June 27, 2009
May 12, 2009
In earlier posts I have mentioned that a turbine capable of harvesting the energy of jet streams would probably be better for newspaper headlines than for an economical approach to wind electricity, since it would probably be cheaper and more effective to build several smaller low altitude turbines than a single monster that could tap into the jet streams. But it got me to realize that there are no jet stream turbines on the Salient White Elephant. This is Salient, to be sure, but is it White Elephant? Certainly not! And already I can hear not a little hubbub from the Canadian Parliament behind me patting their tables and gushing heah heah! So let’s just round things out with a couple of jet stream turbines before tensions run too high and one of the hairs on the head of the Right Honourable Stephen Harper springs noticeably out of place, shall we?
For some reason, I’m usually biased toward using suction rather than high pressure in my flow accelerator ideas. But one advantage of using ram air pressure in the machine proposed here is that it would keep the long fabric tube inflated. This is very significant of course, since one of the biggest challenges in designing an airborne turbine is keeping weight to a minimum. Using high pressure might eliminate any rib-like supporting structure that would otherwise be required for the tube. I guess you’d have to stabilize the fabric tube by attaching it to the tethering cables at various intervals, but who knows… maybe somebody can design a way around this requirement.
Triple Tethered Variation
Multiple Blimps Variation
There are many variations of the ideas proposed here, but let me discuss one in particular. This idea emphasizes a technique I’d like to use to bring these pie-in-the-sky airborne turbines a little closer to feasible. Imagine eight blimps. Each is tethered by at least three cables to keep the blimps from moving around too much. An aerial view would reveal that the blimps are situated at the vertices of a gigantic octagon. It is important to note that the “diameter” of the octagon is far from insignificant. I can’t give you a number… maybe two or three football fields? Each blimp has a parachute and a high pressure tube, just as described above. All of the high pressure tubes converge at the center of the octagon, where they connect to a single larger high pressure tube that takes the jet stream wind down to the ground.
What’s so great about this variation? Well… let me first list what I believe may be the salient objectives of airborne turbine design:
- If possible, no moving parts in the air.
- If possible, no fiberglass, electrical cable, gearboxes, drive shafts, or electrical generators in the air. (Ever notice how the components of a wind turbine that have to do with mechanical and electrical power are about the most dense (heaviest) things known to engineering kind?!)
- MINIMIZE WEIGHT, MINIMIZE WEIGHT, MINIMIZE WEIGHT!!!!!!!!!!!
So the idea here is that instead of having eight different tubes, we attempt to minimize weight by having a single large tube carry wind from the jet stream to the ground. This is desirable because the really long distance is from the jet stream to the ground. Once at the center of the jet stream octagon, it isn’t much further to the blimps. So could we use this trick to reduce the overall weight of the machine?
Well, whether this trick will work or not… I think you see my point. What is needed is a kind of linear programming style optimization that minimizes weight of fabric per kilowatt of capacity.
Can We Really Reach the Jet Stream?
No. The jet streams are like 30 to 40 thousand feet off the ground. (The cruising altitude of jet airplanes!) So we can’t reach the jet stream with the design proposed in this post. But we can certainly reach a higher altitude than today’s state of the art wind turbines! If you want to see a more practical configuration that uses the principles described in this post, check out the Practical Artificial Pressure Differential Wind Turbine.
May 10, 2009
A number of posts to this blog describe turbines (both HAWTs and VAWTs) having blades that are supported at the high-speed blade tips rather than at the low speed parts of the blades. This is usually accomplished by having a blade tip engage some sort of slot that is cut into a blade guiding track, so that the action is somewhat reminiscent of the way a rail guides the wheel of a train. One of the biggest problems with this approach is how to come up with a simple, reliable way of converting the kinetic energy of the blades to electricity. The conversion apparatus should not be unweildy or cumbersome, and should not require too much hardware. (For example, in some of my earlier posts I have suggested distributing generator windings all along a very lenghty blade guiding track. This is clearly undesirable because it would make the tracks very heavy and very expensive.)
I think I may have stumbled on a good way to deal with this problem just a minute ago while writing the post entitled: Skyscraper with H-Rotors. I didn’t do a very good job of describing the counter-rotating drive idea in that post, so I’ll attempt to do a better job of it here. (Although the technique described here may be applied to many of the HAWT and VAWT turbines proposed on the Salient White Elephant, you might want to read Skyscraper with H-Rotors first, since I’ll draw the diagrams and everything assuming that we’re applying the counter-rotating direct drive idea to that particular turbine.)
This turbine produces power in pulses. Each time two blades that are traveling in opposite directions pass each other, their generator components (permanent magnets and coils) pass close to each other as well. So a pulse of power is produced when two blades pass each other. Obviously, it would be better for a turbine to produce power at a smooth constant rate. This is desirable for many reasons. For one thing, producing power in pulses applies a cyclic fatigueing load on the mechanical components, and this is obviously bad news. For another, the electricity is easier to process and manipulate if it is produced at a smooth regular rate. But I am hypothesizing that the design proposed here may be a good one because it allows blades to be supported at both blade tips, even as both tips travel at high velocity! This is a tremendous advantage. But the main advantage of this design is that although it allows blades to travel long distances guided only by slots that are cut into blade guides, it does not require for these long distances to have generator components (magnets and/or coils) distributed along these long portions of the blade guides. Instead, the generator components are compact, and are attached to the ends of the airfoils. You can think of all of the airfoils that rotate (say) clockwise as comprising the generator “stator”, while all of those rotating counter-clockwise comprise the generator rotor. Of course, another disadvantage of this approach is that slip rings would be required to get the power away from the blades and into the electrical system. But there’s another advantage as well – the fact that the generator’s rotor and “stator” rotate with equal and opposite rpm’s effectively doubles the relative speed with which the coils and magnets pass each other.
