What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

TKATC - Report

BillL

Well Known Member
I have been checking NTSB site every few days for a year. Today the site says the final report will be released on 6/10/2015 (estimated).

This is not to start discussion (yet), just an advisory note. Link(s) will be posted when released.
 
Last edited:
An in-flight breakup, but I can't see a probable cause. Did I miss it?

You should hit the supporting documents link above and read the trajectory and structural reports. That will give the general idea until the probably cause is released. Possible flight dynamics of some maneuvers are beyond my experience, so it may or may not be clear.
 
Last edited:
I read all of the supporting docs and as far as I can see, it doesn't sound like the plane fell apart on its own. That would be good news in re: worrying about Vans design, if confirmed.
 
Sad

A sad and unfortunate read. Reads a lot like the c-gndy accident several years ago up in Ontario where I lost a buddy.

Stunningly similar. Fly save guys. Great planes...just not mx2s.
 
An in-flight breakup, but I can't see a probable cause. Did I miss it?

The first couple of paragraphs of page 1d and page 1e of the factual report are pretty telling of the information that will likely be in the probable cause report.
 
I have a question to pose to the engineers, and others in our group with metallurgy experience.
The report detailed "45? cupping" at the failure point on the crankshaft flange. The photo of the crankshaft clearly shows the "cupping". My question is, what causes a failure such as this? It is interesting to note that the failure is quite linear and clean, and does indeed have a 45? beveled appearance at the edges of the failure point.
Just trying to learn something here!
 
Well, the Probable Cause is released. Nothing that was not already said in the docket listing. Overspeed, 220 kts. Inflight breakup.

Reading and comparing the results back to the Canadian 7 ( http://www.vansairforce.net/safety/G-GNDY.pdf ) with the same results, the VS/Rudder failure appearance seems eerily the same. The same in terms of the physical separation. Flutter was not stated in the NTSB report, but I don't know why. Maybe it "is the meaning of is".

This being the second occurrence of the same root cause, and at a lower airspeed than 234 kts has me wondering what physical factors go to the direct cause of decreasing flutter from about 250 knots to 220 knots.

We can see how Tony rebuilt the rudder and had several layers of paint, but there is no information available that I can find that relates the rudder mass and CG to any change in flutter speed. Further, that aspect was not even mentioned in this report.

I thought that 15% margin for "any condition" was supposed to be provided (regs) for Vne to any analysis or testing for margin. Does this reduction to 10% due to unknown cause concern anyone?

Probable Cause Full Narrative
Docket -listing of all supporting information
 
Last edited:
I thought that 15% margin for "any condition" was supposed to be provided (regs) for Vne to any analysis or testing for margin. Does this reduction to 10% due to unknown cause concern anyone?

No, because:
A) Vne is/was determined by multiplying Vd (design dive) by 0.9. See FAR 23.1505.
B) Full/abrupt control inputs anywhere near the speeds discussed are so far outside the aircraft's design envelope that it's essentially irrelevant. Such control inputs aren't appropriate above Va, let alone well past Vne.

If you are aware of another piece of reg that discusses 15% in relation to Vne, I'd be interested to learn more.
 
No, because:
A) Vne is/was determined by multiplying Vd (design dive) by 0.9. See FAR 23.1505.
B) Full/abrupt control inputs anywhere near the speeds discussed are so far outside the aircraft's design envelope that it's essentially irrelevant. Such control inputs aren't appropriate above Va, let alone well past Vne.

If you are aware of another piece of reg that discusses 15% in relation to Vne, I'd be interested to learn more.

I am not a regs expert, not even close. My statement was related to what I read in FAR 23.629 - it may not apply and I am just flat wrong, but that is the reference.

edit: it also says "(1) Adequate tolerances must be established for quantities which affect flutter, including speed, damping, mass balance, and control system stiffness; and" Adequate tolerances, implies that the mass balance, torsional stiffness etc should have some known variation curves that show the envelope. This is what would be very helpful in not reducing the (flutter/Vne) margin
edit 1 -fat finger correction - the "reg" was actually an AC 23-629, not 23.269 as previously posted. I was looking at an older version. The new version does not mention any ratio with V dive. I stand corrected on the reference to V dive and Vne. Thanks
 
Last edited:
Flutter may not have been an issue. A large control input, especially when exceeding Vne, could result in a static overload. See AA587.
 
