Structural Dynamic Loads on Wind Turbines

[Rhombuss]

This is probably a long shot, but I was curious if anyone had any experience (either theoretical or practical) about analysis of dynamic wind loads on wind turbines (specifically for wind-power generation).

Should the frontal projected wind area be taken as the stationary projected blade area, or does the fact that the blades rotate increase or decrease loading?

Also, since the blades rotating cause centripedal forces, are they generally large enough to cause 2nd order deflection/bending at the tower mast base?

 

[alphatarget1]

I don't think you need structural dynamics for this problem honestly, but then I'm only an undergrad in civil so what do i know?

The blade is at an angle so the wind pushes it and one component of that create a force tangent to the circumference (if that makes any sense). I don't think they spin *that* fast anyway, but blades will probably have to be properly designed so that the centripedal forces do not cause excessive stress, you don't want blade fragments flying everywhere. Fatigue shouldn't be that much of a concern, either.

I don't really see how the blades would cause 2nd order deflection either, what do you mean by 2nd order deflection anyway? Do you mean combined loading or P-Delta effect (don't know much about P-Delta...) . Just calculate the loading, let's say, 130mph winds, (not sure if these structures are governed by building codes, probably) and throw a factor of safety in there. Make sure the foundation is properly done..

EDIT: you probably would get some bending of the blade about the base also, something to think about...

 

[CycloWizard]

Dynamic analysis of any sort of turbine, whether for power generation or mixing (agitation), has been sort of a pipe dream for a long time. Most of the guys I talked to that worked in this area basically said the dynamics of the situation won't be understood on a theoretical level in their lifetime. That said, you could solve the problem numerically using a combined fluid dynamics/dynamic finite element analysis, depending on what you have at your disposal.

The rotation of the blades will most definitely change the loading. Looking at it from a momentum transfer perspective, the purpose of the tilt in the blades is to change the force from an axial force (front of impeller to the back) to a tangential force. The tangential force will offer maximum rotation speed of the impeller, generating maximum power. How exactly these forces will vary with location I can't say a priori, as it would require solution of a 2-D PDE that is very much geometry-dependent, and that's just the static case.

The second question requires more speculation on my part, but I can't imagine that a part would be designed to undergo bending and so on under normal operating conditions. The centripetal forces generated by the blades should (ideally) cancel each other out, so it should have a net zero impact on the mast.

 

[Rhombuss]

I'm a structural engineer working in the telecommunications industry, so I have experience with communication tower design. This wind turbine project came up, so I was curious of how to treat the dynamic loads.

(1) For this kind of system, I think structural dynamics plays a large role if you want to attain accurate analysis. Because of the rotation of the blades and intersecting winds, the tower will undoubtedly have a variable frequency (depending on wind speeds). Most likely much faster than normal high-rise buildings.

(2) The tapered and twisted shape of the blade isn't in dispute here. Obviously, it's to generate rotation with frontal wind being applied. I'm speaking specifically of the applicable wind loading that can be considered for wind area load, as the area (particularly flat area, not smooth) contributes to increased wind load.

(3) I believe second order effects (yep, P-delta) will definitely impact the base reactions. Generally, the 1.8-2MW generators are over 100 metric tonnes in weight. And the rotors will have a larger wind area than almost any equivalently sized communications tower, so the deflection at the top will be high. Once the rotor and nacelles shift over the vertical axis slightly, that 100 metric tonnes of weight will be literally twisting the base of the tower. Regardless of the size of the mast base, I think the moment caused from that second-order effect will be very high.

(4) Typically centripetal forces (of similarly symmetric arms) cancel each other out if the support is parallel to the rotation axis (right-hand rule anyone? ). That's not the case here. The blades are spinning about a horizontal axis and along a vertical plane, while the tower mast is vertically oriented. I think the downward force of the blade on one side, along with the upward force of the blade as it spins around should induce moments at the rotor/nacelle base support.

Interesting discussion so far, lets keep it up!

