Servo Pitch Control

[Murlin]

Here is an Idea I had about a variable pitch controlled rotor.

The push rod will work through the drive axis with a servo/encoder on the end that is is driven by a ball/lead screw.

This is a different turbine design that incorporates twin taper bearings and a hollow drive shaft.

The variable pitch will be on the front side of the tower pole and the alternator on the back side to balance the system.

Criticism is always welcome, good or bad....

Thanks for your time.

Murlin

[SamoaPower]

I guess it just goes to show that great minds think alike murlin. Ha!

This is just what I'm doing except for the bearing hub. Mine is just a trailer hub that I bored out the spindle for. The pitch push rod is driven by a cut-down satellite dish positioning linear actuator. There's a bearing on the push rod that mates with the pitch collective to allow rotation.

Works fine.

[Walt]

Murlin & SamoaPower

I have been following your post of ideas on furling and yaw control. I like the way you think. It's always good to try better ways of control.

Keep at it,

.

[kamikaze]

Hi Murlin,

Your idea certainly looks like it will work just fine. Just one point however, if you are planning to use it to control the maximum rpm, ie overspeed control either by increasing pitch, or decreasing pitch until stalled, then I believe you should consider a failsafe backup, maybe in the form of a conventional furling tail. The downside of an electro-mechanical system like yours is that if it fails the blades will remain fixed at whatever angle they were at at the time of failure.

 I definitely think it's an idea worth pursuing though. I am thinking about a similar system myself.

Regards, Kamikaze

[RP]

I wonder if a parallel system could be used for backup.  

Maybe a simple second drive system to yank the blades to stall without any of the fine control stuff if it overspeeds past the normal range of the primary system.  This could be separate motor or maybe a big honking solenoid (you'd have the amps to drive it in overspeed).

[commanda]

In the first picture, the grey coloured right-angled piece needs to have an articulated joint at the elbow, so the brown coloured link piece can move through an arc.

Amanda

[Murlin]

Yes Amanda you are correct, I was thinking about that last night.

At first I had thought the brown arm would be slotted to allow for this but I think a joint at that intersection would be better.

Murlin

[Murlin]

"he downside of an electro-mechanical system like yours is that if it fails the blades will remain fixed at whatever angle they were at at the time of failure."

If the pitch on the leadscrew is coarse enough, and if the mechinism is spring loaded, the ball nut will have zero friction and will return the blades to zero pitch.

What got me thinking about this was when I was moving one of my larger CNC mills.

If you did not chain down the table it would roll all over the place when it was tilted out of level.

The only other thing would be axis runaway.  If and then it would run to either limit and blow a fuse, loose power and the blades would bleed to zero pitch, forced y the springs.

The springs are not shown in the picture.

[kitno455]

murlin- why cut windows in the back plate? that just makes those gray pieces require a much larger right angle design. i think with a more traditional flyball governor design (straight gray link, with hinge at outer end), you could even use weights as the primary system. your servo would just take the place of the spring. you would want to use a pretty coarse leadscrew, though.

allan

[Murlin]

The reason I was leaning towards a twin bearing design was that the push/pull rod was actually a leadscrew.

The ball nut would be on the front where the tri arm is.

A servo motor would drive the screw by a rubber gromet as to be totally insulated.

The encoder will also be in its own closed loop system and have an on board battery to supply power to the wireless.

This is a far more complicated system than the trailer hub, but not so complicated as to be impractical.

The blades I want to use are going to be a foam core laminated fiberglass.

I will be making a mold for that.

So since the blades are going to be more flexible than wood, they need to be moved out from the tower.

If this works, I will be making several units.

That is why I am making a mold for the rotor blades....I figure the 10' rotor blades would weigh in at about 10 lbs each..

I was going to do alot of testing for airfoils, but decided that you were right and there was not much I could do that hasn't already been done by someone else more qualified and smarter than I.

So I have chosen the SG6050 airfoil.

My mold will be similiar to the resin-transfer mold posted a while back, except that it will probably not be vacume driven, have water running through it,  and just be made out of wood and fiberglass.

Any air bubbles that cause surface blemishes will just get filled in prior to painting the blades.

I look forward to your alternator tests on your new machines.

Murlin

[Murlin]

I have never seen a flyball design and this is what I came up with on my first attempt at the concept.  There is room for improvement to be sure

The reason I cut the poviot axis through the plate was to be able to use the largest lever possible as to minimize servo load.

