RV-12 Fatal Crash - Control Failure 6/6 Auburn, WA

Another important lesson should be how essential it is to be properly trimmed at all times. A lot easier to control using power and rudder if the aircraft already wants to fly straight and level.

For example, there's a Piper SB concerning a potential weld failure in the T-bar assembly. That would leave you with no pitch (and most likely no roll) control from both sides. If you fly one, it's a good reason to practice power and rudder approaches on a regular basis, just in case. But, as seen in this accident, that training mindset should be present regardless of the airframe you're flying.
 
Another important lesson should be how essential it is to be properly trimmed at all times.
The trouble is, for the large number of light aircraft, this is impossible.

Most aircraft have pitch trim (mine doesn't), some aircraft have rudder trim, but trim controls for roll are relatively rare. And if you're flying solo in an aircraft where the seats are side-by-side, the airplane is going to want to roll to the left and you've got no way to trim that out.

Ron Wanttaja
 
Can’t tell you how many times I’ve said they really could have used that time better, but part of me thinks that people in this situation need or at least seek a little contact with the outside world.

A radio call was wise and necessary. He needed to clear the pattern and runway, both for his own safety and others. This was not at all a comfort call.

I just can't figure the Pan Pan. This was a Mayday if there ever was one.
 
you're flying solo in an aircraft where the seats are side-by-side, the airplane is going to want to roll to the left
And that was the accident mode in this case. Gradually increasing left roll.

The more I look at the photo, the more I think he attempted to use the other stick, but the controls were blocked from rolling right by the loose rod end.

I can't tell from the photo whether the horn is roughly centered in that photo. Perhaps someone with RV-12 experience has insight?

Some degree of roll can be countered with opposite rudder. We do it all the time when we slip. But how much, and for how long?
 
For example, there's a Piper SB concerning a potential weld failure in the T-bar assembly. That would leave you with no pitch (and most likely no roll) control from both sides. If you fly one, it's a good reason to practice power and rudder approaches on a regular basis,
If you fly one, it's a good reason to make sure that the airplane is getting a proper and thorough annual inspection. That's what annual inspections are for, after all. Mechanics on these airplanes should be aware of their weaknesses, and SBs point out those weaknesses. There is no reason whatever for stuff like this to fail unexpectedly.

FAR43 Appendix D lists the minimum inspection requirements for annual and 100-hour inspections. In it we find this:

(c) Each person performing an annual or 100-hour inspection shall inspect (where applicable) the following components of the cabin and cockpit group:

(5) Flight and engine controls - for improper installation and improper operation.

(7) All systems - for improper installation, poor general condition, apparent and obvious defects, and insecurity of attachment.


The inspection requirements for the empennage say the same thing. Inspect those controls for defects.

This requires opening up whatever inspection panels or removing interior furnishings to get a look at control systems. That's work, and time, and costs money, so it's often pencil-whipped instead.
 
If you fly one, it's a good reason to make sure that the airplane is getting a proper and thorough annual inspection. That's what annual inspections are for, after all. Mechanics on these airplanes should be aware of their weaknesses, and SBs point out those weaknesses. There is no reason whatever for stuff like this to fail unexpectedly.

FAR43 Appendix D lists the minimum inspection requirements for annual and 100-hour inspections. In it we find this:

(c) Each person performing an annual or 100-hour inspection shall inspect (where applicable) the following components of the cabin and cockpit group:

(5) Flight and engine controls - for improper installation and improper operation.

(7) All systems - for improper installation, poor general condition, apparent and obvious defects, and insecurity of attachment.


This requires opening up whatever inspection panels or removing interior furnishings to get a look at control systems. That's work, and time, and costs money, so it's often pencil-whipped instead.

This was EAB, so the builder can do the inspections. The same guy who made the error.

IMO this is a strong case for the use of torque seal as standard practice.
 
This was EAB, so the builder can do the inspections. The same guy who made the error.
No. This was a kit-built ELSA. Builder cannot perform the yearly condition inspections.

However, in this case, the builder did take the separate course that allows an ELSA owner (not necessarily the builder) to perform their own condition inspections.

Ron Wanttaja
 
No. This was a kit-built ELSA. Builder cannot perform the yearly condition inspections.

However, in this case, the builder did take the separate course that allows an ELSA owner (not necessarily the builder) to perform their own condition inspections.