 

[Minerva]

SWECS rule of thumb is plan for full swept area (as if a solid disk) when running at service speed, aka full power! That's an incredible load. Above this feathering becomes important or your structure will fail.

 

[Rhombuss]

Hmm, I find it difficult to believe, since most turbines have a nominal rotation of approximately 15 rpm. Anything beyond that and most turbines will have an auto-lock feature to prevent excessive winds from damaging the blades from over-rotation. At 15 rpm, that's about a rotation every 4 seconds. There's no way you can conservatively justify taking the entire sweep area of the blades.

Could you provide a link where you got this info, I'm really curious as to it's validity.

 

[Minerva]

Keyword is plan so it's over engineering. It came from a textbook; sorry no reference off the top of my head but I'm sure I can track it down. This was specifically covered in design of the collector supporting structure.

 

[CycloWizard]

Originally posted by: Rhombuss
I'm a structural engineer working

Fair enough... You're well more qualified than me in this area, so I'll just shut up.

The only advice I have then is that maybe you could find something looking into the design of liquid/gaseous agitators. I know the technology is very similar, though the implementation may be different.

 

[Rhombuss]

Yeah, it's probably a 'safe' design strategy for very small turbines of less than 10m in diameter. But for something larger, say 75-100m diameter, it's literally impossible to design a shaft mount and foundation that could take the entire sweep area for wind loading.

For a basic estimate: Say the wind loading is 500Pa (about average for an offshore installation). Tower height about 100m, so the elevation factor would be about 1.5 or so. If the entire sweep area is taken for a 100m diameter rotor, that's 7854m² area.

So roughly, we have a wind load force of about = (0.5 * 1.5 * 7854) = 5890 kN or 1320 kips for those of you used to imperial.

With a tower height of 100m, we get about 589000 kN-m or 433070 kip-ft of bending at the base (doesn't even include wind loading on the mast or bending from second-order effects). That is INSANELY huge. According to the Canadian steel standard, a required section modulus for that moment is:

Mr = 0.9 (resistance factor) * 350MPa (yield stress) * Sx, so Sx = 1.870 x 10^9 mm^4. Wow, a lot of second-moments of area (I) values for very large tubular sections don't even get close to that large, and this is a SECTION modulus. For sure this may be applicable for small turbines, but for large turbines, I think more liberal analysis is required for economic (and structural feasibility) considerations.

 

[bonkers325]

Originally posted by: Rhombuss
Yeah, it's probably a 'safe' design strategy for very small turbines of less than 10m in diameter. But for something larger, say 75-100m diameter, it's literally impossible to design a shaft mount and foundation that could take the entire sweep area for wind loading.

For a basic estimate: Say the wind loading is 500Pa (about average for an offshore installation). Tower height about 100m, so the elevation factor would be about 1.5 or so. If the entire sweep area is taken for a 100m diameter rotor, that's 7854m² area.

So roughly, we have a wind load force of about = (0.5 * 1.5 * 7854) = 5890 kN or 1320 kips for those of you used to imperial.

With a tower height of 100m, we get about 589000 kN-m or 433070 kip-ft of bending at the base (doesn't even include wind loading on the mast or bending from second-order effects). That is INSANELY huge. According to the Canadian steel standard, a required section modulus for that moment is:

Mr = 0.9 (resistance factor) * 350MPa (yield stress) * Sx, so Sx = 1.870 x 10^9 mm^4. Wow, a lot of second-moments of area (I) values for very large tubular sections don't even get close to that large, and this is a SECTION modulus. For sure this may be applicable for small turbines, but for large turbines, I think more liberal analysis is required for economic (and structural feasibility) considerations.

strangely enough, i understood exactly what you just said. how hard was it for you to find a job after you graduated? im looking for entry-level structural engineering positions and was wondering what they would have me do for my first year or two. i'd imagine busy-work and proofing data/reports and other equally menial tasks...

