Diamond Rollover: Freak Accident Or A Design Flaw?


I doubt if there are many AVweb readers who, upon reading an accident report, don’t have one of two reactions. (a) Hey, I already knew this and I’d never do that or (b) I didn’t know this, I’ll be careful not to do it in the future. “This” is whatever happened in the accident being described. In my case, upon reading the freak fatal ground accident of a Diamond DA42 in Orlando earlier this month, my reaction is, “hell, I already did that myself.”

Before I get to that, here’s the preliminary report on the accident. A student and instructor were taxiing out for departure at Orlando Executive with a thunderstorm building to the north and east. During the taxi, the wind was 150 at 11 knots, gusting to 16 knots. That’s hardly a no-go. Sixteen minutes later, while they were in the runup area, what appears to be a microburst drove the wind to 54 knots and flipped the aircraft over in place. The student—an experienced pilot from Nigeria—was killed, the instructor with him seriously injured.

Naturally, from the comfort of an armchair, the righteous “how could they have been so stupid?” is understandable. Well, let me tell ya how. Florida leads the U.S. in thunderstorms and lightning activity, equaled only by equatorial Africa, where, ironically, the student was from. It is fairly common in Florida to operate near thunderstorms if the plan is to fly away from the weather to a known clear area, or if weather isn’t severe or building. Everyone will make a different judgment on that and there is no one size fits all. What you might be comfortable with is not the same tolerance I might have.

The FAA recommends 20 miles separation from strong storms, but no closer than 5 miles. I suspect many pilots in Florida are between the two, biased toward closer when conditions allow. As long as clear weather is ahead, I’ve flown within 10 miles of storms on the upwind side. There’s a lightning and turbulence risk, of course, but a low risk of hail in Florida because the freezing level is so high. Looking at the radar loop, there was a large area of convection to the northeast, building over Orlando International. I honestly can’t say I would have departed in such conditions. Can’t say I wouldn’t, either, because I once did.

Coincidentally, it was at Diamond Aircraft’s factory in London, Ontario. Ontario doesn’t get as many thunderstorms as Florida, but those it does get are often in mesoscale lines, with tops in the 50s and 60s and high hail potential. I was busy working on a story and photos when one of my hosts mentioned an approaching line that looked pretty bad. They suggested I hurry up to the FBO and taxi the Mooney down to the factory hangar. I got a ride up the ramp and just as I was getting into the airplane, the temperature dropped and the breeze picked up. The gust front. Then I realized I was rushing things and had to get back out to see if I’d removed the chocks. Nope. In my haste, I threw them in the airplane, thus stealing the FBO’s clearly labeled chocks.

The taxiway down to the factory is about 4000 feet, give or take. The wind, which soon gusted to the mid-30s, was a quartering tailwind from the right. It was levering the tail up and lifting the right wheel a little. I didn’t quite know what to do about this. Should I speed up the taxi and reduce the relative wind? Do the taildragger trick of diving away from a tailwind? Stop and turn into the wind? I’m not sure what I did, other than skid and bump along the taxiway with, to paraphrase a famous Canadian, the wind turning the minutes to hours.  

Diamond DA 40 and 42 aircraft have an integrated roll bar between the front and rear seats. The removable rear hatch aids escape.

My salvation was the long taxiway to the factory ramp, which is sheltered by a woodline. It was contributing windblown leaves and small branches but I dodged all that. I taxied the Mooney straight into the open hangar and chocked it. They pointed out I had stolen the chocks. In times of dire stress, morality is situational on bits of triangular lumber. It rained like hell, the wind blew, but no hail.

Having cheated death, was I more emboldened to repeat the feat or chastened into a never-again moment? I have no idea. It depends on the situation, I guess, which is the essence of GA flying. There’s no cookbook on risk assessment and I reject the argument that there is. Anyway, I was pretty sure even if the airplane upended, it wouldn’t kill me, unless it’s possible to die of embarrassment.

Did the two pilots of the DA42 at Orlando think that? Or did they even consider it at all? Would you? And here a word about Diamond’s crashworthiness. In 2011, I undertook an exhaustive safety comparison of popular GA aircraft crash fatality rates. Diamond swept this field with a 0.35 rate for the DA40 and 0.54 for the DA42. That compares to 1.2 (then) for GA as a whole and 3.25 for the Cirrus line. Since 2011, however, thanks to improved training, the Cirrus rate is much lower. I have no reason to believe the Diamond rate is any different now, but I’ll make that conditional upon reviewing the intervening 10 years of accident data.

A poster in my Facebook feed said there is good evidence that Diamond aircraft aren’t as crashworthy as other models. I would contest this, given the company’s overall excellent safety record, low number of crashes and virtual absence of post-crash fire. Still, in the Orlando crash, someone died who shouldn’t have expected to in a stationary rollover. What happened here?

