Full Accident Report - T-51 Fatal Crash, NZ

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

don january

Well-Known Member
HBA Supporter
Log Member
Joined
Feb 10, 2015
Messages
3,037
Location
Midwest
Good reading for a lesson to be learned. But very sad reading for the pilot and aircraft and also the Mazda power plant. Thanks for sharing this report.
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,732
Location
Port Townsend WA
I guess the airworthiness inspection didn't catch the improper shoulder harness attachment to the seat.
The shoulder harness should be linked to a substantial hard point with cables. Not the lower seat frame as they proposed.
A low point can cause spine crushing.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
2,041
Location
Republic of Texas
I guess the airworthiness inspection didn't catch the improper shoulder harness attachment to the seat.
The shoulder harness should be linked to a substantial hard point with cables. Not the lower seat frame as they proposed.
A low point can cause spine crushing.
Nearly all T-51's have this seatbelt/harness configuration. And, there have been other off airport crashes, but never with a seat frame failure.

What kind of vertical velocity would be required to generate a 20 G impact? I wonder if the plane was actually at or near a stall, as the right wing dipped, and just came down in a near wings level attitude.

The forward G load was a much more manageable 7.8 G's.

These T-51's are way over built. But how many other aircraft could survive a near 20 G impact?

Did the vertical G load cause the seat frame to break?
 

BBerson

Light Plane Philosopher
HBA Supporter
Joined
Dec 16, 2007
Messages
14,732
Location
Port Townsend WA
Never attach seat belts to the seat. The seat pulled loose and he hit the panel. Look at your car, they are attached to the car body.
This was a basic flaw.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
2,041
Location
Republic of Texas
Never attach seat belts to the seat. The seat pulled loose and he hit the panel. Look at your car, they are attached to the car body.
This was a basic flaw.
We had a Mercedes Benz 230E. My wife was involved in an accident with this car. The seat failed, and she impacted the steering wheel, and the windshield with her head. She survived the accident. She was wearing her shoulder harness, and lap belt.
 

12notes

Well-Known Member
Lifetime Supporter
Log Member
Joined
Aug 27, 2014
Messages
1,169
Location
Louisville, KY
The seat belt was attached to the upper bar of the seat back frame which is welded to the fuselage at 6 points, although the lower one is no help in this situation. The airworthiness notice stated that the seat belt should only be attached to the lower tube of the frame. If you look at the bracing, you can see that the upper bar would act as a lever on the attachment points, where the lower bar would be much stronger from the location alone, and has a brace at it's attachment point.

Titan mustang seat brace.png
Titan mustang seat brace crash.jpg
 

rv6ejguy

Well-Known Member
Joined
Jun 26, 2012
Messages
4,206
Location
Calgary, Alberta, Canada
19.6 G vertical MIGHT be survivable but you'd likely have serious spinal compression. Many people die from this aspect rather than frontal G forces.

I watched some of this owner's videos a few years back on YouTube and one thing that struck me then was the dizzying procedure he went through to check operation of the ECUs and backup systems. In an emergency at low altitude, you don't want to throw more than one switch. No time to sort out complex switchology. Then after changing the switch layout (maybe he also thought the original setup was too complex), you want to rehearse the new procedures and switch placement a lot and just prior to takeoff each time.

No mention of a runup preceding the flight. That might have saved him discovering the problem before he was moving.
 

Riggerrob

Well-Known Member
Joined
Sep 9, 2014
Messages
1,880
Location
Canada
I would still want more diagonal braces leading back to upper longerons. Even steel cable would do.

On a related note, a friend is building a wooden Chilton. Planes say to attach shoulder straps to a flimsy wooden spacer between upper longerons. He concluded that it would be reduced to kindling during a forced landing. So he ran a cable back to rear fuselage and bolted the cable to steel tailwheel fittings.
 
