Author Topic: Christmas Windmill Time (Read 28358 times)

kitestrings · « **Reply #33 on:** April 14, 2020, 08:31:56 PM »

Your blades look awesome. Congratulations.

Do be careful though...500-600 rpm on that step ladder with tent stakes; scary. Otherwise, all that hard work ends up "in broken pieces on the ground."

taylorp035 · « **Reply #34 on:** April 14, 2020, 10:26:31 PM »

Kitestrings - My 8 foot windmill ended up in pieces for that very reason. My brother and I were holding a 6 foot aluminum ladder in 2 feet of snow on a 30-40 mph day.... we tried turning it out of the wind since it was pushing harder than the 2 of us could hold it and the blade hit the front of the ladder, sending pieces flying about 100 yards in all directions. While spinning, it sounded like the blade on a turbo prop airplane at full tilt. Probably 1500+ rpm since it was completely unloaded and the blades were particularly flat/fast. Luckily everything exploded outwards when it hit and not towards our faces.

This most recent round of testing was done on a day with ~15 mph max winds and we were standing in front of it. I need to find a new way to build a tower and a way for it to pivot. I was highly concerned on this windmill since the treadmill motor had been heavily abused testing Briggs and Stratton engines for probably 8 years since this was last a windmill, so I didn't know how straight the shaft was. Also these new blades are at least 3 times heavier than anything I had carved in the past.

I'm waiting on some electrical parts to arrive and then I'll have a better setup on the charging side of the project with less alligator jumper wires. This week I'm testing how many times I can charge my phone from the 18 recycled Dewalt 18650 cells (~1700 mah) I have and so far it's looking like 8-12 times. Once everything is sorted, I want to buy a batch of 100 cells and make two 48 cell (2600 mah), 3s16p packs that can be easily be hot swapped for charging on the windmill and also reconfigured for the 48s1p pack for charging. They can also double as a portable battery for my 400/750W inverter. I'm currently using a 12v cigarette adapter for a 2A USB port. I might search for a 2 port one so my brother can charge his phone too. Also I want to find a quiet 75 or 150W inverter so I can plug my laptop in and take it off grid too that I currently leave on drawing ~5-10 watts 24/7 to charge my phone since my phone was made in the UK and didn't have a US wall adapter/charger.

taylorp035 · « **Reply #35 on:** April 17, 2020, 12:17:01 PM »

I decided to sand off the tips of the blades to see if that helps high speed performance. I also leveled the blades along the shaft axis. I'm going to wait to decrease the pitch angle on the blades until after I've done some more testing and watch how well it starts up and hits cut in voltage with the diode in the system since that will take another volt.

ElSuperFantasma · « **Reply #36 on:** April 17, 2020, 10:41:00 PM »

Those blade tips remind me of my favorite pillow.

Blades looking good.

TSG

taylorp035 · « **Reply #37 on:** April 22, 2020, 08:40:33 PM »

"It would make me feel better if you used two hub plates, on the front and back." - Sparweb, you win the best comment from the audience award today...

So last weekend, I took the windmill out again and left it up over night to take advantage of some 18 mph winds that were building up towards lunch time the next day since my 18650 cells were almost dead. At 6:43 AM, I decided to look outside at the windmill in the back yard and was trying to make out if the blades were still there and if they were spinning. Well, about 10 seconds later, I got to witness the whole thing explode and throw the blades in all directions. The front leg on the ladder took quite a beating from the one blade and the other two blades shot off into the woods in either direction. When the blades flew their 100 feet in either direction, they struck the trees and sounded like wooden baseball bats being thrown into a pile.

Upon closer investigation, I've deduced that the blades did not hit the ladder first, but rather the hub had ripped in half along the grain by centrifugal force. I had confirmed this by comparing the dent in the blade to where it hit the fiberglass ladder.... seeing that the blade had to be about 6-8" lower than where it normally would be. Impressively, all 3 blades were in really good shape afterwards and they are definitely re-usable. The shaft on the motor appears to be straight still, but I'll have to use a dial indicator at a later point. The ladder on the other hand didn't fair so well and had lost all of its structural integrity... so that's going to cost me about $200 for a new one. The batteries flew out of their holder and about 4 of them were driven into the mud 1". I figure it was spinning at about 1,000 rpm when she let go.