Ron Wanttaja
Ok, thanks for the clarification.

In this case, were the builder, owner, inspector, and pilot all the same person?
 
No. This was a kit-built ELSA. Builder cannot perform the yearly condition inspections.

However, in this case, the builder did take the separate course that allows an ELSA owner (not necessarily the builder) to perform their own condition inspections.

Ron Wanttaja
I thought it wasn't amateur built.
 
Ok, thanks for the clarification.

In this case, were the builder, owner, inspector, and pilot all the same person?
The aircraft would have required a DAR for the initial certification inspection (which counted as the first Condition inspection). Beyond that, once he had received his Light Sport - Repairman certificate, the builder, owner, inspector, and pilot would all have been the same person.

I highlighted that one section, as we do not know when the builder received his LS-I certificate. The plane is listed as a 2021 completion; I have the August 2021 downloadable airman/mechanic database, and he's not listed in it. In any case, one thing that's necessary to understand: The course for the LS-I certificate is just 16 hours long, and has no experience requirements like those required for a full A&P certificate.

While I'm doubly-hesitant to speculate in this particular case, I believe it does well illustrate the differences of the Experimental Light Sport Aircraft vs. the traditional Experimental Amateur-Built category, and the shortcomings of the ELSA category.

The RV-12 is a good example. Van's sells the RV-12 kit, which can be built as an Experimental Light Sport Aircraft (ELSA), OR as Experimental Amateur-Built (EAB). The RV-12 qualifies under the "51% rule" for EAB, in that the builder must perform the majority of the tasks involved in constructing the aircraft.

In contrast, there IS no 51% rule for ELSAs. Companies are permitted to sell kits at whatever level of completion. Vans could sell an ELSA RV-12 kit that requires less construction time, if they wanted to.

Where they differ is when they roll out of the shop door into the hangar. The builder of an EAB RV-12 can make whatever changes he or she wants. Auto engine? Sure! Steam gauges instead of electronics? Yup. Change the ailerons to a cable-operated system? Certainly.

But if the builder wants to license it as Experimental Light Sport, ABSOLUTELY NO CHANGES are allowed. To get the ELSA certification of the kit, Van's had to build an example (an exemplar) and fully certify it in the Special Light Sport Aircraft category...the category that covers ready-to-fly light sports. When presented to the DAR for certification, the ELSA must *exactly* match the exemplar aircraft, except in cases where the company does allow leeway. You don't have to paint the airplane the same as the exemplar, for instance, and the company may allow variation in avionics. But the company does not have to grant that leeway, and if they license the exemplar with a Narco Mark 4 as a radio, they can demand all examples have the ol' coffee grinder on the panel.

The builder of the EAB RV-12 presents his aircraft to the DAR, who checks the paperwork, potentially winces at any changes, but eventually signs it off. The builder can be awarded the Repairman Certificate for *that* particular aircraft.

In contrast, when the ELSA RV-12 is presented to the DAR, the DAR is supposed to ensure that the plane exactly matches the exemplar.

In the real world, of course, this is impossible. The DAR is not going to be an expert in RV-12s. They may not detect changes.

From the NTSB Preliminary, it appears the aileron control system was modified from the original. If this had been done during construction and the DAR detected the change, they could have demanded it be restored the exemplar configuration.

But it was a subtle change; one would not expect the average DAR to detect it (though I bet DARs from now on are going to be looking....).

However, once the ELSA aircraft has its Experimental certificate, the match-the-exemplar system goes out the window. Once the certificate is signed off, the owner of an Experimental Light Sport can modify their aircraft to their heart's desire. The owner of an ELSA RV-12 can install an R-985, extend the wings to make the plane a motorglider, whatever. All that will happen is that the plane has to go back into its test period.

Was N914JN built with the allegedly modified aileron system? We don't know, yet. But if it was, it's something that should have been caught by the DAR...but I don't fault the DAR. Probably would have been a hard thing to notice. It's one of the drawbacks to the ELSA system; the aircraft are expected to match the exemplar, but without both aircraft sitting side by side, changes are difficult to detect.

However, once the aircraft had its certificate, the owner would have been free to make the modification, just like the owner of an Experimental Amateur-Built.

