientation (about y’ orientation I think, its directed along TRAILING edge only for clockwise rotating turbine as mentioned in standards), they follow the right hand coordinate system rule and now the moments from FAST and direction of forces make sense.

Do you agree to this system for using FAST results from a clockwise orientation to anticlockwise?

I am not using any horizontal shear and yaw error is set to be 0 degree for the structural analysis I am doing on a blade. So I am not sure, how to implement the conventions from FAST input-side. Hence, I was thinking to “assume” the changed coordinate system of FAST during post processing (like I mentioned above) to get the same moments and forces as I would have got for a clockwise orientation.

Please let me know if it makes sense for you.

Thanks a lot,
Neelabh

Dear Neelabh,

If analyzing a rotor that spins in the anticlockwise (counterclockwise) direction (when looking downwind), you shouldn’t change the blade coordinate system at all. For example, a force tangential to the chordline and directed toward the trailing edge won’t point towards the leading edge when the rotor spins in the opposite direction. In FAST, you must model the rotor spinning in the clockwise direction and adjust inputs (such as the yaw angle) accordingly, as described well in the YawDyn User’s Guide attached to my prior post.

Best regards,

Dear Jason,

  • Infact, if you observe, I didn’t change the coordinate system from clockwise to anticlockwise orientation (my last post). Its still the same as clockwise orientation; y’ axis changes as leading edge becomes trailing edge (TE) and trailing edge becomes leading edge (LE) as we move from clockwise orientation of blade to anticlockwise (from my understanding). But on the whole, its still ‘same’ . From what I understood, if LE was under compression and TE was under tension for clockwise orientation, then LE comes under tensile stress and TE under compressive when blade has an anticlockwise orientation. The force still acts in same direction for both the cases. Is my understanding correct?

  • I read YawDyn.pdf: in my model I am using full field wind files(.wnd files) with .sum files for simulating wind field (in turbsim). Now, I can see Horizontal shear in hub height file (.hh file: horizontal shear is zero here), but i cannot read binary files (.wnd) and am not sure how horizontal shear convention is changed here (or if needs to be changed since its zero in hub height files) ?

-For Nacelle yaw angle, I am still using 0 degree constant final yaw (‘NacYawF’ input in FAST). In future, if I put +ve value when I want a -ve one and vice versa, will it serve the purpose of anticlockwise orientation?

Thanks for you patience and am sorry for these long posts!

Neelabh

Dear Neelabh,

Your understanding is not correct. The tension/compression of the leading edge (LE) and trailing edge (TE) will not be reversed if the rotor spins in the opposite direction (assuming no shear or yaw error).

You only need to be concerned about the shear and yaw conventions if you are trying to model a specific set of shear/yaw conditions. For example, in Figure 5.5 of the YawDyn User’s Guide, it shows how one should reverse the direction of the horizontal shear when modeling in the clockwise direction when that actual rotor spins counterclockwise. If you are simply using TurbSim to generate turbulent wind, you are not trying to match a specific horizontal shear.

I’m not sure I understand you last question regarding nacelle yaw angle.

Best regards,

Dear Jason,

My line of though might be incorrect then. This is how the blade I am modeling looks like ‘from tip’:

Blade_view_from_tip.png

Generally, I observed (for clockwise rotating turbine), the airfoils have opposite orientation to the one showed above when looking from tip (i.e. LE and TE are swapped with orientation of pressure and suction side same) . Hence, I thought this orientation, being opposite, is for an anticlockwise rotating turbine (for the picture shown above).

  • But, as I understand from you answers, the rotation direction of turbine has nothing to do with airfoil orientation? This was the origin of the doubt and might be very silly and stupid as I look at it now.

  • If I apply the coordinate axes as you mentioned for the above shown airfoil section (i.e. x’ towards downwind direction, y’ towards trailing edge, z’ towards tip), they don’t follow the right hand coordinate system rule. Hence, force direction were altered which should not be the case.

  • Regarding my last posts: It was based on YayDyn.pdf recommendations:: So it is not possible to manipulate horizontal shear and nacelle yaw using TurbSim and FAST?

Thanks a lot,
Neelabh

Dear Neelabh,

Your image of the blade when viewed from the tip for a rotor spinning counterclockwise looks correct. And yes, a rotor spinning clockwise will have a blade that is the mirror image of what you are showing (the leading edge on the opposite side).