So before closing, let me address one of the biggest disadvantages of the idea proposed here – that power is produced in pulses. First of all, the fact that generator rotor and “stator” components are counter-rotating means that more pulses per second are produced than you might otherwise expect. (The more pulses the better. If we had enough pulses then they’d all bunch together and we’d have continuous power. As a matter of fact, three phase power is produced in pulses as well, yet these pulses combine to produce power that is perfectly constant. Might we find a way to exploit this three phase effect to make the power output from this machine constant? Don’t know, and too tired to think about it right now, so maybe I’ll revisit this later. But anyway it may not matter. I’m not concerned about the electrical pulsing – we can easily deal with that using power electronics. I’m more concerned about the pulsating mechanical loads, because these will fatique mechanical components and cause them to fail. On the other hand, the good news is that this pulsating load is confinded with a small space that is enclosed by the slots that guide the blades. This is good, because the more confined it is, the more options we have for dealing with the cyclic load. One option being, for example, just beefing up the support structure in that area. This is possible because, again, this area is aerodynamically shielded from the outside wind because it lives inside the slot.) Anyway, as I said, the because the blades are counter-rotating, they pass each other at a relatively high frequency. So maybe we can just design the machine to have many small blades (i.e. many blades, each having a short chord). Now when all these blades counter-rotate, we may end up with so many pulses that the output power looks like DC with a ripple on top. (Remember that adjacent blades don’t necessarily need to be separated by a constant angle. For example, just because there are (say) 6 blades that rotate (say) clockwise doesn’t mean that each adjacent blade must be separated by an angle of 360/6 = 60 degrees.)
Ring Generator Option
If the pulsating loads turn out to be a showstopper, then we can always fall back on the ol’ ring generator approach. In this case, we have the advantage that the rings are counter-rotating, thus doubling the velocity between magnets and coils.
April 28, 2009
(Note – there are some errors in this post that I haven’t had time to fix yet, but I’m sure that if you know mechanical engineering you can easily correct the errors yourself. I think this idea might have potential once the errors are corrected. Note also that the torque tube will probably remain fixed with respect to the stationary tower rather than rotating around it. Also note that the struts each need to be connected by a vertical lattice (near the stationary tower) to keep them separated… that is, to prevent the load that tends to bend the ends of the struts towards each other from being transferred to the rest of the structure, thereby defeating the fundamental purpose of the idea.)
(Okay, here’s a pic with some errors corrected, but with no explanation:
This is a 3 bladed turbine, but I have drawn only two blades in order to make the illustration easier to understand. And I realize there are a lot of “legitimate” mechanical designs to realize this concept, likely using gears instead of tires and so forth. But I’m not a mechanical engineer, and so I just want to draw something that will give the real designers an idea they can play with.
Because the tower does not rotate, the rotor can be very tall, very slender, and it can spin at high rpm without becoming centrifugally unstable. But can’t the stationary tower can bend just as much as the rotating tower? And if the stationary tower bends, won’t this cause the rotating part of the structure to become centrifugally unstable just as if the tower were rotating? No. To see this, consider what happens when the middle of a rotating tower bows in response to the lifting forces transmitted to the tower from the airfoil by the middle strut. In this case, the middle of the rotating tower bows in the downwind direction, but its rotational axis does not change. Therefore the mass of the rotating tower has been displaced from the rotational axis, and centrifugal force now acts to cause even more bowing, and the rotor has become unstable. But when the middle of the stationary tower bows in the downwind direction, the rotational axis of the middle struts and airfoils moves along with it. And so although the rotor’s axis of rotation is no longer straight, it is at least centrifugally stable.
Another advantage of this design is that the guy wires are not connected to the tower through bearings. This should provide a big reduction in mechanical losses, since the bearings at the top of a traditionally guyed Darrieus bear a very heavy load – the rotor’s overturning moment. Of course, the overturning moment must be supported somewhere by some bearings. This design has bearings inside the rings that the struts attach to. So is there any advantage in this compared to the traditionally guyed Darrieus? I’m not a mechanical engineer, so I don’t know. Maybe there’s no advantage at all, but I’m wondering if the approach here isn’t better because it is easier to influence the bearings at design time. For one thing, you can spread the load over as many bearings as you want, while the traditional design requires two sets of bearings – one at the top of the tower and one at the bottom. For another thing, the guy wires in the traditional design are not only trying to torque the bearings about a horizontal axis, they are also doing this cyclically, from very low torque to very high torque several times a second. Surely this can’t be good. Of course, the present design also places a cyclic load on the bearings – there’s no way to avoid that. But at least it’s a “typical” load in that it doesn’t try to twist the bearings to a new axis. So maybe this is a better approach. It seems to me that mechanical losses will be decreased by eliminating the torquing thing, but again, I don’t really have the background to know if this claim is accurate.
April 24, 2009
The Circular Wind Dam first described in this post may not provide much improvement over the current state of the art. However, it lays the groundwork for some variations that I think may be very viable indeed. These variations are presented toward the end of this post, but in case you’re in a hurry, here’s a general idea of the type of machine that we will be working toward:
Also, make sure you read the section entitled Advantages of the Wind Dam Over Current State of the Art Wind Turbines, as this section presents some very very powerful ideas.
Circular Wind Dam
A circular hallway is made of brick or concrete. It is extremely high, and it has an extremely large diameter. A number of walls block the hallway inside:
Here is a cross-section of the Wind Dam that shows one of the walls that block the hallway. The wall has a hole cut into it, and a wind turbine rotor captures the energy of the wind that flows through the hole:
The outside of the Wind Dam has holes cut into it. These holes have sliding doors that are kind of like garage doors. The doors can block the holes, or they can open the holes to the outside air:
The holes near the upwind side of the dam and the holes near the downwind side of the dam are opened, while all of the other holes are kept closed. Now wind flows through the hallway from the upwind side to the downwind side. The wind turns the turbine rotors that are embedded into the walls that partially block the hallway, and the turbine rotors drive generators to make electricity. The turbine rotors and generators work just like the rotors and generators on a standard wind turbine, except that gearboxes might not be required between the turbine rotors and generators of the Wind Dam. This is true because the Wind Dam and the holes that house the turbine rotors have a concentrating effect on the wind, so that wind flows through the holes and turbine rotors at a much higher velocity than the velocity of the wind outside the dam.
Concentric Hallways Variation
A wall is added inside of the hallway that separates it into two hallways that form concentric rings:
The previous variation of the Wind Dam had holes in the outside wall – that is, the wall that faces the outside of the dam. In addition to these “outside holes”, the Concentric Hallways Wind Dam has holes that face the inside of the dam (the area enclosed by the dam). The holes facing the inside of the dam are controlled in exactly the same way as the holes that face the outside of the dam. That is, only the holes that are near the upwind and downwind sides of the dam are kept open, and all of the other holes are kept closed. Oddly, wind flows through the inside (smaller diameter) hallway from the downwind side to the upwind side.