 

[alphatarget1]

Originally posted by: Rhombuss
I'm a structural engineer working in the telecommunications industry, so I have experience with communication tower design. This wind turbine project came up, so I was curious of how to treat the dynamic loads.

(1) For this kind of system, I think structural dynamics plays a large role if you want to attain accurate analysis. Because of the rotation of the blades and intersecting winds, the tower will undoubtedly have a variable frequency (depending on wind speeds). Most likely much faster than normal high-rise buildings.

(2) The tapered and twisted shape of the blade isn't in dispute here. Obviously, it's to generate rotation with frontal wind being applied. I'm speaking specifically of the applicable wind loading that can be considered for wind area load, as the area (particularly flat area, not smooth) contributes to increased wind load.

(3) I believe second order effects (yep, P-delta) will definitely impact the base reactions. Generally, the 1.8-2MW generators are over 100 metric tonnes in weight. And the rotors will have a larger wind area than almost any equivalently sized communications tower, so the deflection at the top will be high. Once the rotor and nacelles shift over the vertical axis slightly, that 100 metric tonnes of weight will be literally twisting the base of the tower. Regardless of the size of the mast base, I think the moment caused from that second-order effect will be very high.

(4) Typically centripetal forces (of similarly symmetric arms) cancel each other out if the support is parallel to the rotation axis (right-hand rule anyone? ). That's not the case here. The blades are spinning about a horizontal axis and along a vertical plane, while the tower mast is vertically oriented. I think the downward force of the blade on one side, along with the upward force of the blade as it spins around should induce moments at the rotor/nacelle base support.

Interesting discussion so far, lets keep it up!

sweet, I thought there weren't many civil/structural engineers on these boards.

1) I don't know much about wind loads (not that i know a whole lot about earthquake either), but as long as the resonant frequency is low it should be fine... am I right, or am I way over my head?

2) Here is an idea, have a ME draw and run a COSMOS analysis on a blade assembly, scaling down (dimensional analysis) things so the analysis is shorter.

3) 2MW? wow, I didn't know wind turbines generate that much power! That is a lot of weight though, I agree, P-Delta is necessary if there is a significant amount of deflection at the top. How much deflection are you anticipating from the wind load at the top?

Where do you work, if you don't mind me asking? I think California schools emphasize a lot more on earthquake than anything else.

 

[alphatarget1]

Originally posted by: Rhombuss
Yeah, it's probably a 'safe' design strategy for very small turbines of less than 10m in diameter. But for something larger, say 75-100m diameter, it's literally impossible to design a shaft mount and foundation that could take the entire sweep area for wind loading.

For a basic estimate: Say the wind loading is 500Pa (about average for an offshore installation). Tower height about 100m, so the elevation factor would be about 1.5 or so. If the entire sweep area is taken for a 100m diameter rotor, that's 7854m² area.

So roughly, we have a wind load force of about = (0.5 * 1.5 * 7854) = 5890 kN or 1320 kips for those of you used to imperial.

With a tower height of 100m, we get about 589000 kN-m or 433070 kip-ft of bending at the base (doesn't even include wind loading on the mast or bending from second-order effects). That is INSANELY huge. According to the Canadian steel standard, a required section modulus for that moment is:

Mr = 0.9 (resistance factor) * 350MPa (yield stress) * Sx, so Sx = 1.870 x 10^9 mm^4. Wow, a lot of second-moments of area (I) values for very large tubular sections don't even get close to that large, and this is a SECTION modulus. For sure this may be applicable for small turbines, but for large turbines, I think more liberal analysis is required for economic (and structural feasibility) considerations.

The little column I designed for my concrete design project had a peak moment of... 4000 kip-in . I'm assuming you're from Canada, do you use a different building code than we do?

Are you planning to use steel for the tower?

 

[Rhombuss]

Originally posted by: CycloWizard
Fair enough... You're well more qualified than me in this area, so I'll just shut up.