If you have any links to the flyball design, please post them so I can take a look see...

Many thanks....

Murlin

[Murlin]

I love not being able to edit my posts

Anyways   If I didn't have to cut through the plate at all and keep everything on the front, it might be possible to just do all this on the front rotor.....

Maybe you could just use a thicker front rotor plate so as not to warp the mags and scrape the stator...

Murlin

[Murlin]

I wonder if a parallel system could be used for backup.  

Maybe a simple second drive system to yank the blades to stall without any of the fine control stuff if it overspeeds past the normal range of the primary system.  This could be separate motor or maybe a big honking solenoid (you'd have the amps to drive it in overspeed).

OK I have been thinking about my approach and you are correct.

The springs must keep the blades at the highest TSR possible so it would start producing power in the low winds without the servo using any power. Otherwise you would be using more power than you are making to power the servo motor.

Then when the system started producing too much power for the current TSR value, the servo needs to come on and start pitching the blades to a slower TSR and reduce the RPM.

You will be making plenty of power by then to power the servo motor.

A backup solenoid that grabs the pushrod and shoves the whole thing to zero pitch would be nessessary for sure.

Perfect.....now to design the variable pitch hub with reguards to material thickness and such...

Good thinking RP

Murlin teh slowly figuring it out...

I love it when a plan comes together...

[Murlin]

My goodness I have more posts to myself than replies...

Oh well....

One more note on a back shut down.

Flux mentioned an oil/heater overload to shut down the stator when it was getting too hot.

I think this would be the best way to shut it down short of furling.

You could have either a backup rudundant brake or another pushrod running up through the center of the tower yaw axis to manually pitch the blades to zero.

I like the idea of a manual pitch control lever....it could still be run by another motor on the ground so you could shut it down via your computer.

My turbine will be controlled by a program running on my PC.

I will be able to manually input an offset to the running program that is running to adjust the pitch angle.

The console will show the RPM, wind speed, volts, amps, stator heat, angle of the blades ect.

It will have a manual override and shorting out the turbine will be done by a soleniod in the breaker box and run at the console as well.

That is my plan...

[SamoaPower]

murlin,

"The reason I was leaning towards a twin bearing design was that the push/pull rod was actually a leadscrew.

The ball nut would be on the front where the tri arm is."

I don't really see how this is going to work. So, your push rod is not really a push rod but a rotating lead screw. The ball nut would have to be carried in a bearing at the pitch collective to prevent rotation with the rotor hub. But even so, bearing friction would still try to rotate the nut on the lead screw, changing the pitch.

"The blades I want to use are going to be a foam core laminated fiberglass."

I suggest you really pay attention to the spar system. Some others seem to be thinking they don't need a spar and are just depending on the glass skin to provide flex strength. I suggest an aluminum tubular spar since that mates up well with the stub shafts of the VP hub.

"So I have chosen the SG6050 airfoil."

At least two others are using this airfoil (sort of) but I don't think they have been able to come up with the polar plots in order to properly design the blades. You might see if you can find that data before making a final choice. Without the data, it's a crap shoot.

"So since the blades are going to be more flexible than wood, they need to be moved out from the tower."

A better solution is to go to a downwind configuration.

"The springs must keep the blades at the highest TSR possible so it would start producing power in the low winds without the servo using any power. Otherwise you would be using more power than you are making to power the servo motor."

I'm not sure springs are a good idea. The servo motor would have to fight against these springs, increasing it's power consumption. I assume you will have positional feedback from the servo system so setting to a start-up pitch is easy. Also, you will have to go past the start-up pitch setting to allow feathering of the blades for shut down.

"You could have either a backup rudundant brake or another pushrod running up through the center of the tower yaw axis to manually pitch the blades to zero."

Another backup system could be to use controlled yaw and simply turn it out of the wind. I will use this in addition to the electric brake.

[scottsAI]

Hello Murlin,

Keep in mind a well designed system whether mechanical or electro-mechanical will be as reliable. A mechanical furling system has failed many here. The more complicated the more likely a component will fail, well maybe.

By design complexity can make it more reliable, for example planetary gears. Two to three times as many gears, yet is more reliable than the two gears it replaces. How? Two ways: 1.Spreading the forces over a larger area (more contact points) reduces the local stresses. 2. balanced forces dramatically reducing stress on the bearings, resulting in remarkable improvement in system reliability.