The DA40 and DA42 have, in effect, a roll bar consisting of a robust composite hoop that runs from the sides of the fuselage slightly aft of directly over the pilots’ heads and serves to anchor the forward-tilting canopy. Did this somehow fail? Was it not adequate for the loads encountered? The answers await further investigation by the NTSB.

When the Diamond aircraft first appeared, every review we did raised complaints about the egress limitations of canopies and upward opening doors. This is true of the Socata line and, to a degree, the Columbia and Cirrus aircraft. We sometimes worried that these would trap people who might have otherwise escaped and survived. There is no persistent pattern of this happening. Yes, the Diamond airplanes have upended—although it’s not common—but this accident may be the first fatality related to lack of overhead protection.

If that’s so, Diamond may want to look at this design and see how it might be improved. It seems to me that it would take an enormous impact load to fail that overhead bar. Meanwhile, this is a point of departure for the rest of us to consider the state of restraints in our aircraft. One of the Diamond accidents involved the failure of a seatbelt fitting. There’s no reason that should ever happen in a properly maintained airplane.

Other AVwebflash Articles


  1. Forecasts wouldn’t hurt. Thunderstorms are predictable, and pilots know from accident reports specifying severe weather as a causal or contributing factor that they should avoid flying in the vicinity of anvil clouds or CCVD. What was the forecast for that airport, and were wind shear/microbursts emphasized via PIREPs, SIGMETs or AIRMETs?

      • And there doesn’t even have to be a thunderstorm for there to be a microburst. I once inadvertently took off into a rain shower where for a moment I wasn’t sure if we weren’t going to get pushed into the ground. I have since learned to give even non-convective dark green/yellow (radar) rainshowers due respect.

      • I agree – WX systems don’t always just move from A to B, they also comes into existance in places where there weren’t any before, sometimes quite fast.
        It’s important to continuously monitor NEXRAD and other in-flight WX sources, especially on XC flights through states known for violent WX.

      • Thunderstorms paint pretty well on RADAR, and forecasters/observers know their tracks, levels and potential. If RADAR is not present, the Mark One Eyeball will suffice to detect threatening weather.

    • Good morning, fellow aviators. When we were in Air Force undergraduate pilot training in southern Georgia, we used to watch the summer thunderstorms billowing up. We would enter into a dive to their alititude, pull up at 600 kts, and see who could top out first, the thunderstorm or the T-38. The race usually ended around 35,000 feet. The thunderstorm is a climbing thing of glory and power. Back on the earth, it would leave four inches of standing water where, half an hour before, there was none. All in all, we learned not to toy with the weather, tempt it, or compete to win. It could swallow you up, take your life with utter indifference, flick you off and away like a gnat on the tail of a tiger.

    • Florida during the summer is non stop scattered showers in the afternoon. The visibility is usually very VFR outside the rain columns. Some randomly build into a thunderstorm. I lived and flew in Florida for many years. It is just something you learned about the weather.
      I was in the Navy in Florida in the 80s. The Navy had to close a recruit training base in Orlando, because people marching around with flag poles made great lightning rods… and you never knew when one of these random rain shower would turn deadly.
      When I worked for the FAA in Ft. Lauderdale as a tech, it was one of the first airports to get a wind sheer system to test, because of the unusual wind changes during these pop up storms.
      These very strong winded relatively small area storms are not the norm. But, you can see when the bottom drops out. I’ve even seen small tornados that look more like water spouts pop up on one side of an airport, while planes were taking off on the other side of the airport and even the tower didn’t notice it.
      Weather in Florida is something you learn to read and get use too, like flying through the mountains and reading the wind coming over them.

  2. I had not heard about this accident — my condolences to the pilot’s family.

    Was an autopsy performed on the deceased? Do we know what exactly was the cause of death?

  3. While the human body can take enormous trauma and keep going its also very fragile if all the cards fall in the right place. Kinda like an airplane being very strong in flight yet fragile on the ground. May he rest in peace.

  4. At the end of Burn After Reading the perplexed FBI officers are discussing their response to an random situation for which they were mostly just spectators: “What have we learned here?” Um, not to do it again?”

  5. A British DA62 crashed in Dubai in 2019, killing four. However that was due to wake turbulence from a heavy ahead on a parallel runway.

  6. With the DA’s long wingspan, the pivot point of the wingtip is going to put the cockpit much higher during the inversion than most other GA airplanes rolling over a wingtip or for any rolling over over the nose (e.g. mushy ground off airport landing).

