Devil is in the Details (Lean of Peak Aviation)

Investigating and proving what caused a major control system failure in my Sonex Onex.

[JoystickTX]

I use a rod between my tailwheel horn and rudder horn on my model planes, but it does transfer shock loads from the tailwheel to the rudder. That can cause the rudder horn to come loose or it can damage the rudder hinges. I have noticed more damage than the springs that I used before.

On GA planes, springs would be way better and much safer.

I'm glad you found the root cause before you lost rudder control.

People need to really be careful when they change the design, even when they think it's a minor change.

[milehigher]

Our Sonex, has the "upgraded" link with the ball ends, this was chosen for the vary reason that has been mentioned by others. The standard link does not freely allow vertical movement from the flex/spring in the tail wheel shaft. This produces unwanted flex and wear on the cranks/attachment points that could lead to stress failure and free-play in the assembly. The ball end link moves freely with the vertical movement of the tail wheel while providing a slop free connection. If the "upgraded" link has the exact same measurements eye to eye and the geometry matches the standard design as well as being attached as per the plans it allows full and free moment of the control surface. The addition of a washer larger than the hole ensures if a ball ever failed the control rod would stay attached but would now have some freeplay but would not be a catastrophic failure. The argument that the original design has been in use for 30 years without any known failures is valid but it does not mean that it cannot be improved if done correctly.

[SoloRenegade]

awesome you took the time to test the old and new parts to show the differences.

[joseruiz-Joecool]

I am glad you were able to find this mistake on this airplane when you did. Looking foreword to the flying videos in the future.

[davidmangold]

Wow, it’s great you caught that! I’d look for other non-standard parts that the builder may have used. My friend just bought a Velocity, that a great builder built. But, one has to become intimately familiar with what the builder built!

[excellenceinanimation]

Wow it looks so similar! Scary how such a small change can have such a huge effect!

[GlensHangar]

As someone who spent a year tearing down and rebuilding a 1961 Cessna 172B 'certified' aircraft... Don't let 'certified' / 'passed annual' give you all of the good feels, while putting down homebuilders as corner cutters.
We found 60 years worth of non-standard parts and repairs in the 'certified' airplane.

[bobjoatmon]

Having helped build three Experimental aircraft I found that sticking to the plans is hard for the personality of the average builder BUT the plane should be built to spec because the builder probably doesn't know all the engineering decisions that went into the design. The NTSB has many accident reports where an experimental aircraft design was as good as a certificated airplane but poor workmanship or minor changes introduced by the builder lead to crashes and/or fatalities. I walked away from helping one builder because he just cut and threw the fiberglass on any old way (to economize on material evedently) instead of orienting the strands properly for the loads. Luckily he ended up abandoning the project (a divorce during build) and the plane was scrapped.

[crikitaviation]

Builders sometimes forget the saying "Safety in Numbers". Anytime you modify your airplane and deviate from what everyone else is doing, you put your name at the top of the list for problems that may occur because of your modification. Always better for your name to be somewhere down the problem list and hope that someone else will experience a problem or failure before you do. Think about that every time you make a change no matter how small. So glad you caught this. WOW.

[sganzerlag]

Ok so what happens when the tailwheel fork flexes due to landing loads and uneven rolling surface? How does that flex affect the fixed tailwheel control pushrod? I don't see how the new pushrod can account for the fork flex during tailwheel suspension action.

[salamander]

If the forward bolt position (measured at 3:19) is off laterally, maybe it is also off fore/aft. There is no adjustment in the factory link, so if the hole is off the tailwheel may not track straight. Which is maybe why the previous owner made his own link with Heim joints for steering adjustment?

[FlyMIfYouGotM]

Aviation is a very unforgiving environment that won't tolerate carelessness for long. Attention to the smallest details is a must.

[mylesvrooman]

If you don't wish to replace the whole Tailwheel castor, you could use a larger diameter bolt to attach the control arm, or you could use a bushing to tighten the hole up to your specs.

[Wolves]

Hello, nice videos, the shimmy on the tail wheel is not coming from the play on the bushing, is coming from the caster angle-

[danielbasovitch]

In other words, there now are NO Swivel Balls in the connections to allow for vertical rotation of the control arm!

[yetiatlarge]

What grade of connecting hardware did you use are you using Teflon impregnated nuts is there any spec for a sleeve where where that undersized bolt is moving around

[DennisZylstra]

GREAT VID. Loved the testing. Was wondering if you should redesign the way the tailwheel hooks up to the rudder. I like a spring on the RT and LT side to eliminate any vibration or hooping pressure on the rudders control arm produced by the wheel hitting things. Seems to me a very bad design. Keep up the good work!

[louisvanrijn]

1.
2:32.. Play is unacceptable. OK.
Ream the hole in the plate, and turn a small sinter bronze bush that just fits. Play gone. Apply grease to keep moisture away. Idea?

2. 4:33
That assemby is designed to fail that way.

Peak forces on the tailwheel are bringing pounding loads in that rod. These loads under an angle cause the secondary bending moment which is designed to exist in this assembly.
If you apply an small angleplate between the lower bracket/rib and the skin of the rudder, the moments are coupled off. The lost rivets will not come off again.

The strength of the sheared-off rivets is overestimated, or the loads not identified.

Remember that the Avdells, pull-through-rivets or pop-rivets are hollow. These arent solid AD or DD rivets such, as used in in a B-17, Liberator or equivalent, and have only a fraction of the strength.

[peterblacklin]

The surfaces on the link are flat, how does the flexing of the tail wheel arm affect the fit?

[danielbasovitch]

With Just bolts attaching the horn to the wheel assy how does it swivel when a load is applied in a down force to the tail wheel arm?? The tabs are just flat surfaces.