The only advice I have then is that maybe you could find something looking into the design of liquid/gaseous agitators. I know the technology is very similar, though the implementation may be different.

No need to 'shut up', I posted here to get interesting ideas off of everyone, all thoughts are welcome .

Will definitely take your advice on the gaseous agitators.

Originally posted by: bonkers325
strangely enough, i understood exactly what you just said. how hard was it for you to find a job after you graduated? im looking for entry-level structural engineering positions and was wondering what they would have me do for my first year or two. i'd imagine busy-work and proofing data/reports and other equally menial tasks...

I found a very good position at a large firm quite soon after graduation, it's very important to have good relationships with your professors (particularly those with industry contacts). In terms of what you get to do the first few years, it varies from firm to firm. Generally the smaller firms will have you doing more tedious tasks, most of which may not be technical, such as drawing review and preparation. If you get to work for a larger firm with more resources, most of the low-level administrative stuff is given to people hired for that purpose, and you get to do a lot more technical design (and in some cases theoretical analysis) which is what I'm lucky enough to be able to do.

Originally posted by: alphatarget1

The little column I designed for my concrete design project had a peak moment of... 4000 kip-in . I'm assuming you're from Canada, do you use a different building code than we do?

Are you planning to use steel for the tower?

Yes, from Canada. Actually, we have very sophisticated and thorough steel and concrete design codes in Canada. The American steel code primarily follows the ASD (allowable stress design) standard, which isn't the most up-to-date standard available. We follow a strict limit-states methodology, which I think is what the American codes lack, although I'm aware that a steel limit-states design code was issued a number of years ago to no avail.

Definitely using steel. If reinforced concrete was used, the amount of reinforcing required would probably be enough to build a separate tower .

Originally posted by: alphatarget1
sweet, I thought there weren't many civil/structural engineers on these boards.

1) I don't know much about wind loads (not that i know a whole lot about earthquake either), but as long as the resonant frequency is low it should be fine... am I right, or am I way over my head?

2) Here is an idea, have a ME draw and run a COSMOS analysis on a blade assembly, scaling down (dimensional analysis) things so the analysis is shorter.

3) 2MW? wow, I didn't know wind turbines generate that much power! That is a lot of weight though, I agree, P-Delta is necessary if there is a significant amount of deflection at the top. How much deflection are you anticipating from the wind load at the top?

Where do you work, if you don't mind me asking? I think California schools emphasize a lot more on earthquake than anything else.

The sway of the tower back and forth will largely depend on the wind area (which is why I posted in the first place). There doesn't seem to be much literature in how to treat it from a structural foundation point of view. Also, I may have neglected to mention, but this tower is to be installed off-shore. Thus constant wave loading and immense wave loading from storms will be another factor on the lower elevations of the tower. Considering hydraulic forces are much greater than wind forces, it may be significant.

We don't have inhouse mechanical engineering consultants (at least not directly for our telecommunications division, unfortunately), but that would be an interesting thought. Hopefully the manufacturers have already done the leg work in this area, and are willing to provide that data.

Anticipating the deflection to be on the order of 'feet' during max gusts (that's assuming total deflection range, which considers sway back and forth over the stationary point).

I'm from Toronto, which isn't an earthquake design zone. You're correct, west coast institutions both in Canada and the US highly emphasize structural and geotechnical dynamics for earthquake analysis. It's amazing these days what can be done to dampen dynamic shock from earthquakes. I've seen some buildings being supported entirely be hydraulic foundations that reduce impact loading on the building columns - very neat!

 

[alphatarget1]

Originally posted by: Rhombuss

Yes, from Canada. Actually, we have very sophisticated and thorough steel and concrete design codes in Canada. The American steel code primarily follows the ASD (allowable stress design) standard, which isn't the most up-to-date standard available. We follow a strict limit-states methodology, which I think is what the American codes lack, although I'm aware that a steel limit-states design code was issued a number of years ago to no avail.