One of the points I like to hammer on my homebuilt accident talks is the rate of occurrence of accidents vs. aircraft total time. Most builders are focused on the first flight. Once that has been successfully completed, there's an attitude like "Well, I guess it's built properly."

Yet look at this plot of accident rate vs. aircraft total time:
1719592383575.png
Notice the peak at the 50-60 hour range. Some of this is related to the aircraft having completed its test period and starts running out of gas on cross-countries or scud running.

But some of that increase is due to mechanical failure. Just because a part survived the first flight DOESN'T mean it's perfect. Construction flaws could result in gradual deterioration. Parts might be wearing abnormally, and end up failing much earlier than they should.

Building an aircraft gives one knowledge on how the parts go together, but it doesn't teach you how they wear and change in operation.

Which brings us back to this accident. The alleged change in the aileron system obviously worked fine at first, but it appears it was gradually going bad.

Ron Wanttaja
 
I thought it wasn't amateur built.
Depends on your capitalization.

The airplane could be considered "amateur built," since it was constructed from a kit by an amateur.

What it WASN'T was an Experimental Amateur-Built airplane. It had to follow the Experimental Light Sport Aircraft rules, rather than those for Experimental Amateur-Built. See my post above.

Ron Wanttaja
 
Was the modification only that he installed the rod ends backwards? Sorry not the rod ends themselves but the attach bolts.
 
Last edited:
Was the modification only that he installed the rod ends backwards?
Yeah. My impression is this was an installation error (and one that could be easily overlooked) and not a purposeful modification. Now there's an SB to alert others to this issue. This specific kind of mistake is pretty rare in the RV world because the kits are generally engineered in such a way that most assemblies can only be installed one way or it doesn't fit or work properly easily alerting the builder that something isn't right. That's why I'm kind of surprised by the design in this case.
 
On my aircraft, all control system fasteners are wired or cottered. Anything not cottered is torque sealed. Torque seal probably would have caught this.
I must be looking at this incorrectly, because that's not what I'm seeing?

What i see is a failed ball joint, that allowed the male section to rotate in an axis it shouldn't have been able to and unthread from the control rod. An assembly like that threaded at each end can't disassemble itself unless one end is reverse-threaded.

What am I missing?
 
I'm gonna keep suggesting torque seal til someone agrees or refutes it!

Did I mention torque seal?
 
One of the easier and faster ELSAs to assemble was the recently defunct Skyreach Bushcat. Just bolt-together from what I understood during my earlier interest in buying and building one. There’s no A for Amateur in the regulatory category, but that would be one of the simplest amateur-built airplanes out there. If you can “hack” on a car, you can build a Bushcat (maybe erroneously).
 
I must be looking at this incorrectly, because that's not what I'm seeing?

What i see is a failed ball joint, that allowed the male section to rotate in an axis it shouldn't have been able to and unthread from the control rod. An assembly like that threaded at each end can't disassemble itself unless one end is reverse-threaded.

What am I missing?
The placement of the "ball" and the "socket" were reversed.
 
I must be looking at this incorrectly, because that's not what I'm seeing?

What i see is a failed ball joint, that allowed the male section to rotate in an axis it shouldn't have been able to and unthread from the control rod. An assembly like that threaded at each end can't disassemble itself unless one end is reverse-threaded.

What am I missing?

The ball joint didn’t fail. It should have been assembled into the rod. Instead, it was put on the bracket, and the female inserted into the rod, which unthreaded itself; if it would have had a jam nut, which wouldn’t be needed on that fastener, it would have stayed assembled. As well, if the reverse thread end of the rod were inboard, that female would never have fit - either by bad luck or assembler swapping one end of the rod for the other, it fit.
 
I must be looking at this incorrectly, because that's not what I'm seeing?

What i see is a failed ball joint, that allowed the male section to rotate in an axis it shouldn't have been able to and unthread from the control rod. An assembly like that threaded at each end can't disassemble itself unless one end is reverse-threaded.

What am I missing?

The male member of the ball joint was free to rotate axially, independent of movement by the push tube.
 
Ok I see it now.

71511084468598752e2534ba30a41015.jpg
 
The ball joint didn’t fail. It should have been assembled into the rod. Instead, it was put on the bracket, and the female inserted into the rod, which unthreaded itself; if it would have had a jam nut, which wouldn’t be needed on that fastener, it would have stayed assembled. As well, if the reverse thread end of the rod were inboard, that female would never have fit - either by bad luck or assembler swapping one end of the rod for the other, it fit.
Ahhh...got it, thanks.
I was looking at the photo, assumed it was a failure rather than both of the ball joints being oriented incorrectly.
 