We can see how Tony rebuilt the rudder and had several layers of paint, but there is no information available that I can find that relates the rudder mass and CG to any change in flutter speed. Further, that aspect was not even mentioned in this report.

My recollection is that it also had significant quantities of filler applied in spots to smooth it off before the paint was applied... So the imbalance was more significant than *just* extra layers of paint.
 
Flutter may not have been an issue. A large control input, especially when exceeding Vne, could result in a static overload. See AA587.

Yes, but reading the details of the report, it doesn't seem consistent with the other actions. Is there something you saw in there that hints at this?
 
Last edited:
Is this what you are looking for?

All fractures exhibited characteristics consistent with static overload. No evidence of multiple dynamic loading fractures was found, nor evidence of fatigue characteristics or other preexisting damage.

I understand this to mean that the stabilisers were overloaded and basically each one snapped off, in one direction, with no cyclical loads leading to the break.

The whole situation is terrible.
 
Where's the other prop blade? It was stated that it had an MT 3-blade prop, but the report only accounts for two blades?
 
Is this what you are looking for?

All fractures exhibited characteristics consistent with static overload. No evidence of multiple dynamic loading fractures was found, nor evidence of fatigue characteristics or other preexisting damage.

I understand this to mean that the stabilisers were overloaded and basically each one snapped off, in one direction, with no cyclical loads leading to the break.

The whole situation is terrible.

Exactly. That sentence is saying that there was no evidence of flutter. The airplane was simply overloaded.
 
Is this what you are looking for?

All fractures exhibited characteristics consistent with static overload. No evidence of multiple dynamic loading fractures was found, nor evidence of fatigue characteristics or other preexisting damage.

I understand this to mean that the stabilisers were overloaded and basically each one snapped off, in one direction, with no cyclical loads leading to the break.

The whole situation is terrible.

That is what i thought at first, then looking at the VS-rudder parts from C-GNDY, and looking where the rudder weight landed in respect to the other parts I got the impression that there was more than a single one time force that started the process.

Exactly. That sentence is saying that there was no evidence of flutter. The airplane was simply overloaded.

At first, I thought that too. But . . . the separation results of the two incidents look exactly alike (VS/Rudder). Flutter discussed on one, not the other.

Flutter was, oddly, never mentioned in the NTSB report. They did not say it was, or it was not. From where the parts landed, it's almost like the rudder weight came off first, then the other parts.

220 knots for failure seems too close considering 190 knots, IAS, at 8500 is 220 TAS.
 
Tony and I swapped emails about painting last year. I believe he built a completely new rudder and painted it himself. I don't imagine there was any filler.
 
TKATC

VNE is one thing , A Split S at VNE is another . Enjoy your plane , loops and rolls are big fun done within the designed speed range , agressive stick and rudder inputs at VNE might not end well .
 
Did he first repair the damaged rudder and then later build a new rudder?

No repair, just rebuilt. Link on post #11. There was a self etch layer, hi build layer (sanded down) a yellow/gold layer, a red layer, then the clear coat. No mention of rudder weight in the report IIRC.
 
Prop

I've never done a split S. Would a constant speed prop have helped control the speed in this situation or would have made it worse?
 
Prop

I did aerobatic instruction in the Pitts and other aircraft full time for two years, and part time for many more years. When I had a new trainee I would ask what they had been flying and what their goals were. I had a gentleman who had been flying for a long time. He had an RV6 that he had bought, constant speed prop. He admitted very candidly and in some detail that he had been trying to teach himself rolls in the RV. he completely lost control and split s's , reaching near 300 m/h and a lot of G. The thing that I believe saved him was the constant speed prop. As he was accelerating rapidly downhill he final remembered to pull the throttle back. he was VERY close to the ground when he regained control. I have seen this scenario far too many times.
An interesting side point is that in an airplane such as the Sukhoi, with a very high G limit (around 11G depending on model and weight)and a relatively high Vne, a downward looping maneuver can be started at Vne with the speed controlled by pulling/pushing very hard, full power all the way. The Sukhois are relatively high drag airplanes, expecially above 200 m/h. The two seat Sukhoi, from a hammerhead to Vne, vertical down with full power all the way, takes approximately 4000' to reach near Vne. 5 m/h below Vne pull really hard and the airspeed doesn't change.
I am a mostly self taught aerobatic pilot. I taught myself the basics, did basic aerobatics for two years and then got two hours of instruction in the Pitts. NO flat spin training. I did the flat spins on my own.
A pilot who tries to teach themselves aerobatics has a fool for an instructor.
 