Since wind flows through the hallways in opposite directions, we can replace the horizontal axis rotors with a vertical axis rotors. There are many ways to do this, but for the sake of making the diagram easy to draw, let’s use Savonius rotors with flat vanes. Notice that wind will always flow in opposite directions through the hallways regardless of which direction the outside wind is blowing:
Here’s a diagram that shows how wind flows through the dam:
Advantages of the Wind Dam Over Current State of the Art Wind Turbines
- Most of your investment in a current State of the Art Horizontal Axis Wind Turbine (SOTAHAWT) is used to purchase sensitive, short life components that are difficult to design and that are easy to break. The majority of an investment in a Wind Dam purchases the structure itself – basically just a pile of bricks – and this investment will probably still be productive 100 years from today. (Think of the great hydro-dams in the NorthWestern United States.) And if somebody knocks a hole into the wall… big deal – just fill it in with a few more bricks. But suppose instead that an expensive component on one of your SOTAHAWTs fails. Suppose have to replace a turbine blade that is 150 feet long! You buy a new blade, wait 3 months for it to be manufactured, somehow get that 150 foot blade to the site and then weave it in between the other turbines, hire a crane, take the old blade down, put the new one up. You see what I mean – most of your investment might as well be stored in a crystal goldfish bowl that is balanced on top of a 300 foot FM radio tower. My goodness… how can you even sleep at night!? I don’t know how large the turbine rotor blades will be in the Wind Dam, but even if they are 20 feet long… so what? They don’t represent a significant fraction of your investment, and repairing and replacing them are no big deal.
- The Wind Dam is virtually indestructible. It can assume a very low drag profile during storm winds by opening all of the doors in the inside and outside walls of the dam. Its drag profile may be further reduced by installing sliding doors into the wall that separates the two concentric hallways. These sliding doors may also be opened during storm winds to keep the drag profile to a minimum. No large sensitive components (rotor blades, etc) are required to withstand storm wind drag forces.
- Wind turbine rotors, gearboxes (if required), and generators are all located inside the Wind Dam. Thus, these relatively sensitive components are well protected from wind, rain, snow, and other weather. And it would be easy to provide heating for these components during cold weather.
- The Wind Dam is very quiet because noisy components are located inside the dam.
- Electrical components are virtually immune to lightning strikes. This is true because these components are housed inside the dam, and because the dam has lightning rods on top to direct lightning strikes away from sensitive components.
- The life of a Wind Dam will probably far exceed the 20 or 30 year lifespan of your typical SOTAHAWT. It is obvious that this would hold true for the structural part of the Wind Dam (the bricks). What is less obvious is that smaller rotors that turn at higher speeds that drive higher voltage electrical equipment, all of which live in the luxurious indoor environment of the wind dam, will also last much longer than the massive SOTAHAWT components that are stressed, fatigued, and pounded by wind, rain, snow, ice, cold temperature, humidity, and lightning year after year after year. Also, smaller components last longer than larger components because they are easier to design and because they are subjected to less vigorous mechanical abuse than are larger components.
- Since the wind turbine rotors and electrical equipment live inside the Wind Dam, it’s hard to imagine the machine posing any danger whatsoever to the public. Can you imagine what would happen if a 150 foot rotor blade became detached from a SOTAHAWT? I wouldn’t be surprised if it could pierce through the roof of an enclosed baseball stadium. This also reflects in the cost of wind generated electricity. Because of the danger of accidents, SOTAHAWT components are quite significantly over-engineered. But normal safety margins could easily be justified in the design of the Wind Dam, and this will significantly lower costs.
- An oft cited disadvantage of wind energy is that it has a very low energy density. The wind dam solves this problem. I don’t know what the water pressure is at the bottom of a hydro dam, but I know it’s huge. Water is heavy. But the fact that the energy density of water is high means that you need a lot of concrete to hold it back. Maybe your hydro dam is only as long as a couple of football fields, but it has 50 million dollars worth of concrete in it. But the wind dam is holding back something that is much lighter, so the walls do not have to be so massive. Maybe the circumference of the wind dam is 75 football fields, but since the walls are thin, they still have the same total amount of concrete: 50 million dollars worth. Warm up to the idea that the density of investment capital is directly proportional to the density of energy. Spreading energy out over a larger area does not raise the capital cost of the structure required to harvest that energy, it merely spreads the same capital cost out over the same larger area. Getting uptight about the low energy density in wind is like worrying about whether you should stack your money in a pile when you count it, or whether you should lay your dollar bills end to end across the bedroom floor. Actually, if you carpet your house with dollar bills, it looks like a lot more money than if you just put them in a greasy old stack with a rubber band. (But what about storms? Won’t the wind dam have to be massive and expensive in order to withstand storm winds? No. Consider the hydro dam. If more precipitation falls than you expected, you simply open the gates and let the water run through the dam. But exactly the same approach works for the wind dam. If there’s a hurricane, you simply open all the doors and let the wind pass through unimpeded. And if you have trouble making this approach work, remember that the light weight surfaces of a wind dam may be folded up, rolled up, retracted, or even laid down on the ground.)
Real Estate Sharing Variation
The wind dam can be built near areas that have residences or commercial activity, since there’s no possibility whatsoever of an accident, and since the noise is probably mostly contained inside the hallways. Holes can be cut into the walls so that roads can go through. In this case, the air is simply routed over the passageway in an aerodynamically friendly way. You could put a wind dam around a cornfield, and make a passageway big enough for a combine to get through.
Wind Vane Doors Variation
Instead of garage doors that roll up into the ceiling, maybe curved doors could be used instead to further concentrate flow in the hallways:
Very High Altitude Variation
Imagine an extremely tall (500 feet?) circular wind dam. It may have a single internal hallway, or it may have two concentric hallways. The very-high-altitude very-high-velocity wind causes the air inside the hallway(s) to also move at high velocity. This action may be accomplished in any of the ways depicted above: doors that open and close, curved vanes, or via some other appropriate aerodynamic components. However, once inside the hallway, the energy is transferred from the highest part of the hallway to a lower altitude region of the inside of the hallway. This may be achieved by blocking all altitudes within the hallway except for those that are (say) 300 feet or lower. Or it may be achieved by using nearly horizontal airfoils that direct the high-altitude energy downward toward the earth, or by some other kind aerodynamic hanky-panky. Bringing the high energy to the lower elevations allows the turbine rotors, gearboxes, and generators to be positioned at a lower altitude.