Definitely using steel. If reinforced concrete was used, the amount of reinforcing required would probably be enough to build a separate tower .

Originally posted by: alphatarget1
sweet, I thought there weren't many civil/structural engineers on these boards.

1) I don't know much about wind loads (not that i know a whole lot about earthquake either), but as long as the resonant frequency is low it should be fine... am I right, or am I way over my head?

2) Here is an idea, have a ME draw and run a COSMOS analysis on a blade assembly, scaling down (dimensional analysis) things so the analysis is shorter.

3) 2MW? wow, I didn't know wind turbines generate that much power! That is a lot of weight though, I agree, P-Delta is necessary if there is a significant amount of deflection at the top. How much deflection are you anticipating from the wind load at the top?

Where do you work, if you don't mind me asking? I think California schools emphasize a lot more on earthquake than anything else.

The sway of the tower back and forth will largely depend on the wind area (which is why I posted in the first place). There doesn't seem to be much literature in how to treat it from a structural foundation point of view. Also, I may have neglected to mention, but this tower is to be installed off-shore. Thus constant wave loading and immense wave loading from storms will be another factor on the lower elevations of the tower. Considering hydraulic forces are much greater than wind forces, it may be significant.

We don't have inhouse mechanical engineering consultants (at least not directly for our telecommunications division, unfortunately), but that would be an interesting thought. Hopefully the manufacturers have already done the leg work in this area, and are willing to provide that data.

Anticipating the deflection to be on the order of 'feet' during max gusts (that's assuming total deflection range, which considers sway back and forth over the stationary point).

I'm from Toronto, which isn't an earthquake design zone. You're correct, west coast institutions both in Canada and the US highly emphasize structural and geotechnical dynamics for earthquake analysis. It's amazing these days what can be done to dampen dynamic shock from earthquakes. I've seen some buildings being supported entirely be hydraulic foundations that reduce impact loading on the building columns - very neat!

I think the industry here is trying to move to strength design (LSFD, if i remember correctly) for steel. Concrete is all strength design now, much more efficient than ASD, methinks.

How do you plan on dealing with corrosion with steel? I'm assuming that this thing, once built, will stay there for an extended period of time. They have to paint the Golden Gate Bridge all the time (from one side to the other) continuously. When you mention the wave loading, do you mean the foundation?

I have never done stuff with deflection on the order of feet . This is such an interesting topic.

I never really liked the idea of having dampers in the foundations, massive and expensive. They make these dampers for the chevron bracing for steel buildings and they're pretty nice. ease to maintain and everything.

 

[dkozloski]

Speaking as a layman with about fifty years in aviation and aero space, the part about wind generators that fascinates me as I stand and look at that enoumous rotating fan is, what the hell do you do to stop it when the blade pitch control and the orientation mechanism fails? If you can't change the pitch, the torque remains, and if you can't turn the mast to present the edge of the blades to the wind you're going to have to wait for a calm day to lassoe the blade as it goes by or shoot it off with a bazooka. I can't imagine any brake mechanism that could have any serious effect on slowing the spinning blades without burning up. Look at that tiny hub, realize that most of the work takes place near the tips of the blades, you're absorbing 2mw of heat, and it looks impossible. To me this would be a real head scratcher, like putting out an oil well fire in a hurricane.

 

[dkozloski]

Annother thought comes to mind. Do you design for a blade failure where you lose part or all of one blade? Do you just let the remaining mechanism disintegrate?

 

[dkozloski]

At this stage of the game, why would you be designing more of these things? It would seem to me that there must be a multitude of good basic designs in existence already. The sizing and wind velocity parameters must be pretty well set in stone by now so all an entrant to the field would have to deal with is variations in the building site and possibly inclement weather like icing conditions. Maybe engineers just like to build monuments to themselves.

 

[alphatarget1]

Originally posted by: dkozloski
Annother thought comes to mind. Do you design for a blade failure where you lose part or all of one blade? Do you just let the remaining mechanism disintegrate?