Soo - use a 6 tail?

. . .He had an RV6 that he had bought, constant speed prop. He admitted very candidly . . . he completely lost control and split s's , reaching near 300 m/h and a lot of G. . . .

Ya know, there is more than one of these stories, RV6 Vne is 210 mph. His VS/rudder did not come off. How do we assure that on a 7, get a 6 tail? I have one, actually the early 7, .020 skin, will it give me more margin? Was he braggin' or is it just more tolerant? Anyone know? PM?
 
My Rocket has a standard issue RV-8 tail group and thanks to an unfortunate flying experience I don't want to expound on further, I believe there is significant flutter margin in a properly constructed tail. I have personally been well in excess of 270 knots (followed by a 6 G pull at gross weight) with no damage of any kind. Based on that experience as well as the number of Rockets flying with -8 tails, I think there is more to this than a "simple" overspeed.
 
No repair, just rebuilt. Link on post #11. There was a self etch layer, hi build layer (sanded down) a yellow/gold layer, a red layer, then the clear coat. No mention of rudder weight in the report IIRC.

From the Accident Docket: "Witness Statements and interview Summaries":

Page 1 of I
Todd Gunther
Air Safety Investigator
Eastern Region
Date: September 21, 2013
Person Contacted: Amy Kelly (Spouse)
NTSB Accident Number: ERA13FA424
RECORD OF CONVERSATION
According to Mrs. Kelly, Her husband had the last annual inspection on the airplane done in
Texas. The plane was there for 7 days and while sitting on the ramp, the rudder was damaged by
the wind, prop blast, or jet blast. The rudder had a hole in it but her husband was able to fl y it
back to Cross Keys in March, and then built a whole new rudder.
He then got a new gustlock, to
keep the rudder from being damaged again.
Page 1 of l
Todd Gunther
Air Safety Inves tigator
Eastern Region
Date: September 21,2013
Person Contacted: Butch H arold Delaney (Friend)
NTSB Accident Number: ERA13FA424
RECORD OF CONVERSATION
According to Mr. Delaney , the airplane had previously had its rudder damaged not too long ago,
and the pilot had duct taped the rudder, and flown it home. He then built a new rudder to replace
the damaged one.

This was his second accident. Both occurred in Hamilton Township .
During the first accident he had run t he airplane out of fuel.
Mr. Delaney also believed that the pilot was building another RV airplane when this accident
occurred.
 
My Rocket has a standard issue RV-8 tail group and thanks to an unfortunate flying experience I don't want to expound on further, I believe there is significant flutter margin in a properly constructed tail. I have personally been well in excess of 270 knots (followed by a 6 G pull at gross weight) with no damage of any kind. Based on that experience as well as the number of Rockets flying with -8 tails, I think there is more to this than a "simple" overspeed.

One difference is your rudder has a single piece skin and isn't double-flush riveted like this one (and the Canadian -7 IIRC) was. Could that be significant?
 
My Rocket has a standard issue RV-8 tail group and thanks to an unfortunate flying experience I don't want to expound on further, I believe there is significant flutter margin in a properly constructed tail. I have personally been well in excess of 270 knots (followed by a 6 G pull at gross weight) with no damage of any kind. Based on that experience as well as the number of Rockets flying with -8 tails, I think there is more to this than a "simple" overspeed.

One difference is your rudder has a single piece skin and isn't double-flush riveted like this one (and the Canadian -7 IIRC) was. Could that be significant?

Yes, and the 8 has a .020 skin, with smaller area and formed trailing edge with a counterweight. The failed rudder(s) is(are) .016 skins with the riveted TE. This "8" rudder is also known as the 7 "short" rudder shipped with very early 7 kits. Significant? good question indeed.
 
The sequence of failures is the thing I am curious about.

If that first failure had not happened perhaps there may not have been other failures and a recovery may have been possible.

The lower third of the rudder was the first piece that apparently separated but I have doubts that it was the first part to fail.

Is there a way to know if the vertical failed first and then the left horizontal or vice versa?

Perhaps if it was possible to know where the first failure occurred it may have been possible to gain enough safety margin with a design modification to survive the incident.

An event like this will likely occur again someday so learning all we can about this may save someone else.