As stated in another post on this blog, Synopsis of the Best Design Tricks Developed to Date, if you can’t put the turbine up into the high altitude high energy wind, then bring the high altitude high energy wind down to the turbine. (I wrote that synopsis of best design tricks post in a hurry… don’t be disappointed because it’s actually not a very good post. Hope to have time to rewrite it later.)
Energy Exchange Variation
For ideas on how to design a very tall H Rotor (Straight-Bladed Darrieus), see High Speed Centrifugally Stable VAWT. Of course, any VAWT may be used with the energy exchanging version of the Wind Dam, and the Savonius may be a good choice as well.
Another approach would use H Rotors with a horizontal axis. One end of a rotor’s axis would be anchored to the smaller diameter wall, and the other to the larger diameter wall. The rotor’s axis should be parallel to a line that extends radially from the center of the circle formed by the smaller diameter wall segments. (This is, or course, the same point that is the center of the circle formed by the larger diameter wall segments.) And, or course, the H Rotor’s axis would also be parallel to the ground. Anyway, you stack these horizontal axis H Rotors one on top of the other so that they form an “aerodynamic barrier” that extends from the ground to the height of the two walls (smaller and larger diameter walls). Since the axes of these rotors are all parallel, they can all be connected with a common chain drive. The chain extends to a sprocket which is on the generator shaft, and the generator shaft is on the ground.
Furthermore, if it is determined that HAWT rotors would provide better efficiency or this design, simply replace the 3 bladed H Rotors above with walls that can rotate through an angle of 180 degrees. Now cut round holes in these walls, and install HAWT rotors in the holes. Now rotate the wall by 180 degrees if the wind passes through the holes in the direction that is opposite the direction that the HAWT rotors are intended to act upon.
Rotated Energy Exchange Variation
I wonder if you could put some kind of a smooth geodesic dome structure on top of a Walmart to smooth out all the turbulence generated when wind pushes up vertically from the walls and tries to make the right angle turn to flowing across the roof. And I wonder if you could build a wind turbine that looks a little like the one above and put it on top of that dome roof.
Controlling Vortex Shedding
High Altitude Variation
Suppose the wind turbine blades extend from an elevation of 50 feet up to an elevation of 150 feet. We would like for the walls of the wind dam to reach an altitude of 300 feet. This allows us to extract some of the energy from higher energy density wind at altitude. There are a variety of ways I can think of to do this, but I’m wondering if all that is necessary is to make the walls of the wind dam lean in one direction or the other:
At first, this solution would seem to be sensitive to wind direction. But I’m wondering if maybe that isn’t the case. Imagine, for example, wind flowing from right to left in the diagram above. Intuitively, it would seem that high energy high altitude wind would be driven downward towards the earth by the slanted portion of the wall. This is reminiscent of a flow concentrator. But if instead wind flows from the left to the right, high energy high altitude wind ramps over the slanted portion of the wall, reminiscent of a flow diffuser. In this later case, wouldn’t some of its energy still be transferred to lower altitude wind, albeit through the action of suction rather than through the action of compression?
Of course, it is also important to note that a certain amount of the high energy high altitude wind will be deflected downward (toward the earth) even if the entire wall is vertical (rather than having the upper part slanted).
In any case, if a leaning wall proves to be useful, note that it is easy to build one by anchoring fabric to the guy wires of a tower:
Better Diagrams of High Altitude Variation?
I’m not much of an artist, and I’m using 2D software to boot. But here’s an attempt at rendering one of the variations that captures energy from high-altitude winds without using tall turbines:
Circular Wind Dam Pseudo-High-Altitude Variation #2
Circular Wind Dam Pseudo-High-Altitude Variation #2 Stacked on top of Rotated Energy Exchange Variation
Just stack the high altitude part:
On top of the low altitude part:
Put the towers inside the brick walls, and make the lower ends of the tarps come approximately to the tops of the brick walls.
Circular Wind Dam Pseudo-High-Altitude Variation #3
Using HAWTs instead of VAWTs
Of course, HAWTs may be substituted for VAWTs in the above designs. Just block the regions of accelerated flow with walls, cut holes in the walls, and put HAWT rotors into the holes. The only other necessary modification is that you’d have to figure out a way to “yaw” the HAWT by 180 degrees. This is so because the wind may flow in either direction through the hole, depending on the direction of the ambient wind.
If a non-circular path is properly designed, the wind dam will still be equally effective (or almost equally effective) regardless of wind direction. For an explanation of how to design a non-circular path, see the earlier post 20 Megawatt Direct Drive Darrieus.
The diagram looks silly, but don’t dismiss this post too quickly – I think this is a terrific idea:
There are many variations, but as we explore those variations, keep in mind the salient points of this machine:
- direct drive
- not only is the generator at ground level, the turbine rotor is on the ground as well!
- incredibly scalable
- low maintenance.
Suppose we have five towers in the shape of a giant pentagon. Cables run between adjacent towers, and the shroud is attached to those cables. In this case, only the shroud is yawed:
Or it can look like this:
Instead of making the shroud out of fabric like nylon, it can be solid. It would look kind of like the top half of a radar dish. In this case, it might be supported by a number of lattice structures on wheels.
Single Tall Tower Variation
Yet another variation would use a single very tall guyed tower. Because the bottom of the shroud is like a circular arc that sweeps through 180 degrees, the diameter of the arc at the lower end of the shroud can be made larger than the diameter of the anchor points for the guy wires. A long horizontal tube at the top of the tower supports the top part of the shroud. The turbine rotor and generator are on rails or wheels and trace out a circle as the machine is yawed through 360 degrees.