In the span of less than 6 hours, the windmill generated about 4000-5000 mah. So I must of had the wiring correct, so that's now validated.

So the lesson here is that it is a really good idea to use two hubs to sandwich the blades together to both lower the bending stress in the hub and to distribute the load. It's also likely a good idea to add some additional fasteners. I figure there was a few thousand pounds of centrifugal force on each blade, especially since they were at least 3X heavier than my previous 5, 6 and 8' blade sets. Also it would be a good idea not to use poplar as a hub.

Here is a video of it going relatively slowly in probably 8 mph winds.

And the carnage video afterwards.

ElSuperFantasma · « **Reply #38 on:** April 22, 2020, 10:02:50 PM »

That was like an accident waiting to happen in my opinion. Glad you or others did not get hurt in the process.

Stay safe and healthy.

TSG

SparWeb · « **Reply #39 on:** April 22, 2020, 11:37:50 PM »

I did not intend to, or mean for my comment to be so prescient.

It even took down your ladder in the process. Very sad to see it Taylor. Don't give up!

I could bug you about "orthotropic materials" but you should have heard enough about it already at Uni. - - - and were just looking for the opportunity to make it real to yourself.

taylorp035 · « **Reply #40 on:** April 23, 2020, 07:26:45 AM »

I will probably make a real tower and a better hub in a few weeks once I'm done with a few other projects. Since I'm not driving to work right now, I've torn apart my daily driver to fix some rust and I still have another two weeks worth of work on it to get it done. I'll also have received some more electrical components by then too so I can have a more proper charging setup. The few hours this windmill lasted was enough to prove to me that the concept is worth investing a few dollars into. It also proved to me my new balancing method works pretty good. This latest windmill only cost me about $3 in bolts and about 50-60 hours of labor.

kitestrings · « **Reply #41 on:** April 23, 2020, 04:52:04 PM »

I'm glad you shared this taylor. We all prefer to share the positive results, but often learn as much or more from these events.

Glad to see that no one was hurt, and that the blades are still in good shape. You can replace the ladder, right.

With a better hub plate(s) design, and better tower, you'll be in great shape. The blades look great. ~ks

SparWeb · « **Reply #42 on:** April 23, 2020, 11:10:11 PM »

And in anticipation of a possible future question (if you haven't already picked a finish to use that you prefer):
I get great results from Varathane Diamond Finish, about 5 coats, sanding between coats. It last several years.

frackers · « **Reply #43 on:** April 24, 2020, 12:50:34 AM »

Sounds like a good case for plywood (both sides) rather than straight timber to avoid the grain issue.

electrondady1 · « **Reply #44 on:** April 24, 2020, 10:23:09 AM »

tough luck Taylor, but it sounds like your shaking it off. if the blades came through unscathed your not out the time you invested so, stage two , good to go.

kitestrings · « **Reply #45 on:** April 24, 2020, 01:29:07 PM »

Quote from: frackers on April 24, 2020, 12:50:34 AM

Sounds like a good case for plywood (both sides) rather than straight timber to avoid the grain issue.

I've always thought that wood was a rather poor choice of materials here. I know it is common in many of the text-book designs (Hugh's, the Dans') but seems like a sandwich using two light-weight pieces of sheet metal makes a better Dagwood. ~ks

kitestrings · « **Reply #46 on:** April 24, 2020, 01:34:19 PM »

I wonder too if having only two fasteners in the same line along the blade root might have played a role. It looks to have split along the line that had parallel grain in the hub material. There's got to be some torsional component at high speeds, I would think.

ElSuperFantasma · « **Reply #47 on:** April 24, 2020, 02:06:58 PM »

I’ve seen people using two old circular saw blades to make a good solid blade hub. Get a grinder, cut the circular blades to the shape of the root of your wooden blades, drill some holes and paint them so they won’t get rusty.

I use blade hubs out of old air conditioning fans and also stainless steel face plates for the front.