Last edited:

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,021
Location
US
19.6 G vertical MIGHT be survivable but you'd likely have serious spinal compression. Many people die from this aspect rather than frontal G forces.
NATO studies found a 10% likelihood of serious spinal injury at about 17Gs vertical. I don't know what their assumed population was, but the older pilot in this case might have been considered more prone to back/neck injury than the population of concern in a NATO study.

From the CAA report:

By calculation, the safety investigation determined that the accident forces which the pilot would have been subjected to were within the range for human survival. The forces calculated were approximately 19.6g vertical and 7.3g horizontal. These forces were based on an estimated airspeed of 70 knots when the aircraft struck the ground. The range for human tolerance is much higher than those the pilot was subjected to during the accident sequence4. If the pilot’s shoulder harness had prevented him from flailing forward as designed, he would have most likely survived the accident.
The report criteria used was "survival" rather than "injury", and I'd think most people would agree with the report that 19.6G vertical would be quite survivable.

To calculate the G's to which the pilot was subjected, the investigators must have made some assessment of the distances over which the horizontal and vertical acceleration occurred. With regard to the vertical acceleration, a lot would depend on whether the main gear was in play and the provision for a crush zone under the seat pan (or Conform-type seat cushions).

No mention of a runup preceding the flight. That might have saved him discovering the problem before he was moving.
According to the report, an eyewitness (an associate of the pilot) said there was an extensive runup prior to takeoff, and that this was the pilot's normal practice. No specific mention as to the power setting used or if both ECUs were checked out.
 
Last edited:

BoKu

Pundit
HBA Supporter
Joined
Aug 15, 2013
Messages
2,872
Location
Western US
The seat belt was attached to the upper bar of the seat back frame which is welded to the fuselage at 6 points...
When I measure the photo on my screen (yes, I put the ruler up to the screen), I see that the seat belt is 9mm and the square tube it is attached to is a bit over 2.5mm. Assuming that the belt is standard 2" webbing, simple algebra suggests the tube is somewhere around 5/8" or 3/4".

The equivalent structure in the Vans RV-8 is a massive steel weldment made with 1.5" or so OD round tubing. It has generous diagonal struts near each end, and was clearly designed to accommodate huge forces applied by the shoulder harness in addition to stiffening the cockpit opening between the two seats. In comparison to the RV-8, what's shown here for the T51 appears woefully undersized.

--Bob K.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
2,041
Location
Republic of Texas
Here is something else to consider...the complete and total lack of any blood spatter, on the panel and glare shield. Head trauma injuries are notoriously bloody, when the victim is alive.

Several have proposed that the pilot was dead before the crash, 80 years old, suffering a cardiac event in concert with the engine problems. This would explain the fact that there is no visible blood in the photos.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
2,041
Location
Republic of Texas
When I measure the photo on my screen (yes, I put the ruler up to the screen), I see that the seat belt is 9mm and the square tube it is attached to is a bit over 2.5mm. Assuming that the belt is standard 2" webbing, simple algebra suggests the tube is somewhere around 5/8" or 3/4".

The equivalent structure in the Vans RV-8 is a massive steel weldment made with 1.5" or so OD round tubing. It has generous diagonal struts near each end, and was clearly designed to accommodate huge forces applied by the shoulder harness in addition to stiffening the cockpit opening between the two seats. In comparison to the RV-8, what's shown here for the T51 appears woefully undersized.

--Bob K.
There have been a number of high G load, off airport arrivals in the T-51. As far as I know, never before has the seat frame failed.

In addition, I would wager that a good flight helmet would have saved his butt.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,021
Location
US
Here is something else to consider...the complete and total lack of any blood spatter, on the panel and glare shield. Head trauma injuries are notoriously bloody, when the victim is alive.