Alternatively, the guy wires are used not only to support a very tall tower – they also support the shroud. Suppose there are 5 guy wires. A cable connects the 5 guy wires so that an aerial view of it looks like a pentagon. There are several levels of these horizontal pentagonal cables, so that an aerial view of them looks like several pentagons of different sizes. The shroud is connected at each side to each of these cables, and rides these pentagonal cables as it yaws:
Just like the single tower variation just described, except now the entire surface traced out by the guy wires is covered with fabric. It looks like a giant American Indian tipi. Now all of the fabric that runs the circumference of the tent from an elevation of (say) 20 feet to 30 feet is separated from the rest of the fabric. It looks like a giant pentagonal wedding ring. This ring has two holes cut into opposite sides of it, and a fabric tube connects these two holes. The tube has the turbine rotor inside of it. The tube and the ring yaw, and the rest of the tent is stationary. The turbine is near the tower (or coaxial with the tower) so that it doesn’t have to move much when yawing.
April 21, 2009
I’ve posted a number of wind turbine ideas to this blog that have chain or belt drives, so I want to dedicate this post to a multi-speed drive for these turbines. It’s too hard to draw chains, so the diagram below illustrates the idea using belts and pulleys:
The purple wheels are behind the yellow wheels. Each set of coaxial orange and green wheels are mounted on the same shaft. The green wheels engage the front belt, and the orange wheels engage the rear belts. The diagram shows all three sets of coaxial (orange and green) wheels engaging the two belts at the same time, but this is just for the sake of clearly illustrating the geometric relationships. In reality, only one set of coaxial (orange and green) wheels would be lowered down until they engage the two belts, and the other two sets of coaxial (orange and green) wheels would be raised up a little so that they do not engage the belts.
I guess this drive might be useful for other applications, such as for a multi-speed drive for a bicycle.
Ideal Path for Wind Turbine Blade
Doesn’t it seem like the ideal path for a wind turbine blade would have the entire blade moving at “tip speed”, spending the majority of its time traveling in a direction that is orthogonal to the direction of the wind?
Description of the Machine
(Note that the word “power cable” in the diagram below does not refer to an electrical cable, but rather to a cable that carries the mechanical power to the pulley wheels at the generators.)
Drag on Moving Cable
How much aerodynamic drag does a moving cable add to the system compared to the drag of a strut or the low speed part of a troposkein Darrieus blade? Consider an H Rotor supported by a strut. The instantaneous velocity of a given radial location on the strut is at right angles to its longitudinal dimension. For this reason, it is able to leave a semi-evacuated space behind it as it moves. This space is filled with turbulent air. But the velocity of a given cylindrical cross-section of a cable is along its longitudinal dimension. Because there is always more cable following any cylindrical cross section, it is impossible for the cable to leave a semi-evacuated space in its wake. The only semi-evacuated, turbulated space it can leave is on its downwind side, and the force thus produced cannot drag down the system because it is orthogonal to the velocity of the cable. The only aerodynamic drag that is possible is that drag that is created by the cable’s ability to “stick to the air around it”. I’m not an aerodynamicist, but I’m guessing this might turn out to be pretty small. Also, the moving cable in the machine presented here has small diameter because it moves at “tip speed”, and because it is only required to carry the torque at tip speed. Because tip speed is high, torque will be low, and the cable will have small diameter.
One obvious disadvantage of the oblong path is the need to yaw the track. We can eliminate this problem by providing the turbine with a polygonal path that approximates a circle:
Airfoils Could be Slowed to Go Around Tower
To avoid inflicting excessive centrifugal force on the blades as they go around the tower, maybe the control system could sense the point at which blades are nearing towers, and slow them down somewhat as they make their way around the tower. Alternatively, mechanical support could be provided on the opposite side of the airfoil to prevent excessive bowing due to centrifugal force:
Low Drag Airfoil Variation
The airfoils need to have gaps near the regions where the load carrying cables penetrate them. This is necessary to accomodate the mechanical supports for the apparatus that steers the airfoils around the tower. But if the airfoils carry a small generator, like the generator that rubs against a bicycle tire in order to power a light for the bicycle, then these gaps could be filled with streamlined shapes during the long path between the two towers. The streamlined shapes would be pulled out of the way as the airfoil makes its way around a tower. The actuating mechanism is powered by electricity that is generated from the small generator just described that is located inside the blades.
Alternatively, maybe the streamlining shapes could be pushed out of the way by some mechanical means as the airfoil goes around a tower.
I don’t have near enough time to do justice to this idea. I’ll just post it in very abbreviated form and leave the rest to your imagination.
The chain-like structure that protrudes from the inner perimeter of the rim of the spoked wheel is essentially just a bunch of rollers whose axes are parallel to the rotor axis of rotation. The fact that these engaging mechanisms are rollers (rather than just fixed pegs) permits very efficient transfer of mechanical power from the rotor blade to the chain drive.
The high speed generator shaft can be attached to a second smaller (aerodynamic) rotor. This permits more efficient conversion of energy flowing through the axial region of the rotor disk, and eliminates the big fat twisted small-radius section of the larger rotor blade.
The chain drive may be replaced with a high-speed rotating shaft whose axis of rotation is coaxial with the longitudinal dimension of the blade (coaxial with the blade spar). The rotating shaft goes down the center of the blade just like the chain drive. A tire at the blade tip engages a ring that looks like a giant washer and that protrudes from the inside of the rim of the spoked wheel. The tire drives the high-speed shaft that goes through the center of the blade. Assuming a 2 bladed turbine, the two shafts will be counterrotating. One of these shafts drives the generator rotor and the other drives the “stator” in the opposite direction. In this design, the generator rotates with the blades and slip rings will be required to transmitt power from the rotating generator.
Yet another variation has the chain drive or high-speed rotating shaft running longitudinally inside the spar of an H-Rotor Darrieus. In this embodiment, either the generator can rotate with the blades and transmitt its power out through slip rings, or else another sprocket near the tower protrudes from the spar and drives a rotating ring which in turn drives a generator that is fixed with respect to the turbine foundation.
April 11, 2009
April 1, 2009
There are many variations of the turbine described in this post. I’ll just go through the ones I have time to present one by one.
Two Cables Move With Symmetrical Airfoils
This variation may move along a polygonal path that approximates a circle.