I think the generator used isn't designed by him, he's just trying to design a structure beneath it.

what about height reduction? It boils down to a big cantilevered beam with a big axial load on it, a big concentrated load on the top and also a big distributed load on it. how much loss will the turbine get if you reduce the height by 20 meters? it should reduce on deflection (p delta effects), structural weight, wind load and stuff... just a thought.

 

[dkozloski]

It doesn't look to me that there is very much clearance between the ground and the blades to start with. Also there is a sharp rise in wind velocity with a small increase in height plus the turbulence from ground eddys diminisnes.

 

[Rhombuss]

Most of the turbines I have researched thus far have locking mechanisms that prevent the rotor from spinning at sub or super-critical velocities. So yes, they do have effective breaking mechanisms in the nacelles. Icing also isn't much of an issue, as many of the manufacturers install the turbines with heating systems to prevent condensation from freezing.

In terms "standard" tower masts for these turbines, it's very much the kind of situation where you try NOT to over-design, and have it relatively stream-lined for economic considerations. Soil foundations, local statistic wind speeds, elevation requirements, and rotor/nacelle mechanical specifications all contribute to the design process. It is possible for manufacturers to make generic designs, but I guess the multitude of configurations is more than one might think. Also, another big reason is the liability and responsibility of the structure's lifespan. They'd rather have the onus on other engineers to take the liability of the tower mast. Almost every industry involving structural engineers is similar. A lot of time can be saved, but engineers ultimately get additional work just for taking responsibility incase of some failure.

The turbine we're currently looking at is approximately 80m tall, with 45m long blades (90m rotor diameter). This is an offshore installation, so the blades must be kept a reasonable height off the tower base to avoid high waves during storms.

 

[dkozloski]

Good answer Rhombuss, as a practcal matter a generic fitzall mast could be designed and manufactured but it would have to be so overbuilt in so many areas it would be uneconomical. The masts I've seen in Kotzebue, Alaska would probably also work in Southern California but not offshore. The incorporation of a locking device to prevent rotation of a stopped rotor is one thing but stopping an out of control rotor in a wind storm would be a real trick. I'd still want to be there to see the fireworks show.

 

[kevinthenerd]

Originally posted by: Rhombuss
This is probably a long shot, but I was curious if anyone had any experience (either theoretical or practical) about analysis of dynamic wind loads on wind turbines (specifically for wind-power generation).

Should the frontal projected wind area be taken as the stationary projected blade area, or does the fact that the blades rotate increase or decrease loading?

Also, since the blades rotating cause centripedal forces, are they generally large enough to cause 2nd order deflection/bending at the tower mast base?

You get three loading vectors on the blades: the frontal area (direct hit from drag) and the induced force (rotational force from acceleration). The third one is centripetal force. Gravity should be nearly negligible, but depending on the geometry of the blade, it might be worth a quick check with a lower F.S. When the wind dies down, the natural friction of the blades slow them down, producing an opposite rotational force.

but that's just for the blade. This is something I'd definitely want to do numerically.

 

[Resnik]

Rhombuss -
May be I didn't understand the question correctly !
Are you analyzing (or looking for the analysis) Turbine blade, Rotor or the Column that hold the structure?
In any case dynamic Analsys can be carried thru FEA(Finite Element Analysis) and latest software like COSMOS supports the analysis.

I did developed some software back in 92 for Dynamic Analysis of Active Magnetic Bearings.

 

[kevinthenerd]

It's always more FUN to develop your own software, but I've already learned in my limited professional experience that writing your own code isn't the best way to meet deadlines. The best thing is to write your own code BEFORE the project is assigned to you. (I'm trying to write a library of various situations that I'll encounter for maximum efficiency without the expensive use of outsourced software, such as ANYSYS, etc.)

 

[kevinthenerd]

I'm unique as far as ME's go. I LOVE writing code. It's not just a necessary evil for modeling for me. It's a major hobby and often a distration from real mechanical design.