Crack in Rotor Hawk Hub Bar

[Jim Edwards]

Originally Posted by kolibri282

Jim,
just the other day I came across the pitch block design and immediately thought that this must be the least reliable design of all. I'd prefer a design with straight straps any time since you have an undisturbed force flow.

Juergen, I'm hoping I'll be able to find another system with a straight bar, but the bar I'm now using has milled-in coning angle.
Jim

[Jim Edwards]

Originally Posted by Redbaron

Thanks for sharing, jim. I have the same hub bar as you. Its hard to tell from the pics if you have any pre cone built in but I would guess yes since mine has a couple degrees. Will definatelly pay more attention to the hub bar on pre flights.

Jeff, as I said to Juergen, there is built-in pre cone. The crack progressed through the bolt hole and part way through the adjacent blade-side edge. There was not much metal left to hold that bolt! I'll try and get some pictures taken soon so everyone interested can see details better.
Jim

[dragonflyerthom]

Hope this doesn't slow you down Jim

[WHY]

I find it interesting that the crack starts at the trailing edge rather than the leading edge ? Which brings into question, where is the stress coming from and is a "frequency" involved ??



Tony

[karlbamforth]

Its interesting to see that the crack is between the blocks.

If the blocks were torqued down I think the bending load would have been spread over the full surface area of the blocks. This would mean that the peak bending load would have been at the edge of where the block sits.

That is where I would expect it to have cracked.

For it to crack between the blocks it must have been flexing at the point where the vertical bolts fit.

Can you examine, take pictures from above and below with the blocks removed.

If the area is dark grey to black, this will indicate fretting (movement) due to not enough torque.

Another possible cause is swarf or other debris trapped between the blocks, preventing proper torque and creating a high load point for the crack to start from.

[Passin' Thru]

Karl, I agree with your analysis. The joint was getting whip-sawed by the 2-rev pulses and the blocks were not secure enough to share the loads. I'd bet you find signs of fretting between the blocks and bar, and some hole elongation in the hub bar.
Jim, Please let us know what you find.

[Redbaron]

I agree with Pete. Pictures would be great. Speaking of whip sawed joints, curious of what torque the most highly loaded bolts in a gyro were torqued to.

QUOTE=Passin' Thru;472231]Karl, I agree with your analysis. The joint was getting whip-sawed by the 2-rev pulses and the blocks were not secure enough to share the loads. I'd bet you find signs of fretting between the blocks and bar, and some hole elongation in the hub bar.
Jim, Please let us know what you find.[/QUOTE]

[kolibri282]

If the blocks were torqued down I think the bending load would have been spread over the full surface area of the blocks

One problem I see is the very short base length of the blocks (d in my drawing) which could not be remedied no matter how hard you torque the bolts. Since the blocks are elastic a considerable bending load is transmitted to where the bar is weakest, namely at the point where the two bolt holes are. A quick and dirty analysis also reveals that the design results in a considerable stress amplification. Assuming that the resultant blade force acts at about 1/3*R (one third the radius of the blade, this assumption is a bit by guess and by golly, but probably not too bad) and the distance between the horizontal bolts is twice the blade thickness the bending force at the blocks is about ten times the resultant bending force at the blade. From an engineering point of view this is just lousy design which might be improved on by a fully fledged Finite Element Analysis but until someone comes up with an optimized design I would like to propose to switch to a design where the blades are attached via straight straps.

[WHY]

Juergen

An excellent analysis, and a likewise excellent recommendation for the "cure" . If this is indeed a very poorly designed application, I think we in the gyro community have been MOST FORTUNATE to not have had a FATALITY from this potential major problem.

This should be talked up all over the forum and web and especially at Mentone this year, we have had our warning, LET'S NOT IGNORE IT !!!