Alternatively, the entire path (cables, towers, and all) may be yawed to keep its long dimension at a right angle to the approaching wind:
Single Cable Moves With Inverting Airfoils on Circular (Polygonal) Path
This variation moves along a polygonal path that is approximately like a circle. The following diagram shows how the airfoil center of mass is near the center of the cable. Because the gravity moment is small, the orientation of the airfoil will be controlled by aerodynamic forces whenever it is not traveling directly into or directly out of the wind. In these cases, the airfoil has only one stable orientation with respect to its aerodynamic forces. That stable orientation is the one in which the blade tips are downwind from the cable. Any other orientation is unstable. Whenever a change in the direction of the airfoil velocity vector causes it to become unstable, it will flip over to the stable orientation. This allows the use of pitched, asymmetrical airfoils. (Pitched, asymmetrical airfoils deliver better aerodynamic performance than the zero-pitch, symmetrical airfoils that are typically used on a Darrieus turbine.) When the airfoil is traveling directly into or directly out of the wind, its orientation will be controlled by the gravity moment. In these cases, the only stable orientation has the blade tips lower than the cable. From an aerodynamic perspective, we don’t care how the airfoil is oriented when traveling into or out of the wind, since in these cases it cannot produce power anyway.
In the following diagram we can see that a gearbox is not required for this turbine. This is true because the airfoils and the cable are traveling at what we normally think of as “tip speed”. Suppose the wind speed is 20 miles per hour (mph). If the tip speed ratio is 6, then the airfoils travel at 120 mph, or 176 feet/second. If we want the generator to spin at 1,200 rpm, then the wheel that engages the cable must be about 2.8 feet in diameter, which sounds like a pretty reasonable number to me.
The diagram above has been simplified – it does not show how the airfoil is prevented from hitting the tower or the generator. The following diagram adds a semi-circular guide that rotates the airfoil whenever its orientation is going to cause it to strike the tower or the generator:
The wheel that is closest you in the diagram above is the one that is situated at 3 o’clock. (The one with the cable directly behind it.) It is possible that the airfoil will strike this wheel exactly in its center. The design of the closest wheel causes it to move slightly upward and toward the tower whenever this happens, thus diverting the airfoil down the lower part of the guide.
Single Stationary Blade Guide Hangs From Cable, Yawing Oblong Path
When an airfoil is supported at both ends using two blade guides like the ones pictured above, inverting the airfoil isn’t a big deal. It would seem at first to be impossible to support an inverting airfoil with only a single blade guide, since in this case the structure that connects the airfoil to the blade guide will always hit the blade guide’s supporting structure when going around the tower with one of the two airfoil orientations. But here’s how we get around this problem. In the diagram above, the orientation of the airfoil as it approaches the tower is such that the blade supporting structure (black) will not hit the blade guide support (purple). We wait until the airfoil has rounded the tower and is ready to begin its long linear path to the next tower, then we invert the airfoil. In other words, we wait unti the airfoil has passed the blade guide supports before inverting it. Now when the airfoil is approaching the tower on the other end of the oblong path, we invert the airfoil before it gets to the blade guide supports. For this reason, when the airfoil completes its trip around that tower, it already has the correct orientation for traveling its long linear path back to the other tower.
Another interesting twist to this variation would use a solid weight for toggling the airfoil center of mass instead of a liquid weight. In this case, the motion of the weight during inversion can be damped by adding two small holes that vent each end of the cylindrical cavity to the outside air. With this design, the weight moves slowly from one end of the cavity to the other because it must push air out through one ventilation hole and pull air in through the other. If it isn’t practical to vent the cavity without risking the intrusion of rainwater, then the holes at each end of the cavity may be connected with another very small diameter cylindrical cavity. Neither of the two cavities is vented to the outside air in this design. As the weight travels from one end of the large cylindrical cavity to the other, it must first transfer air from one side of the large cavity to the other by forcing it through the small diameter cavity.
March 27, 2009
Although the direct drive apparatus isn’t shown in the diagrams, a written explanation should suffice. Permanent magnets are attached to the blade supporting structure that is at each end of the airfoils. The airfoils drive these magnets at a speed that is analogous to the “tip speed” of a more traditional wind turbine. Generator windings are built into the blade guides (aqua colored components in diagram above). Instead of embedding windings into the entire length of the blade guides, windings are separated somewhat. Blades speed up slightly while traversing the distance between windings.
The blade inverting sections of the track permit the use of asymmetrical, pitched airfoils. (Asymmetrical, pitched airfoils deliver better aerodynamic performance than zero-pitch, symmetrical airfoils. For an explanation of the “helical airfoil inverting section” of the track, see my earlier post: 20 Megawatt Direct Drive Darrieus.)
I wonder if the blade guides can be flexible? If so, the flexibility may provide a number of benefits. For one thing a flexible guide might be cheaper, lighter in weight, and easier to build. For another, it may help to absorb abruptly changing loads due to (for example) wind gusts. It might be interesting to explore the possibility of a blade guide which is flexible enough to resemble, to a degree, a hanging cable, and yet which is still rigid enough to accurately maintain the tight mechanical tolerances that would be required for the direct drive generator components contained within it:
Two Hanging Cables Variation
(If you haven’t yet read about why you might want to invert the blade, click here.)
It might be interesting to explore how the Eye of the Cat Rotor Blades would perform when hanging from two cables. Although there’s no centrifugal forces to balance, remember that the Cat’s Eye Rotor also balances aerodynamic forces. In this case, think about what happens if the blades are longer than the separation distance of the cables (including slack… so that they tend to keep the cables pushed apart).
Single Hanging Cable Variation
In this variation, sensors detect when two airfoils get too close to each other. In this case, the direct drive generator coils for the leading airfoil are switched off until it reaches an acceptable distance from the airfoil immediately behind it.
March 24, 2009
Only the airfoils rotate in this machine. The flow accelerators are supported in the manner of a wheel with spokes. Each end of each airfoil is attached to a cable that moves like a rotating ring inside of its flow accelerating shroud. The cables drive the generators. (Generators not shown in diagram.) Alternatively, the flow accelerating shrouds can house direct drive generators in the manner described in an earlier post entitled Flow Accelerating Ring Generator for Horizontal Axis Wind Turbine.
It might be worth considering a combination of this idea and the one described in Highly Scalable Horizontal Axis Wind Turbine.
March 23, 2009
The diagrams are not very well done, and certainly not sufficiently detailed. But basically what happens is that a blade catching mechanism follows the blade that it catches through the arc shaped guide. After the blade moves past the end of the guide, the blade catching mechanism begins its return trip back to the beginning of the arc shaped guide. The blade catching mechanism will arrive at the beginning of the arc just in time to catch its blade as it comes back around. All three of the blade catching mechanisms behave this way, and they are synchronized in order to catch each blade with little mechanical shock on impact (i.e. a blade catching mechanism is moving at the same speed as the blade at the moment it makes contact with the blade).