Happy Friday!

hiker · « **Reply #48 on:** April 24, 2020, 03:10:45 PM »

https://m.youtube.com/watch?v=1JOOGBqNd4c Vid I posted years ago,,,saw blade rotors,,plywood hub,,angle cut in hub not blades,,strong setup,,still cranks out the power

SparWeb · « **Reply #49 on:** April 25, 2020, 11:56:14 AM »

My first WT hub was a pair of 12" sprockets. Somewhat like Fantasma suggested.

taylorp035 · « **Reply #50 on:** April 25, 2020, 09:02:34 PM »

The two bolts in line were probably not a good idea. I'm a mechanical engineer, so I'll make sure I do it better next time. I'm planning on upgrading to some M10 or 3/8" bolts... should be able to get 10-20k lbf of clamp force instead of my cheap 1/4-20's I had in there before. I might do more bolts this time per blade as well. I will also need something other than small washers to distribute the load into the wood. I'll also make it out of some stronger wood, like maple or maybe get some plywood. I also will try to make the hub an inch bigger in diameter.

It also badly needed a steel hub adapter to the motor shaft. It's was a little loose when I took it out the second time. This could of been another contributor for why it would of broken.

Centripetal force = mass*V^2/r. omega = v/r, so C_f = mass*r*omega^2. Lets say the blades weigh 1 kg. Center of mass is 40% down the blade, so r= 0.426m. TSR of 7 at 20 mph on 42" = 560 RPM.... lets say unloaded that gets you to 2x that (guess on my part... the blades have a rather thick cross section), so works out to 117.333 rad/second or 1120 RPM (280 mph tip speed).

C_f = 1 * 0.426 * 117.333 ^ 2 = 5864.7 N = 1318 lbf per blade at 20 mph unloaded.

If I want the thing to survive 40 mph, the force goes up by 4, so 23459 N or 5,274 lbf per blade @ 2240 RPM.

Since I don't want the bolts in shear, I need the bolted joint to work within its friction envelope. 5,274 lbf / 0.15 friction coefficient = 35,160 lbf is what I need from the bolts. My experience at work says I'm going to need 3 or 4 10.9 M10's torqued to about 50-60 ft*lbs to achieve that.

On the one that exploded, the bending internal to the hub from the blade would of been 1318 lbf * ~1.75" offset from the center of mass of the blade to the center of the hub = ~ 20 ft*lbs of torque... probably was a sizeable contribution to overall stress in the hub. And since the blades were mounted to the front of the hub, this bending stress was added to force of the blades as they were being pushed back (not sure how much force that would of been).

I guess all the math suggests the initial design was going to break sooner or later.

kitestrings · « **Reply #51 on:** April 26, 2020, 07:57:40 PM »

I might be I'm missing what you're thinking here taylor, but if you are torquing either thru-bolts with washers against the wood blades, or sandwiching the blades between two plates, it seems like 50-60#' on the fasteners is going to be way too high. I get the concept of the friction fit, but I think the limiting value here is going to the allowable compression of the wood perpendicular to the grain. I don't think you need to be anywhere near that limit, mind you, to get the desired results.

On a different topic, I'm curious if you noticed any discernable difference in blade noise with the rounded blade tips?

~ks

SparWeb · « **Reply #52 on:** April 26, 2020, 10:47:22 PM »

Just gonna go nuts on an element of engineering here:

Quote

...Since I don't want the bolts in shear, I need the bolted joint to work within its friction envelope. 5,274 lbf / 0.15 friction coefficient = 35,160 lbf is what I need from the bolts. My experience at work says I'm going to need 3 or 4 10.9 M10's torqued to about 50-60 ft*lbs to achieve that.

Exactly backwards.
Never design bolted joints to rely on friction. Never. Never ever.
I'm shocked that this misconception could be taught in university today.

You do account for friction on various ways in bolted joint design.
All correctly designed bolted joints will work if friction is zero.
They do rely on strength of the fastener through the shear plane at the interfaces of joined members, and bearing of the members against the bolt around the shank.

I know, I know. That was pretty pedantic, but I've seen you post engineering stuff before and it's never been so far off the mark as that.

MattM · « **Reply #53 on:** April 26, 2020, 11:53:45 PM »

Why not tap a couple of wood screws that utilize floating cap washers? It's still a good idea to use a plate front and back.