Automotive uses what is cheap, light, and reliable. Planetary gears are used in the transmission.

Just how far do you carry fail safe? Any one component failure can take down a system. Complexity rises dramatically when dealing with fail safe.

For a 3 bladed hub bearing support, no stronger shape than the triangle. I to like two bearing supports.

When first looking at the pictures I thought you were using gears on the end of the blade shafts driven by the lead screw (third picture). Upon further study I see a lever? Any reason your not putting a gear on the blade shaft and turning directly? The force on the lead screw is the same for both, the gear is much simpler. Of course more complex to make the gear, should be able to buy. If I could think of a way to use a planetary gear I would. So far a worm gear comes to mind. Concerned about gear slop.

My 10kw design

The pitch controller is on the hub, with the drive motor. Using induction motor/gen with stall and RPM limiting control. Small PM gen to supply power to pitch controller. No wires connect. Lead screw motor will have planetary gear reduction. Wireless will report blade angle, RPM, temperature anything else I can think of. Using ISM band. Will have wireless brake control, brake will hold with no power.

The blade pitch will rotate near 90 deg, will be the furling system by turning the blades directly into the wind to stop the blades. Fail safe is the brake, considering making it a completely separate system.

Love Hugh Piggots designs. Has a simple elegance that works. Planed to make one. Then came along a request for a 10kw design. Commercial systems cost too much, they were willing to build it.

Realized a huge problem with PM gen with the output vs RPM issue. Controlling RPM will limit voltage. Power storage for a 10kw gen is costly so grid tie is required. Power electronics double system cost, so induction gen was the low cost obvious solution, which needs RPM control. Have a book for a stall controlled induction gen, with a special blade design. Pitch control sounds more seasonable.

Looking forward to your design, even comments:-)

Have fun,

Scott.

[Murlin]

This whole idea was just an eppifiny I had driving in my truck a couple days ago, and I hadnt put that much thought into it.

As far as being able to create the loft for the blades I think that PDF on resin transfer molding had just about all the information to do it.

I copied it to my desktop.

It showed what they did for the spar.

They just used fiberglass lamanents for the spars.  Stacked and staggered, top and bottom them and stapled them on the core and wrapped the core in glass... the laminates were bout a quarter inch thick at the root, and ending up with just one layer at the very end of the blade

They did stress and flex tests... man they did all the tests...

Do a search and find the PDF and have a look see and tell me what you think.

"I don't really see how this is going to work. So, your push rod is not really a push rod but a rotating lead screw. The ball nut would have to be carried in a bearing at the pitch collective to prevent rotation with the rotor hub. But even so, bearing friction would still try to rotate the nut on the lead screw, changing the pitch."

This was the first thing I came up with to keep servo from overworking.  look at how easy your jack screws work under extreme pressure and load.  I have a power lever longer than most strokes of any hotrod and so the force on the servo will be very small.

The springs would not be a huge acting force against the servo.

They would just employ enough power swing in favor of pitching the blades to zero pitch in a power failure.  They dont have to snap the blades instantly back in place.  Heck the wind might just push the blades back without any help from springs at all.

The nut would travel up and down on some guides in the nose cone to keep it from spinning, and to keep the stress off the hinge points.  I would use a nose cone to keep off the weather and to add structural support so lighter guage materials could be used.

Anyways it is a first draft and so must have all the bugs worked out.

I am convinced though that computer control is going to be just as good a way as any to control your system..

Heck, the market value alone would be priceless....lol

[Murlin]

Howdy scott I thank you for your comments.

"When first looking at the pictures I thought you were using gears on the end of the blade shafts driven by the lead screw (third picture). Upon further study I see a lever? Any reason your not putting a gear on the blade shaft and turning directly? The force on the lead screw is the same for both, the gear is much simpler. Of course more complex to make the gear, should be able to buy. If I could think of a way to use a planetary gear I would. So far a worm gear comes to mind. Concerned about gear slop."

Slop was my first concern.  Although backlash compensation offset will be included in the operating system.

Yes actually the 3 gear design was the first thought I had.

I thought about the slop, and I thought about the amount of force required to spin the gears and the surface friction of a bushing vs the cost of three large bearing/gears, thought about the building process ect..

About the best I could come up with was the "lever"...I know I cheesed out on that one, but it would allow for a single servo motor to operate pitch mechinism quiet easily and at the same time keep the design in reach of the DIY'er as far as materials and building expertise were concerned.  I was gooing to use a ball lead screw/nut but you could just use all thread and a nut the same way.  a little more force on the screw.