What is the g-load to failure the the vertical and horizontal?

Glenn Wilkinson

P.S. Just thinking out loud...I know about the overspeed, maybe abrupt control inputs etc.
 
The lower third of the rudder was the first piece that apparently separated

I don't think there is any evidence to support this theory.
In fact I think there is specific evidence to suggest it was not.
There was extensive damage done to the fuselage side skins where the rudder cables exited the fuselage (not sure if it is visible in the photos posted in the docket, but it was clearly apparent in some of the photos I viewed during the initial investigation). The rudder cables sawed large holes in the skins before the cable attach points on the rudder horn finally failed (probably only took a few seconds). It could have been during these violent gyrations that strong vibrations were induced in the fuselage which caused a loss of the canopy and the wing tip.... but that is pure speculation... we will probably never know precisely what happened.
 
He admitted very candidly and in some detail that he had been trying to teach himself rolls in the RV. he completely lost control and split s's , reaching near 300 m/h and a lot of G. ... I have seen this scenario far too many times.

I read a lot about people losing control and "falling out" during a roll... how does it typically happen?
 
I read a lot about people losing control and "falling out" during a roll... how does it typically happen?

Im not being flippant here, but sadly enough it happens when people try to teach themselves aerobatics. Get some real trainng and learn how to do them correctly and also how to safely recover from a botched maneuver. Most peoples first reaction that have had no unusual attitude training is to just "pull up" when things go wrong. Upside down and fast that just may get you hurt.
 
I read a lot about people losing control and "falling out" during a roll... how does it typically happen?

+1 on what Widget said. To answer your question specifically, folks with no training make two mistakes when teaching themselves how to roll. The first is starting without getting the nose up. The second is maintaining back-pressure thruout the roll. As a result, the nose could be as much as 45 deg low when reaching the inverted position. Then panic sets in and bad things happen.

Tony
 
Yes, and the 8 has a .020 skin, with smaller area and formed trailing edge with a counterweight. The failed rudder(s) is(are) .016 skins with the riveted TE. This "8" rudder is also known as the 7 "short" rudder shipped with very early 7 kits. Significant? good question indeed.

Bill, you have both right? Want to hinge them horizontally and measure trailing edge weight? Put another way, do they have the same balance factor?
 
+1 on what Widget said. To answer your question specifically, folks with no training make two mistakes when teaching themselves how to roll. The first is starting without getting the nose up. The second is maintaining back-pressure thruout the roll. As a result, the nose could be as much as 45 deg low when reaching the inverted position. Then panic sets in and bad things happen.

Tony

+2 on Widget and Tony. A few years back I went to Acro Dynamics in Mooresville, NC, and took an upset training course with Mikey Mathews in his Citabria. He told me the first step in recovery from an unusual attitude is to roll wings level. He said the fist thing newbies do is pull on the stick, it is instinctive.

I thought, not me, I'm an engineer and I have all this worked out in my head, all the vectors and velocities, etc. Hah!!

He got us inverted with the nose about 30° up and said "Your plane, recover." What did I do? Without thinking I yanked on the stick, but it didn't go anywhere! Fortunately he had his hand blocking the stick, he said "My plane" and rolled us level. Then he said "Let's try that again", and we did fine.

I still think about that event when I fly, and when I am doing aerobatics (loops and rolls) I try to always remember "Wings level when in doubt..."

That course was some of the best money I have spent....
 
Last edited:
+1 for the formal aerobatic training advice, preferably in an RV or a similar handling aircraft. I have never flown a Pitts, but I have a decent amount of time in a Super D, and it handles a lot like an RV, probably closer than a Pitts.

I really believe a lot of this discussion is rooted in denial. The report specifically states that there was no evidence of metal fatigue. It is a fairly cut and dry case of pilot error. If you read the radar data section, you'll notice that he apparently entered the maneuver above the aircraft's Va. It is really not all that hard to imagine control surface damage if abrupt inputs are made with the aircraft over Vne. I don't mean to be callous, but trying to compare this to other accidents, speculating on things like flutter, and looking for design flaws or poor construction (which there is no evidence of) is just making excuses for the pilot.

I didn't know Tony, but I have heard lots about him from Vlad. The whole thing is very sad, and I understand this community lost a great asset with his passing. I know there was a lot of speculation about the airworthiness of the empennage when this happened, and it led to inspections and an eventual AD. As it turns out, according to this report, those cracking incidents were unrelated. The takeaway is that the design is a solid one, but that proper technique and training is essential.
 