Tony

[GyroCFI]

Originally Posted by kolibri282

One problem I see is the very short base length of the blocks (d in my drawing) which could not be remedied no matter how hard you torque the bolts. Since the blocks are elastic a considerable bending load is transmitted to where the bar is weakest, namely at the point where the two bolt holes are. A quick and dirty analysis also reveals that the design results in a considerable stress amplification. Assuming that the resultant blade force acts at about 1/3*R (one third the radius of the blade, this assumption is a bit by guess and by golly, but probably not too bad) and the distance between the horizontal bolts is twice the blade thickness the bending force at the blocks is about ten times the resultant bending force at the blade. From an engineering point of view this is just lousy design which might be improved on by a fully fledged Finite Element Analysis but until someone comes up with an optimized design I would like to propose to switch to a design where the blades are attached via straight straps.

I was just going to say that...lol

[Redbaron]

Originally Posted by kolibri282

One problem I see is the very short base length of the blocks (d in my drawing) which could not be remedied no matter how hard you torque the bolts. Since the blocks are elastic a considerable bending load is transmitted to where the bar is weakest, namely at the point where the two bolt holes are. A quick and dirty analysis also reveals that the design results in a considerable stress amplification. Assuming that the resultant blade force acts at about 1/3*R (one third the radius of the blade, this assumption is a bit by guess and by golly, but probably not too bad) and the distance between the horizontal bolts is twice the blade thickness the bending force at the blocks is about ten times the resultant bending force at the blade. From an engineering point of view this is just lousy design which might be improved on by a fully fledged Finite Element Analysis but until someone comes up with an optimized design I would like to propose to switch to a design where the blades are attached via straight straps.

Hey jeurgen my blades are lightly loaded do you think I should switch to new rotors and a new hub?

[Jim Edwards]

Originally Posted by karlbamforth

Its interesting to see that the crack is between the blocks.

If the blocks were torqued down I think the bending load would have been spread over the full surface area of the blocks. This would mean that the peak bending load would have been at the edge of where the block sits.

That is where I would expect it to have cracked.

For it to crack between the blocks it must have been flexing at the point where the vertical bolts fit.

Can you examine, take pictures from above and below with the blocks removed.

If the area is dark grey to black, this will indicate fretting (movement) due to not enough torque.

Another possible cause is swarf or other debris trapped between the blocks, preventing proper torque and creating a high load point for the crack to start from.

Took some pictures for everyone today, and assembled the blades with new parts. I now need to string, balance, and adjust pitch as needed, while I look to the wisdom of this forum to determine a proper future direction to pursue (new Dragon Wings? - I always wanted some of those, a straight hub bar?, etc).

http://www.flickr.com/photos/jimedwa...7629901873917/

Sorry, but I haven't figured out to embed photos from Flicker onto the Forum.

[Passin' Thru]

Originally Posted by Redbaron

Hey jeurgen my blades are lightly loaded do you think I should switch to new rotors and a new hub?

Jeff, I wouldn't lose any sleep over it! Just inspect as you would any critical componant.
With all due respect to Jurgen's very academic analysis, it leaves out a few vital factors concerning loads, directions of force and distribution. Jeff, the greatest bending load that that joint ever sees is when the gyro is on the ground and the blades not turning. Except in a crash!
It's too late at night for me to go into all the details, so let's just "cut to the chase"; the proof is in the pudding.
This arrangement has been popular since the late 60's. It has been used on untold hundreds (maybe thousands?) of Brock / Bensen blades plus a considerable number of other makes. The cumulative flight hours must be in the hundreds of thousands of hours, some hellacious crashes and roll overs, and this is the first failure I have ever heard of in over 40 years of experience.
You can destroy any joint by improper assembly. If you permit in plane lead - lag motion in the joint it will lead to failure exactly as shown in Jim's photos.
When you disassemble the blades from hub, never remove the vertical bolts. They should be press fit and torqued to specified values. Remove the span-wise bolts and inspect each time they are removed.
Now, having said all that, There are probably better ways to make an adjustable pitch hub bar. Take a look at the way Sport Copter does it.
As for the rest, do it any damned way you please. It's no concern of mine.

[brett s]

With those new photos of it disassembled it's pretty clear everything was moving around - something was loose.

[Penguin]

One significant difference between the Sport Copter and the Dragon Wing hub bar is straight bar v. coned bar. What advantages has each approach?