The mechanical system that incorporates the three blade catching mechanisms also drives the (off-the-shelf) generator.
March 19, 2009
This machine has a blade tip that simultaneously performs three functions. It is a kind of elongated, arc-shaped vortex spoiler, and it is pitched relative to the oncoming wind so that it also accelerates flow through the rotor disk. It’s third and final function is to streamline and support the permanent magnets of a direct drive ring generator:
How can the (generator) rotor pass through the stator without hitting it and still maintain the close tolerances that are required? I don’t know, but perhaps some sort of guide system can be designed that routes it through the stator. The following diagram gives the general idea. It shows a wheel approaching a guide. The wheel is not necessarily aligned with the guide, but the funnel-like shape of the guide causes it to make any necessary adjustments in position.
One disadvantage of the Flow Multiplying Anti-Vortex Drive is that the structure housing the stator coils must yaw with the nacelle and rotor. However, if the drive idea is used together with the Highly Scalable HAWT idea, then the stator coils may remain fixed, and do not need to yaw with the rotor blades.
March 2, 2009
The Highly Scalable Wind Turbine is remarkable in that guy wires assist in supporting all of the large tower loads that are carried by the machine. These loads include:
- overturning moment on central (vertical) tower tube (due to thrust of operating rotors or drag during storm winds),
- moment that tends to bend the ends of the upper (nearly horizontal) tube in the downwind direction (due to thrust of operating rotors or drag during storm winds), and
- moment that tends to bend the ends of the upper tube down towards the ground (due weight of nacelles and blades).
Note that the configuration of the Highly Scalable Wind Turbine allows guy wires to attach to the tubes they support at any desired location along the tube’s longitudinal dimension, including at the very end of the tube. This feature facilitates the design of a very large turbine that does not require a correspondingly massive tower. It also permits the use of very high towers on sites that have favorable wind shear.
The top tower tube is just long enough to allow the rotors to clear the guy wires that support the central tower tube. Tubes or lattice structures may be substituted for guy wires if greater control over tower resonance is required, or if an upwind design is selected.
Note that the nacelles do not rotate with respect to the top tower tube. Instead, the nacelles and the top tube rotate as a unit about the central vertical tube in order to regulate yaw angle. The top tube is bent in order to lower the center of gravity. This way, the upper structure will naturally want to correct any tendency for one rotor to dip lower while the other rises up. In fact, it may even be desirable to allow limited motion in this direction so that the structure doesn’t have to bear the associated load. (Of course, using a bent tube in the upper structure rather than a straight tube also results in a tendency for the rotors to rock “out of the page” in the upwind view above.)
Although the machine in the diagrams has two-bladed downwind rotors, it is readily apparent that this configuration would be equally effective with upwind or downwind rotors, regardless of how many blades they have. (Some guy wires will be replaced with tubes or lattice structures if an upwind design is selected.)
Teetering to Shed Loads
The Highly Scalable Wind Turbine has a remarkable ability to yield in response to sudden changes in loads. If a wind gust hits one of the rotors but not the other, the machine will compensate by yawing. If two bladed, downwind, teetering rotors are used, then each rotor can teeter in response to variations in the wind that flows through its rotor disk. In selecting turbine configuration options, one must remember that providing more degrees of freedom will render the machine more difficult to understand and model, and may also lead to instabilities.
It may be useful to explore the option of letting the upper structure teeter about both of the axes that are orthogonal to the yaw axis. In this case, the upper structure “balances” on top of the vertical tube.
Tilt Down Option
The highly scalable turbine has yet another extraordinary feature – it is naturally counter-balanced! This means that a tilt-down version may be designed without adding expensive, heavy components. Both nacelles and rotors may be serviced at ground level, or at least near to ground level.
Here’s a folding tilt-down version:
I guess this is getting pretty outrageous, but imagine the controller folds the turbine not only to facilitate service and repair, but also in response to storm winds! The procedure might go something like this:
- Both rotors are feathered,
- the brakes lock both rotors at appropriate angles,
- the machine is yawed to an appropriate angle, and finally
- the upper structure is folded as depicted above.
Might the controller do all this without human intervention? Seems pretty far out, but if this can be achieved, what a fantastic high-wind drag profile!
Another high wind shutdown option would simply have one rotor produce (say) 50% power while the other rotor produces only (say) 20% power. In this case, the rotor operating at 50% capacity would be dragged to the most downwind location, and the machine would maintain this somewhat low drag profile throughout the storm. This action is reminiscent of the way some small HAWTs turn their rotors to the side in high winds. It offers the advantage of quickly and continuously yawing in response to changes in the direction of the storm wind.
Advantages of the Highly Scalable Design
I am hypothesizing that the scalable design presented here is superior for several reasons. The tower of the traditional horizontal axis turbine must support the total aerodynamic drag on the rotor and nacelle during storm wind conditions. Since this force is applied with a lever arm equal to the height of the nacelle, the tower must be large, heavy, difficult to transport, difficult to erect, and it must be expensive. The vertical tube of the scalable machine is strengthened by guy wires, allowing it to be lighter, more manageable, and less costly. The penalty is the addition of another tube (the upper structure tube) that must support the weight of the rotors and nacelles. But this force is applied with a lever arm only slightly longer than the length of the rotor blades, and there are guy wires above the tube to help carry this load. Making the worst case assumption that the scalable turbine is not folded or tilted down, and that its upper structure tube is at a right angle to the direction of the wind, then the scalable turbine’s upper structure tube must support the same storm wind aerodynamic drag as the one supported by a traditional turbine. But again, the lever arm is only slightly longer than the blade length, and there are guy wires on the upwind side of this tube to help. There are also guy wires on the upwind side of the vertical tube to help carry this load, and these guy wires may attach to any place on the vertical tube, even at the very top of the tube!
In the early 1990s, Carter Wind Turbines, Inc. developed a downwind, tilt-down, two bladed 300 kW turbine. This machine was considerably lighter and more flexible than the more common upwind, three bladed turbine. The lightweight, small diameter tower was possible because it had guy wires. It is important to realize that this tower was strong enough to:
- withstand storm wind aerodynamic drag loads applied with a lever arm equal in length to the height of the nacelle, and
- support the weight of the nacelle and rotor (applied with the same lever arm) when the tower was in a nearly horizontal position. (This would be the case when tilting the machine down.)
