back up, first feathering action

[jacquesm]

So, I'm back from my trip, and back in the shop !

First order of battle was to figure out what happened when the machine oversped and started to vibrate like crazy. It turned out that it had spit out one of the blade shaft support bearings.

The bearing was covered with snow, so we got another one, disassembled the governor (after testing if we could get the shaft to 'bind' by pushing at the blade tip, and we found that was not it).

We drilled a hole in the end of each shaft, tapped it for a 5/16" bolt to hold the bearing on with a lockwasher.

We also changed the weights so the bolts that hold the governor on the main shaft can be reomoved with the shafts & weights in place. This makes it much easier to work on the machine, just loosen up six bolts and the central support and you can pop the governor + blades right off.

After that we put it all back together again, but with a big change to the way the blades are mounted.

Originally the blades were mounted in the 'thickest' point of the blade profile. Now we have flipped the blade supports around, and the blades are supported about 1/3rd from the trailing edge of the profile at the root of the blade. This causes the air 'scooped' by the front of the blade to help the blade to try to feather.

We also painted the blades, one red, one white and one blue. No nationalistic signicificance there, we'd already red tipped one blade, so that one became red and the rest of the machine was white & blue. This hopefully will help us with analyzing balancing issues while looking at the machine from the ground.

So, after reassembling the whole works and roughly balancing it we put it back up.

We ran it unloaded for a bit to see if we could get the feathering to work or not and within 10 seconds of spooling up there was this 'clack' and the whole machine lurched forward !

I thought something had gone horribly wrong and quickly hit the kill switch.

After a minute or two of talking to Johannes about maybe having missed something we suddenly realised that nothing had gone wrong, the machine had feathered for the first time, just as we intended. The 'feed forward' that you get from those blades trying to scoop is tremendous though, and it causes the feathering sequence to go tremendously fast once started. This causes the whole machine to temporarily function like an airplane propellor driven by the momentum stored in the rotating mass, which causes it to screw itself into the wind.

It looks like we'll need to dampen that and tighten up the spring a bit to get it to work a bit more smoothly, but it does seems to work !

No pictures today, it's just way too dark...

[JW]

Sweet,

 Glad that your still at it. One thing that you might consider Jacguesm, is relocating your hinge bearings, in addition to adjusting spring pressure. The way I invision the counterweights, is that they are designed to "sweep" under cyntrpedal force that increases with rotor (shaft rpm) speed. To get ideal 1 to 1 forces that are highly liner and predictable, the hindge points(on the counterwieghts) need to be optimized for 90* (anglular) fron the shaft itself. Most designs of convinence will locate the hinge point at 45* beware of this posibility, as the load from the counterwieght will be non-linier or like 1 to 5 ratio and you'll play hell tring to match that. Are you using counterwieghts?

JW  

[JW]

Just read through again,

 I see you are using counterweights in your govener design. I cant recall your configuration with the hub. What im trying to get at, is that the counterwieghts have to be optimized for radial forces, since you are infact constructing an axial machine, the blades hingepoint is optimized for axial, and the counterweight should be optimized for radial forces. If the hingepoints are "hybidized" at all, you will encounter irregular load force variation at speed. Which would make it very difficult to turn the blades out of the wind at the desired speed, or with the consisency that your looking for.

JW

[jacquesm]

the hinge points are actually exactly at the right spot, this is done by offsetting the blades a bit on the governor base plate.

No 'compound' forces on the weights and a slightly compound force on the shaft that holds the blades is the result.

for a picture of the governor look at this image (very large!).

it shows the offset pretty clearly.

[richhagen]

I see the blades are holding up, I was a bit worried about how thin they were near the root.  When the blades feathered that had to put a tremendous strain on the blades as the machine slowed.  Once you dampen that it should ease that strain I would think.  Anyway, it would seem that your design passed a good, if not intentional stress test.  Keep having fun, Rich Hagen

[LEXX]

That's real pretty and all but, forgive me, what the hell is it supposed to do??  Looking at it is confusing me, please explain!

LEXX

[ghurd]

I'll give it a try.  I hope this is right.

The blades are mounted in the black squares with the holes, so they can piviot.

The gold parts rotate on their own little axle.

The small pins in gold are connected to the blades and control the angle of the piviot.

The large gold cylinders are held towards the center with the springs.

As the RPMs goes up, the weight of the gold cylinders is pushed outward, compressing the springs, and the blades piviot to less of an angle of attack.

So for low wind, the higher angle of attack lets it start working sooner.

For higher wind, the now lower angle of attack can make more power more efficiently.

Like a self controlled variable root angle. The best of both worlds, and everything in between.

How'd I do?

G-

[ghurd]

""The gold parts rotate on their own little axle.""

 Maybe that should read something like 'flop back and forth', instead of rotate.

G-

[JW]

Its not a complete drawing so I understand the confusion. As the assembly spins cytrifigal force causes the brass parts (counterweights) out from center, as this happens it establishes movement. This movement is transfered in to the rotor pivots. prefferably with tie-rods(not shown). actually the pins on the back of the counterweights should be on the side rather than the top. Anyhow I was just trying to give a representation of what I was describing to Jacques. In this representation three springs should be used to keep the counter weights together, only one spring is partially drawn there. This is different to how Jacques is using one spring in the center. I just wanted to provide an alternative design, for disscusion purposes.

JW  

[JW]

Thats about right Ghurd. I dont think this design is really any better than Jacques design, I just wanted to make a conversation piece out of it.

JW

[monte350c]

Nice drawing!

Came across this site last night (typical - was looking for something unrelated...)

Anyhow there's a nice photo of their brand of furling.

http://www.eoltec.com/English/pictures_eng.htm

Ted.

[dalibor]

this is preciously how i had been thinking about this problem. you just have to adjust with how much strenght springs will influence speed reducing, according to average wind speeds

[jacquesm]

Wow JW, that's a lot of work to go through ! I can't even get a decent line drawing done, and here is all out cad stuff. I'm impressed...

Now, about the design itself, I was tinkering with this for a long time before making the final design of the governor, and one thing we have found is that the more linkages you've got the more play you will have the more inaccurate and noisy the governor becomes. That is the reason why the blades are at an angle relative to the 'normal' line coming out of the hub, and why they share the same shaft with the weights. If you look carefully you'll see that the weights are exactly in the right spot, but the blades are a bit offset. This reduces the need for one more linkage, and a few thou of play.

I know that seems a tiny little bit,  but when you are looking at angular changes it doesn't take a whole lot of play in a linkage to get appreciable 'slop' in the shaft attached to the linkage.