I personally feel the design is good. But, I aggonized over this accident for a long time. Tony was a customer, and even though I never met him, we talked on the phone several times. After the accident, I discussed the tail flutter theory with members of Team AeroDynamix ( whom I sponsor), and would probably KNOW what the limits were. I was jaded some, because I fly backseat in a F1 Rocket. So----you can see my concern. I was told by ALL of them to stick to the published speed ratings, and I'd be ok. One member, ( nameless) stated he had been 250 kts and no adverse affects. Well, I'm not one to 'like' to fly inverted:eek:, so I felt with my particular misson, I would be ok.

I am saddend (still) by Tony's passing, but have learned alot by this. We have all come to our own conclusions to the root cause, but I also think we all have learned alot.
Tom
 
I really believe a lot of this discussion is rooted in denial. The report specifically states that there was no evidence of metal fatigue. It is a fairly cut and dry case of pilot error. . . . . but trying to compare this to other accidents, speculating on things like flutter, and looking for design flaws or poor construction (which there is no evidence of) is just making excuses for the pilot.

Just because we agree that the pilot could have done a better job and kept the aircraft within the envelope, and no one has said different, does not mean that there is not room for discussion of the flutter stability of this rudder.

Here is a graph from AC 23.629 1b, if (yes speculation) this rudder follows curve 4, then slipping above the dotted line could immediately generate a failure with little or no indication of fatigue. Hopefully, separate from the pilot actions, there is room for healthy discussion and fact seeking for discussion of a factual comparison of the rudders. Seeing three failures (one conveyed to me via PM) of this "9" rudder at flight speeds regularly seen by the "8" rudders on the Rocket leave me curious. Maybe others are too.

Such a discussion of improving the margin, or understanding what factors could lower the Vne-Failure margin is separate from improving aero skills to avoid such an unfortunate situation.

Flutter%252520Instability.png
 
+1 on what Widget said. To answer your question specifically, folks with no training make two mistakes when teaching themselves how to roll. The first is starting without getting the nose up. The second is maintaining back-pressure thruout the roll. As a result, the nose could be as much as 45 deg low when reaching the inverted position. Then panic sets in and bad things happen.

Tony

I would teach people to aileron roll in several steps in the Pitts. "Dutch rolls" back and forth in level flight, starting with about 45 degrees bank and increasing to 60 degrees. Then practice pulling the nose up to around 30 degrees above the horizon and letting go of the stick. Then pull the nose up to 30 degrees, let go of the stick and place one finger against the right side of the stick and push the stick to the left. 100% success rate even with people who had never touched the controls of an airplane. When they would not roll fast enough I would just say push harder. I rarely ever touched the controls.
This is so much better than the 30 hours or so that I spent split sing out of attempted rolls in the Cassutt at 270 m/h. I did have a chute and did have brains enough to stay at a safe altitude.
 
Ya know, there is more than one of these stories, RV6 Vne is 210 mph. His VS/rudder did not come off. How do we assure that on a 7, get a 6 tail? I have one, actually the early 7, .020 skin, will it give me more margin? Was he braggin' or is it just more tolerant? Anyone know? PM?

He was not bragging. He scared himself very badly and knew that he was extremely lucky. I knew another guy that did the same thing in a RV4 and survived. I think there was some damage to the airplane in that case.
 
Aerobatics

It is good to remember that Van in the early days advocated self taught aerobatics in the RV3 and 4. I once had a lengthy discussion with Van about this. I was very strongly opposed to him promoting that.
His position evolved. The Vans "epistle on aerobatics" is a must read for anyone interested in aerobatics in any type of airplane.
I think the CA RV8 structural failure accident was a turning point. I have always believed that this was nothing more than a crop duster pilot used to relatively slow airplanes with relatively heavy control pressures yanking WAY to hard the first time he touched the controls of an RV.
Read Dan Horton's excellent article in Kitplanes on the Wittman Tailwind. The Tailwind is very sensitive on the controls, especially the elevator. Dan's brief PIO in typical of someone flying the Tailwind for the first time.
While the RV's are not as sensitive, when the issue is compounded by falling out of a roll and looking at the ground much to close, 20 G's happens much too easily.
 
Back
Top