    Pure conjecture…the design may be adequate for over the nose, but not designed for over the wing…and even if structurally designed for, the impact g-load may push survivable limits due to higher g decel, head contact with structure or belt not tight enough to prevent trauma. A lesson learned may be the need for a helmet and 4 or 5 pt belt to survive this scenario. Race car sanctioning bodies are very particular about rollcages being padded with specifically engineered material (not pipe insulation) even with a helmet. Sad outcome, and perhaps with knowledge of this scenario more thought goes into avoiding/mitigating.

    • It seems unlikely that the airplane flipped directly over the wingtip. It probably went over a diagonal axis similar to a ground loop.

      • I don’t know, based on the pic of the plane on-scene (click through the pic above and zoom in), I think Rich’s scenario is pretty plausible. The entire top of the plane is pancaked, such that the cabin is not much taller than the rest of the fuselage. It wasn’t just the roll bar that failed – the entire cabin was crushed. That to me suggests the whole fuselage slammed into the ground inverted rather than just rolled over onto it’s top diagonally.

    • Let’s say the wind is at 3 o’clock and strong enough to lift the plane 10 feet up then over. If the wing is 20 feet long, it’s likely going to prevent a flip, correct? OTOH, if the wind can lift a plane 20 feet, it doesn’t matter how much less the wing is, it’s going over, and very well may go 20 feet anyways before tipping.

      I may have this wrong, but it seems the wing length is an advantage more often than not. This is back of the napkin physics though.

      • In that scenario the dihedral allows a fair bit of roll before tip ground contact, don’t know if that geometry is better or worse than a typical low wing…that upwind wing is a longer arm with the dihedral putting the wing higher for the wind to act on, especially if already lifted off the upwind gear.

        No idea on this mishap given brief description, but could see a plane getting blown sideways off pavement, gear digs, tips onto wingtip which also digs in, somewhere in there it weathervanes and instead of going over nose/wing it goes over tail/wing

        Given my acft lack dedicated rollover protection, I’m counting on wingloading, geometry, prudence…and luck.

        • If the flip is on the axis of the tip and tail, the extra few feet of wing aren’t really going to add much height are they?


    • Unencumbered by experience, I’ll have to guess here, but to flip a plane sideways, over its wingtip, would (I imagine) require a crosswind. And wouldn’t a crosswind act on the tail as well as the wing and cause the plane to weathervane nose forward before going sideways?

      Hopefully there’s airport surveillance video of the accident that can shed more light on what happened. Perhaps a design flaw, or perhaps a one-in-a-million freak accident of timing, wind angle, and control inputs. Or a mini-tornado/dust-devil.

  7. Quite many years ago, I landed at a S. Fla airport with a thunderstorm rapidly approaching and very near the airport. I landed just fine and was cleared to taxi to the ramp that was approximately a mile away. I got about 1/3 of that distance when the gust front hit and the heavy rain was not far behind. I was in my first plane, a Tri-Pacer and realized I probably wasn’t going to make it to the ramp the way things were going. So I called ground and requested to hold position at a fairly large taxiway intersection. I turned into the wind and used down elevator along with enough power to counteract the wind. Basically I was flying the airplane while the wheels were well planted on the ground. When needed, I adjusted power for the wind and changed direction to always be headed into the wind. It took about 15 minutes, but the storm finally subsided and I could continue my taxi to the ramp. Haven’t had to use that technique since, but it is there if needed. Can’t help you with hail.

    • Something similar happened to me. Was dropping someone off and was hoping to beat the thunderstorms coming. I landed and saw the gust front coming and realized I wasn’t going to make it. So I quickly taxied to the ramp, pointed into the wind, and added power in case I had to go flying and keep the plane from slipping backwards and a little nose down pressure to keep the plane from flying to begin with. That was the wildest 10 minutes or so I’ve experience on the ground.

  8. Hmmmm. Sleepy warm summer afternoon New England ramp and we were all in the airport diner. Planes sitting in the parking spots. None tied down. As I was getting ready to leave the roiling dark mess to the north presaged a summer pop-up.

    With no intention of going flying – I still fired up and taxied into a clear spot on the ramp. My mariner instinct kicking in – seeking clearwater to maneuver.

    I spent 5 mins flying the plane on the ground with her head to wind and a couple of times was light on the tires as the gusts tried to fly her. The popup didn’t hit the airport – but it skidded by just outside the airport fence and as the wind veered I was able to taxi head to wind again and ride out the hail and hammer blows – still light on the tires and still flying her on the ground.

    The wind veering is what caused the issue in parking. Planes were rotated off their parking spots, rocked, rolled, and a couple of wingtips bruised each other. An empty Cherokee ended up on its back.

    I know this because 5 mins later it turned into a lovely summer afternoon again and I got out to join others inspecting the damage – after which I departed.