Several have proposed that the pilot was dead before the crash, 80 years old, suffering a cardiac event in concert with the engine problems. This would explain the fact that there is no visible blood in the photos.
The pilot's shoulder restraints failed and his head struck the instrument panel under a horizontal acceleration of less than 8g's. Is that acceptable? It didn't lightly tap the panel, it mangled switches. This isn't about helmets or heart attacks (anymore than it is about "Mazda" engines).
An honest evaluation of this accident and effective responses are the best way to honor the memory of this T-51D builder.
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
2,041
Location
Republic of Texas
It is about the entire event, not just limited to a single phase of the crash.

We have been discussing multiple angles...and the total load on the plane was nearly 20 G's. That certainly would have caused a failure of the frame. The horizontal G factor should not have contributed here.

But here is my question for you pilots. How do you get to a 20 G deceleration load? The pilot obviously did not flare into the crash. The ground run had to have been very short. The photo clearly shows the wing is broken. Trust me, the T-51 is way overbuilt, and to break a spar would require a massive load and/or force.

I just got an email from a close friend of the deceased pilot. He stated that the landing gear was in a transit position, as it was folded up and back, underneath the wings. This very well could have been a contributing factor.

The overall goal is for all of us to learn from this tragic event, in hopes of being able to prevent this from happening again.
 

lr27

Well-Known Member
Joined
Nov 3, 2007
Messages
3,822
The vertical load probably wouldn't cause a failure of the upper shoulder belt attachment, I should think.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,021
Location
US
We have been discussing multiple angles...and the total load on the plane was nearly 20 G's.
No, that is not what the report says. It says >the pilot< was likely exposed to about 20g in the vertical direction and about 8 g in the horizontal direction. These forces on the pilot would be different from the forces experienced by various portions of the airframe. It is all about the change in velocity of each component and the time/distance over which it occurred.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
6,021
Location
US
The CFRs say certified aircraft must have restraints and seats that protect passengers in crashes with forward accelerations of 9g. It is a very reasonable standard. The NZ accident report indicates the welds in the accident acft were adequate. This does not appear to be a construction issue.
 
Last edited:

Daleandee

Well-Known Member
Joined
Sep 11, 2015
Messages
1,322
Location
SC
The pilot's shoulder restraints failed and his head struck the instrument panel under a horizontal acceleration of less than 8g's. Is that acceptable? It didn't lightly tap the panel, it mangled switches.
This is a correct assessment given what is stated in the report:

If the pilot’s shoulder harness had prevented him from flailing forward as designed, he would have most likely survived the accident.
Some time back there was a thread about cockpits designs (I could not find it) and in that thread I mentioned that although some Sonex builders install vertical panels and have all kinds of stuff sticking out from the panel, including a sun-shield, I had refused to do that.

My thinking is that in a forced landing there is reason to believe that I may have to deal with some very high horizontal deceleration forces. In order to add safety I choose to keep the panel as far from my face as reasonably possible i.e. (as Dick Fisher would say) "what do the plans say?"

Looking at the accident photos and comparing his panel with other photos of the same found on the net, it appears there are many variations as some have large overhangs and others not so much.

Something else to consider as you build is not only good secure seat belt attachment points but what your face gets to hit if they don't work as planned.

FWIW,

Dale
N319WF
 

TXFlyGuy

Well-Known Member
Joined
Apr 25, 2012
Messages
2,041
Location
Republic of Texas
Consider that the 20 and the 8 may have added up to more G load, perhaps 23 (+/-), after adding in the force vectors. A 200 pound pilot would generate 4600 lbs of force on the seat/frame. That is approaching/exceeding the limits of most light aircraft seat types.

Also, this example was an early SN, and had been modified in the field. It strongly appears that there was hydrogen embrittlement where the seat frame was welded to the upper frame stringer. The manner in which the members just "snapped off" outside the heat affect zone of the weld points to this.

From my experience, normalized 4130 will tear and rip, but will not snap off as shown in the photo. That was an early frame, with the seat more vertical. And the welds were not properly normalized. Again, this was a mod done in the field.
 
Top