This specific kind of mistake is pretty rare in the RV world because the kits are generally engineered in such a way that most assemblies can only be installed one way or it doesn't fit or work properly easily alerting the builder that something isn't right.

Changing the size of one of the threaded ends or having the threads on the push rod instead of the ball joint (like Sonex) thereby making one end a female coupling, would eliminate this assembly error.

I also wonder why the excess threads above the plate didn't trigger the builder to investgate further.

I bought a plane some years ago that was not completed. In the shop I went over the plans and the plane. Even still it was at a followup inspection that I noticed an aileron rod improperly connected inside the wing as it was under the bell crank on one side instead of on top of it like the other side. I thought at first that perhaps this was proper being on opposite sides but a consult of the plans proved that it had been assembled improperly. I hated to admit that not only did I miss this in my first inspections but had also flown the plane that way.

Would it have ever made a difference? Likely not, unless the cotter pin was ever lost and the nut backed off. Being on top means the control rod might come loose but likely not separate from the bellcrank. On the bottom, if the cotter pin and nut were lost from it would have fallen between the ribs and caused major trouble.

It's good to have more eyes on the build even after it is completed and flying ...
 
I'm gonna keep suggesting torque seal til someone agrees or refutes it!

Did I mention torque seal?
It only helps if it is inspected regularly. How many get that sort of care?
 
It only helps if it is inspected regularly. How many get that sort of care?
Well E-AB/E-LSA aircraft have to receive a condition inspection with the scope and detail of part 43, Appendix D (IOW analogous to an annual) every 12 months. How well those inspections were done is anyone’s guess. This particular error could have easily been overlooked. Torque seal (which I use prodigiously throughout my own plane) might have helped identify the problem assuming the issue didn’t start after the last condition inspection.
 
However, once the ELSA aircraft has its Experimental certificate, the match-the-exemplar system goes out the window. Once the certificate is signed off, the owner of an Experimental Light Sport can modify their aircraft to their heart's desire. The owner of an ELSA RV-12 can install an R-985, extend the wings to make the plane a motorglider, whatever. All that will happen is that the plane has to go back into its test period.
Well… kind of. Any modification that takes the airplane out of the LSA envelope (gross weight, speed, prop, gear, and so on) would essentially result in a lawn ornament. It can’t be an LSA, ever again. It can’t be EAB.

You could maybe go for one of the other Experimental certificates, I suppose.
 
Well… kind of. Any modification that takes the airplane out of the LSA envelope (gross weight, speed, prop, gear, and so on) would essentially result in a lawn ornament. It can’t be an LSA, ever again. It can’t be EAB.

You could maybe go for one of the other Experimental certificates, I suppose.
Yes, good point. Guess I'll have to sell that R-985. :)

Ron Wanttaja
 
It had at least 2 annuals.
As an aircraft mechanic, I was into many airplanes that had obviously not been properly inspected for a very long time, even though there were annual inspections dutifully recorded in the logs for many years. Why else would I find:

-Seized control cable pulleys, seized because the factory lube in their bearings had hardened long enough ago that the cable had chewed a section of the cable groove deeper

-Cracks in bulkheads, firewalls, engine mounts, stab and control surface spars, and so on

-Fuel strainer bowls seized to the head casting, not having been off in many years. Corrosion pitting in the bowl, nearly all the way through the aluminum

-Engine shockmounts hard, aged, shrunken and rattling in the engine mount; many sagging mounts

-Wheel bearings that were rusty and making audible grinding noises when the airplane was moved by hand

-Bolts that had missing nuts. Saw that on the lower wing strut attachments on a 172, along with some 130 other serious snags on that airplane two weeks after it had an "annual." Same airplane had its horizontal stabilizer forward spar broken completely through. Stab was flexing mightily and would have failed in some high-G maneuver. Only the skin was holding it. Elevator cables were really loose, too

-Broken stiffener channel rivets in a 185 fuselage belly bulkhead, letting the aileron and flap cables go loose as their tension pulled the bulkhead web forward into a curve

-Cracked exhaust components, a deadly flaw, a few hours after an "annual;" the cracks were 3/4 of the way around the tubes, two exhaust riser tubes on the same airplane, and didn't happen in the last four or five hours. Cracked and leaking mufflers in some airplanes

-Oil and fuel hoses with certification tags indicating that they were manufactured 40 and 50 years earlier. Hard as wood, and broke when flexed. Wire braid in them all rusty

-Magnetos far beyond economical repair because they had not been opened for inspection since the airplane had been built 30 or 40 years and thousands of hours earlier. Internally corroded, bearings shot, etc.