What I was disagreeing with you about earlier was your desire to change the coordinate such that y’ axis points to the LE instead of the TE without changing the sign on the loads output by FAST. For example, if you wish to model your counterclockwise-spinning rotor in FAST, you must model it rotating clockwise and change the signs of some outputs. The relationship between the loads output by FAST and the loads in the x’/y’/z’ coordinate system you defined in your Feb 22, 2013 post above for the local blade loads at station 1 of blade 1 would be:

Force directed along x’ for a counterclockwise rotor = Spn1FLxb1 from FAST for a clockwise rotor
Force directed along y’ for a counterclockwise rotor = -Spn1FLyb1 from FAST for a clockwise rotor
Force directed along z’ for a counterclockwise rotor = Spn1Flzb1 from FAST for a clockwise rotor
Moment about x’ for a counterclockwise rotor = -Spn1MLxb1 from FAST for a clockwise rotor
Moment about y’ for a counterclockwise rotor = Spn1MLyb1 from FAST for a clockwise rotor
Moment about z’ for a counterclockwise rotor = Spn1MLzb1 from FAST for a clockwise rotor

(Notice the change of signs related to the force along y’ and moment about x’.)

My point regarding TurbSim was that a user cannot specify the horizontal shear when running TurbSim. So, you wouldn’t be using TurbSim if you were trying to model a specific horizontal shear. Regarding yaw, the wind direction and nacelle-yaw angle are not specified in TurbSim. If you wish to model a specific yaw error, than you’ll have to reverse the sign on the nacelle-yaw angle in FAST because you must model the rotor spinning in the clockwise direction when the actual rotor spins counterclockwise (as described in the YawDyn’s User’s Guide).

I hope that helps.

Best regards,

Dear Jason,

Just to be on the same page, I will reiterate some of the points I understood. Please correct me, if I am wrong:

  • x’/y’/z’ don’t need to follow right hand system for anticlockwise orientation of airfoil.

  • While the flap-wise bending moment is positive when the blade is bent in downwind direction for clockwise rotating turbine, its positive when blade is bent in ‘upwind’ direction for anticlockwise rotating turbine as per the conventions you mentioned (to maintain consistency with y’ axis).

  • When I am not trying to match a specific horizontal shear (using TurbSim) and yaw error is zero, nothing needs to be changed for this particular scenario.

Sorry for theses long chain of posts, but this coordinate system seems so confusing when I relate it to reality.

Thanks a lot,
Neelabh

Dear Neelabh,

While I agree with your third point, I disagree with your first and second points.

I agree with the x’/y’/z’ coordinate system you proposed in your Feb 22, 2013 post above whereby you reversed the direction of the y’ axis to make the x’/y’/z’ coordinate system right-handed.

The flapwise bending moment is positive when the blade deflects downwind whether the rotor spins clockwise or counterclockwise. For a clockwise rotor, the flapwise bending moment is the moment about the positive y’ axis that is directed toward the trailing edge. For a counterclockwise rotor, the flapwise moment is moment about the positive y’ axis that is directed toward the leading edge. As I said in my prior post, the moment about y’ for a counterclockwise rotor = Spn1MLyb1 from FAST for a clockwise rotor.

Best regards,

Dear Jason,

I understand your points above and agree to them. One last thing, though (as we agreed upon, I am considering coordinate system from 22nd February post is correct):

While I understand that flapwise and axial moments remain same, what I don’t get is why the sign has to be reversed for edgewise bending moment for anticlockwise oriented blade (Moment about x’ for a counterclockwise rotor = -Spn1MLxb1 from FAST for a clockwise rotor). From what I know, it is caused by gravity forces and they would always act downwards. Now when the profile of the airfoil changes (LE comes to TE) for anticlockwise orientation of blade, gravity force would still act down, but now the nature of forces in TE and LE should swap. This would require keeping the same sign of edgewise bending moment. If we change the edgewise moment sign, then we have the same nature of LE and TE forces for both orientation of airfoil, which I don’t understand?

Thanks and Best Regards,
Neelabh

Dear Neelabh,

I agree that the gravity moment would change the nature of the LE/TE forces when considering a clockwise versus counterclockwise blade in the same “time” on the clock (looking downwind). However, as we’ve discussed here, you must model your counterclockwise rotor spinning clockwise in FAST. To do this, you must mirror the azimuth convention when considering the azimuth angle. For example, the “2 o’clock” position (looking downwind) for a clockwise rotor represents the “10 o’clock” position (looking downwind) for a counterclockwise rotor (the blade is rotating downward in both cases). Because these different “times” represent the same event in both rotors, the sign of the moment about the x’ axis is reversed.

I hope that helps.

Best regards,