I experimented with up to ten foot blades on 12" wide 3/4" plywood rotors without a single one breaking like that. On a side note there were plenty of bent blades, just never had any launch even under some terrible storms. I should mention there also was a 26 or 24 gauge galvanized steel flat opposite the blades and an even distribution of capped wood screws in from both sides. Steel loves the constant tension, whereas it fatigues when alternating tension and unloading it. I suggest save the steel rotor for when you're ready for the true production.

taylorp035 · « **Reply #54 on:** April 27, 2020, 12:36:55 PM »

My words must of been interpreted differently than I thought. I know the bolts aren't there to be used in shear through the shank of the bolt. On the windmill that blew up, I can guarantee the joint didn't have enough clamp load and thus the bolts were probably in shear right before she let go. I think the added shear load added stress to the holes in the hub and probably was the initiation site for the hub the break apart.

My university didn't teach me about bolts. And even though I resented that while I was in school since I was more of a hands-on type of guy, about 3-4 years later, I appreciated more that they spent their time teaching other things since there are a million real world applications and only a finite amount of classroom time. Now I get to do this full time as my job designing things for Cummins 15L engines, which in a lot of ways is a bunch of bolted joint problems. A grade 10.9 M10x1.25 is good to about 100 ft*lbs before it breaks and about 17-18k lbf of tension. 55 ft*lbs and about 10k lbf will be a good value to aim for. It's interesting to look at these things when you have a fancy ultra-sonic bolt stretch gauge. I was probably lucky to have 1 k lbf on the low grade 1/4-20's.

For over-compressing the wood and then loosing my clamp load, I know I need some type of metal plates to distribute the load better than my 3/4" dia. washers. At which point I should just go out and buy myself a 1/4" thick sheet of steel and cut two 10" circles.

taylorp035 · « **Reply #55 on:** September 04, 2020, 10:14:22 PM »

It's been a slow recovery, but I finally bought the hardware for the new hub to get this windmill back up and flying. I decided to go big with (2) 10" diameter, 1/2" thick steel plates. I'm also going to have lots of clamping force with a total of (9) grade 10.9 M12 x 100mm flange head bolts and associated nuts. I was working on a project for work this week that had these and they were measured at 23k lbf of load each. I'm probably not going to go that tight, but it's nice to know I have the ability if I want to. This should be a 20-40x improvement in clamping force over the last (6) 1/4"-20 bolts.

I think I mentioned earlier that I measured the pitch on the tips of the blades and all made them equal and a bit faster. It started up so well before that I figured I could go for a bit more speed. I think math says ~20% faster, but we shall see. It went from about 5.5 degrees to about 4.5 degrees IIRC. I can quickly add a half degree back if needed. It should also gain 1/2" in diameter due to the slightly larger hub. I tried welding my own hub a week ago out of a old chain sprocket I had been eyeing up for months, but my 30A stick welder wasn't up to the task of getting enough heat into both pieces of steel.

I also bought the batteries and finished off the wiring of the electronics. Should be good for 20A charging at 4.2 volts and 40A continuous/75A peak @ 12v discharging. The batteries are (39) 3200 mah 10A 18650 cells and I have them all balanced and ready to be wired up in their 13 cell parallel packs. They will all fit into a plastic shell from my old LiFe motorcycle battery for better safety and ease of moving around. It should be a nice pack for running my 400w inverter, particularly when I want to tune my car for a few hours on the road. If the system works well and I want more, I can go and buy more recycled 18650 cells for cheap.

SparWeb · « **Reply #56 on:** September 05, 2020, 09:45:05 PM »

Hi Taylor,
Good to see you back.

I'd love to try out your ultrasonic bolt gauge!

You might be overdoing it on the hub using 1/2" plates. Since you have 'em, okay, but my 10' hub is two 3/16" plates and probably over-cooked at that. Too bad your stick welder is giving you a fuss. If you have an acetylene torch, you can pre-heat the parts before turning the stick on them. Just be sure that whatever you use to hold the parts together during the heating is (a) not flammable and (b) keeps them straight while the heat is applied.

taylorp035 · « **Reply #57 on:** November 17, 2020, 09:34:59 PM »

Lots of hardware has been made and painted since my last update!