-Vacuum hoses hard, cracked. Aluminum vacuum lines rotted through by moisture

-Mouse nests of great age (straw and other organic debris rotted to dust) behind interior panels and inside wings and tails

-Seat belts so old the nylon has gone stiff. Their production tags indicate 40 or 50 years of age. They would never restrain anyone in a crash

-Fraying control cables in bellies and wings

-Carb airboxes coming apart, cracked, sometimes fixed with blind rivets that can release the stem into the engine

-Loose hinge bolts that have wallowed out the hinge brackets. They are supposed torqued up so as to be clamping the "bushing" or "spacer" between the bracket ears, and that spacer is to rotate in the bearing. The bolt is not supposed to be the hinging surface

-Propellers dressed well below manufacturer's minimum dimensions. Nobody referred to any manuals there

-A Continental IO-360 engine leaking oil from behind the alternator/tach drive gasket. The oil was coming from the oil pump idler gear shaft, which was loose in the accessory cover casting. Someone had gooped RTV over it to try to stop the leak. The shaft was wallowing out the hole and the oil pump was not far from failure, meaning catastrophic engine failure

-Outstanding ADs, despite the assurance in the inspection report that "All ADs complied with." Stuff that had been missed for decades sometimes, yet signed off every year

-Amazing amounts of debris and dirt and crud in aircraft bellies, fouling lots of stuff and plugging drain holes. Rusting seat belt anchor brackets. Missing backup plates. Dark blue oil stains under leaking fuel fittings. Once found a mound of the blue stuff an eighth of an inch thick under a fitting that I loosened with my fingers. Nope, nobody was looking for a long time.

No, the minimum inspection requirements are not being met in many airplanes. I could write a book.
 
We get it— not all inspections are accomplished IAW the regs. But most likely this was not the result of a pencil whipped condition inspection, although there’s no way to be 100% sure. The bottom line is this particular installation error could very easily been overlooked even in a thorough inspection.
 
The bottom line is this particular installation error could very easily been overlooked even in a thorough inspection.
With that rod end unscrewing itself, the control sticks would no longer be parallel to each other. That should have raised red flags.
 
With that rod end unscrewing itself, the control sticks would no longer be parallel to each other. That should have raised red flags.
I disagree— the control geometry IMO doesn’t support that conclusion, Also we have absolutely no idea how much time it took for this to happen orif it would have manifested in something discernible to the naked eye.
 
Changing the size of one of the threaded ends or having the threads on the push rod instead of the ball joint (like Sonex) thereby making one end a female coupling, would eliminate this assembly error.
There are three detail drawings on the assembly sheet for the yoke installation. While it's harsh to say so, if one cannot understand what the drawing shows, perhaps they shouldn't be building an airplane.
 
There are three detail drawings on the assembly sheet for the yoke installation. While it's harsh to say so, if one cannot understand what the drawing shows, perhaps they shouldn't be building an airplane.

You'll get no argument from me on that logic but making the assembly a bit more idiot proof (for want of a better term) just might save a life ...
 
For those who think the DAR would find something like this, my one data point is that my DAR barely looked at my plane. He spent 15 minutes looking at my paperwork and no more than 10 minutes looking at my plane. He has one correction - add a missing label on the canopy. He never looked at the control inspection panels, even though I had opened them up. This was an Oakland FISDO DAR 6 years ago. He did take my $800 in cash after being at my hangar for an hour..

And yes, I understand some DARs are more thorough.
 
You'll get no argument from me on that logic but making the assembly a bit more idiot proof (for want of a better term) just might save a life ...
Building an aircraft isn't for everyone. Van's placed three large detail blowups on the assembly drawing. To me that's idiot proofing, but I guess there are some persistent idiots out there.
 
Back
Top