If the provision of guy wires can make the lighter, less costly Carter design feasible, then it can make the lighter, less costly scalable design feasible as well. This is so because the scalable turbine must support exactly the same loads. In fact, the lever arm for the scalable turbine load is likely to be shorter than than the lever arm for the load on a traditional machine. Furthermore, it is very important to realize that the Carter machine achieved all this with guy wires that could reach no higher than the height of the tip of a blade in the 6 o’clock position. The scalable design enjoys the luxury of guy wires that can attach to its tubes at any desired longitudinal position.
Now that utility scale turbines have reached a size of two or three megawatts, transportation of turbine components and construction of the turbine have become very significant cost and technology issues. A turbine’s maximum size may be influenced by some combination of the following factors:
- cost of transportation,
- feasibility of transporting the turbine’s largest components to the wind farm site,
- cost of construction,
- maximum weight that can be lifted by a crane,
- maximum height that can be reached by a crane,
- cost of the crane.
The industry is attempting to design larger and larger wind turbines because increasing scale potentially lowers the cost of energy. However, the benefits of scale may be eroded by any or all of these factors. The highly scalable turbine proposes to solve this problem by reducing the size and weight of tower components, and by facilitating the design of a tilt-down machine. The tilt-down design allows tower top components to be assembled either on the ground, or at least at a much reduced elevation. Furthermore, since the scalable turbine has two rotors rather than one, it provides twice the power of the traditional horizontal axis machine for a given blade length. Of course, it also has two rotors, two gearboxes, and two generators. But it has only one yaw drive, transformer, and foundation. And finally, it makes more efficient use of tower resources. You can see this by noting that the scalable turbine extracts twice the benefit from every extra meter that is added to the height of the vertical tube.
If the pitch of the rotor blades is variable, then the machine is easily yawed as follows:
- If a rotor is feathered, the machine will yaw in the direction that will move the feathered rotor in the upwind direction.
- If a rotor is stalled, the machine will yaw in the direction that will move the stalled rotor in the upwind direction.
If the pitch of the rotor blades is fixed, then equipping each blade with a spoiler will allow the machine to be yawed in a manner similar to that just described. That is, if the spoilers on the blades of one of the rotors are deployed, then the machine will yaw so as to move the rotor with active spoilers in the upwind direction.
Another yaw drive option would employ a tail fin. The highly scalable turbine configuration certainly has plenty of room for a yaw fin. The yaw fin may be mounted on a supporting tube that extends in the downwind direction from the vertex of the bent tube in the upper structure.
Yet another possibility would use a yaw rotor that is similar to a helicopter tail rotor. Like the tail fin, the yaw rotor would be suspended in a position downwind from the vertex of the bent tube. It would rotate about a horizontal axis that is parallel to the plane of the upwind view in the first diagram. The yaw rotor has symmetrical airfoils with zero pitch (i.e., the airfoils are symmetrical about the plane of rotation). Such a rotor will generate a strong correcting yaw moment whenever the yaw angle deviates even slightly from its correct value. If the yaw rotor airfoils have variable pitch, then the turbine can be yawed even when there’s no wind. This may be useful for untwisting the power cables, and for adjusting the yaw angle to a value that is appropriate for tilting the machine down.
Untwisting Power Cables
One final yaw drive strategy that might be effective for untwisting power cables would use one of the rotors as a propeller. Suppose there’s no wind. One of the generators drives its rotor like a propeller, yawing the upper structure until the power cables have been untwisted. If the machine does not have any other active yaw drive system, then this same trick may be used to adjust the yaw angle to a value that is appropriate for folding or tilting the machine down.
Synchronized Drive Variation
Since the rotational axes of both rotors are parallel, I wonder if they could by synchronized with a sprocket and chain drive (kind of like the drive on a bicycle). In this case, both rotors could drive a single generator that is situated somewhere near the top of the vertical tube. This has the added advantage of reducing the amount of weight that the upper structure tube must support, and it would probably also stabilize the upper structure somewhat, reducing any tendency to tilt from side to side. In fact, maybe the generator could be lowered somewhat in order to further lower the center of gravity of the upper structure.
Semi-Direct Drive Option
The following semi-direct drive option is possible because the nacelle is fixed with respect to its supporting tube. This design eliminates the gearbox.
Or the streamlined ring can be pinched between two tires:
February 24, 2009
The wind turbines once manufactured by Carter Wind Turbines, Inc. utilized innovative fiberglass technology. The spar for the blade of the Carter wind machine was made of fiberglass. It looked like a giant wooden popsicle stick. It was spun from fiberglass thread. The fiberglass thread was wound around a mold much in the way fishing line is wound up on the reel of a fishing rod, except that shape of the object supporting the strand was oblong instead of cylindrical. The gears for the fiberglass gearbox would be manufactured in a similar way. First, fiberglass thread is wound into a thick disk shape. Next, a metal ring with teeth is mounted on the rim of the fiberglass disk. The woven strand fiberglass technology was originally used to make pipe. A long cylindrical shape was spun from fiberglass thread in a manner similar to that described above. The outer case of the fiberglass gearbox might be cylindrical in shape, and might be spun from glass in the same way. The shaft of the gearbox might also be spun from glass.
Advantages of Fiberglass Gearbox
The fiberglass gearbox may prove lighter and less costly than the traditional metal gearbox. Also, objects spun from fiberglass strand tend to be robust and flexible. If the components of a fiberglass gearbox do not prove to be overly flexible, then the added flexibility may provide damping and increased resistance to damage from mechanical shock.
I had the idea for the fiberglass gearbox in connection with the Scalable Tower for Very Large Wind Turbine. The fiberglass gearbox may prove to be lighter and more robust than the metal gearbox currently used for wind turbines. However, if such a gearbox is feasible, it would not provide much economic benefit when used in a turbine that has a traditional HAWT tower. This is so because a lighter (lower cost) nacelle does not result in a lighter (lower cost) tower. But if the scalable tower proves viable, then perhaps it would be worth looking in to the possibility of making a gearbox from fiberglass.