Don't try designing a windmill like this.... None of my windmills have ever had a furling mechanism because I've never put that much effort into a windmill. I had one last for 18 months without any issues and we usually get 50-60 mph winds when the lake effect snow winds happen in the winter. This time around I decided to make the hubs strong enough to withstand any imaginable wind speed with the blades un-loaded. This windmill will also be far away from any people or houses if it does let go.

I purchased a bunch of hardware to make this easy for me and to limit the amount of machining I had to do. Everyone has their own skills and equipment and welding or having access to a plasma cutter isn't one of mine. All of this was done with a drill press, a ruler and a protractor. I ordered 10" diameter, 1/2" thick 1020 steel off McMaster and it came in at about 0.6"... which made things rather heavy. I then took it over to my friend's house where we just put together a homemade sand-blasting cabinet. I cleaned the rotors up real good and then primed and painted about 15 layers to make it extra pretty.

Next I drove 2 hours to another friend's house to go use his more capable lathe to make a 15/16" section for the hub on the end of the ~26" long x 1" high lead steel shaft (~60 ksi yield). I also ordered some new pillow block bearings and a full set of grade 10.9 M12 x 100mm bolts and nuts.

The new rotors are about 1" larger in diameter, so hopefully this offsets the slight decrease in pitch on the blades when I milled all the roots to the same thickness. The blades are now about 1/2 to 1 degree flatter and much more consistent between blades. I think the tips are about 4.5-5 degrees for the center line of the airfoil.

kitestrings · « **Reply #58 on:** November 18, 2020, 10:30:40 AM »

That looks pretty robust. Still, I would consider some sort of overspeed protection... furling, electric braking or even manual shut-down. Otherwise, it is only a matter of time, if it is up in the air anyway, before something fails or burns out. You've put a lot of work into it. Why risk that?

Hopefully you have a more substantial tower to experiment on? Best ~ks

bigrockcandymountain · « **Reply #59 on:** November 18, 2020, 06:16:19 PM »

Looks good. What holds the hub on to the shaft?

SparWeb · « **Reply #60 on:** November 18, 2020, 11:09:55 PM »

Nice blades.
I look forward to seeing how you mount this (and I'll add my voice to the chorus: these are worth protecting)

taylorp035 · « **Reply #61 on:** November 19, 2020, 11:57:42 AM »

Quote from: bigrockcandymountain on November 18, 2020, 06:16:19 PM

Looks good. What holds the hub on to the shaft?

There is a 1/4" dowel pin. I'll have a hose clamp or shaft clamp on the front once I'm done to make sure the hub doesn't slide off the end of the shaft. The shaft is stepped, so the hub can't slide backwards.

SparWeb · « **Reply #62 on:** November 20, 2020, 02:06:53 AM »

Consider one of these:

taylorp035 · « **Reply #63 on:** November 20, 2020, 12:10:37 PM »

This morning I decided to bench test the motor with my full electrical system after straightening the mounting bracket of the motor after the hub explosion.

What I found was my 10 amp 1S BMS unit while wrapped in electrical tape was really only good for ~3 amps continuous before blowing smoke out of it. I've since taken the tape off of it for better cooling. After testing the motor, I think I'll want at least 3 of these in parallel (I already had 2 more on order) and probably more like 5.

Here are the results with and without my diode in-line. I had a broken connection to the BMS board after I un-wrapped it, so I'm not sure if this data is any good. I'll rerun this later once I fix it.

bigrockcandymountain · « **Reply #64 on:** November 20, 2020, 02:33:54 PM »

X2 on what sparweb has pictured.

Ill let others do the electrical math. Something seems a bit out to me, but I don't have the brain power for it just right now.

MagnetJuice · « **Reply #65 on:** November 20, 2020, 02:57:34 PM »

X3 on the suggestion by SparWeb.

The aerodynamic loads due to centrifugal, gravitational and gyroscopic forces on that blade/hub assembly are going to be high.

That round dowel is not going to hold that blades/hub assembly in place.

The last thing you need is for that rotor to come loose while spinning at close to 1000 RPM. It could take of like a helicopter and land a few thousand feet away on top of somebody's house.

Find a locking hub similar to the one SparWeb suggested and make sure that you use a square key to go into that square keyway.

Nice work so far. Safety first.

Ed

News:

Author Topic: Christmas Windmill Time (Read 28358 times)