These wind machines are totally new to me. So It is my first attempt at the challenge.

Sure sounds like you know what you are doing though...

I wish I could say the same lol.....

I will post some more pics when I get the design refined as far as material sizes and such..

Murlin

[scottsAI]

Hello Murlin,

I checked your photos, nice shop! Nice work.

I am an electrical eng. strong mechanical background.

I like simple designs, get down to the fundamentals to make it work.

Keep It Simple Scott.

Missing from the diagrams, the lever must travel in an arc.

The simple mechanism shown does not support this arc movement.

The rotation of lead screw to deg change in pitch is now non linear.

OK, if I use slop to replace backlash? Covers wider group of precision issues.

Simple mechanism is a slot, which will have lots of slop, more than a gear.

Or, a second joint can work (much better) less slop than a slot.

Each joint will have slop... A gear looking any better? :-)

Cost for the worm gear and screw were a concern, so went looking:

http://www.atmsite.org/contrib/JSAPP/wormgear/wormgear.html

I have no idea how good the gear, but looks like DYI could do it assuming they have a lathe.

I like to put the lead screw motor upwind of the hub. Hollow tube for the drive shaft may be ok but it must go through the generator or whatever is at the other end.

My Pitch control is based only on RPM, so not reason to put it anywhere else. Will need a small PM gen to gets it's power, which will be reporting RPM (coils on hub, fixed magnets on base). Wireless to report conditions. I will now go talk to your pitch control diary.

Have fun,

Scott.

[Murlin]

"Missing from the diagrams, the lever must travel in an arc.

The simple mechanism shown does not support this arc movement.

The rotation of lead screw to deg change in pitch is now non linear."

There is alot missing from those initial sketches/models.  This was only the concept design and not the final working model.

The nut on the top of the grey tri push/arm is missing and travels in roller guids to keep from spinning.

As you and Amanda pointed out, there is a break point missing at the end of that same grey tri push/pull arm.

Also, the arms need to be standing up and not laying down for strength.

The materials are all too beefy and not to scale.

The return spring system is missing also.

I will perfect the actual structural design later, as to what sizes and guages the materials need to be.

As far as slop is concerned,  There will be hardly any slop in those joints.

The pins will be 65 RC and the actual plates that do the work will be hardened shimstock.

I am hoping that the springs will take keep the slop pushed to one side just enough so keep the blades from fluttering and would pitch the blades to zero should the nut and screw strip out(something that Will happen eventually unless you use a ball bearing nut and screw.)

"I like to put the lead screw motor upwind of the hub. Hollow tube for the drive shaft may be ok but it must go through the generator or whatever is at the other end."

Yes it will, but I cant see any problems with that at the moment since I am doing an Axiel Flux design.

Sure you Have to pull the servo off to take the alternator assembly off but that's ok with me.

For this first design I am going to use simple materials that will do the job.

IE  Stainless all thread for the screw and a two pieces of tapped brass for the twin nut design that will allow adjustmentto take out some of the mechanical slop that will develope in the screw/nut.

Brass bushings in the prop arms.

And as I mentioned, the movment levers to be very hard tool steel.

All of the parts will be cut on the waterjet, except for the hinge points, they will need to be precision reamed holes.

There will however be a couple of items that one will have to take to someone with a lathe and do but it will be simple stuff, and won't cost very much.

The goal on this design is for the average DYI'er to be able to build the machine easily with simple tools and minimal machine shop work.

The starting base design can be improved upon by using ball bearing control and hinge points, but it will make the machine more costly to build.

It will be up to the builder how much money they want to spend.

Murlin

[Murlin]

I am leaning more towards steppers motors rather than servo's to keep the price down.  They are a tad noiser but they cost about a 3rd as much to the same job, sometimes less.

There are some pretty inexpensive devices here:

http://www.weedtech.com

Everything you really need including the software.

That is all I have come up with so far.

Another good link here:

http://www.excitron.com

If anyone has any info or prefrences please post any links.

Thanks

Murlin

[Murlin]

I am rethinking having the servo motor at the back by the tail.  I don't know what I was thinkng....

Hrmmm...

Ruh Row....E/O change...

I would have to make too complicated a driveshaft to let the servo motor rotate.

I will prbably just pull on the rod like I had origonally planned and put the nut in the back.

Murlin