    So who was stupid? Those who hunkered down in the restaurant (NO!) or the guy who got in his plane and flew it on the ground. (PROBABLY). But my plane was undamaged. We all KNOW that there is always insurance. But we still have a tendency to try and protect property. I may even have been subliminally influenced by many a marine policy I had sailed under. First clause – “You shall act at all times as if uninsured”. An interesting way of asking you not to take unnecessary risks but which then predisposes you to take them when close to marginal situations – if even only to try and get out of them.

  9. There’s an imaginary axis on a tricycle gear airplane between either main gear and the nose gear. In a strong enough quartering tailwind (or, if the airplane is rolling forward and a gear gets hooked on something), the airplane wants to roll over onto it.

    Taxiing in a T-34 at Edwards AFB years ago, the always strong afternoon SW desert winds suddenly increased trying to blow me over that way. It got so bad I decided to shut the engine down (to save it) and hope for the best. It’s a lot like the old trike ATV’s that they no longer manufacture for the very same reason … rollovers. It’s not clear from the Preliminary here if that’s what happened but … something I always consider as a result of that scare.

  10. Here are two cautionary tales about thunderstorms. About 4 years ago my son’s father-in-law called me and said “phew, I’m glad you answered!”. He proceeded to tell me about a small aircraft that crashed in his nephew’s field in SW Michigan. It turns out the aircraft was a trike and the gondola had separated from the wing. I shutter to think of the ride down. The experienced pilot was shuttling the trike from Illinois to an airport in Michigan. I was planning to go flying that day but decided against it because the weather seemed unusual. Shortly after the accident I looked at the weather radar and saw a small dot of rain that developed from green to red and then disappeared right along the aircraft flight path.

    The second story is from 15 years ago when a pilot that lived behind us was flying from Detroit to Newtown, CT in a Cessna 180. According to the accident report, the wings were found 4 miles from the fuselage somewhere in the middle of PA. There were thunderstorms in the area. This pilot had spent years in flying in South American documenting rain forest deforestation.

    Build updrafts are invisible and flying into one can have disastrous consequences.

  11. I will note that flight attendants have been seriously injured by hitting to cabin ceiling in extreme turbulence (as in broken necks, partial paralysis, etc.)
    The decedent may have simply hit the canopy in an unfortunate way; taxiing, were the shoulder harnesses on? Everything very snug? You do sit a little in front of the “hoop” in some Diamonds.

  12. Three Diamond incidents of note here.
    1. There was a wake turbulence case that flipped a DA20 causing it to land inverted. Occupants walked away. I think the plane was repaired, but not sure.
    2. There was a DA 20 engine failure with an experienced pilot and son. They tried to make a neighborhood park, but pancaked on a residential street. IIRC, Son experienced a broken leg, and dad died at the scene. The cockpit was intact, but he had a similar brain injury to Dale Earnhardt (I may have the name wrong) a race car driver who famously died because the geometry and velocity of the impact even while all the safety equipment prevented external injury.
    3. This final story was told to me by the retired crop duster pilot who was in the DA40 he had rented from a school on the field. I believe he was 70 at the time of the incident.
    He was flying into Georgetown, Texas when an experimental aircraft (Thorpedo) impacted the top of the DA40 on final approach. He does not actually recall anything after that until he awoke in the plane minutes later, but we know both aircraft impacted the runway and ended up in the grass on either side. He noted the stick had broken off in his hand when he got out and looked over at the other plane and pilot. They communicated the usual “I’m okay, how are you?” And the pilot of the Thorpedo collapsed. The Thorpedo was in pieces and it’s pilot had only been standing due to adrenaline because he was about to spend months in the hospital with numerous broken bones and other injuries. Our crop duster pilot took a ride to the ER after a mild protest. He got a bandage for a sprained wrist likely saved by the weld on the stick designed for just this kind of impact, and a few bandages. Treated and released, as they say. The Diamond was back on the line after what seemed like a very short time and is likely still flying today after it’s midair collision. I believe the DA40 and DA42 share the same cockpit design with some difference in the seats and the same parts for the windows and doors.

    Now, it’s possible there is some quirk in the DA42 design that this wind managed to uncover, but over all, Diamonds are winners when it comes to safety. I doubt very much that a regular spam can design would better protect the people inside as well.

    Diamond now belongs to the CCP. I believe the strange decision we have made forcing new designs to compete with antique designs that would not or should not be certified today has killed, and will continue to kill many pilots and passengers.

    • Earnhardt’s suffered a basal skull fracture in which the skull breaks off of the spinal column – sometimes referred to as an internal decapitation. The proximate cause was running into the wall, but it was a relatively mild impact which should have been survivable – he hit at an angle, not head on. The impact WAS enough to whip his head forward, yanking it loose.

      The sad thing is that a strap and collar arrangement called a HANS device was readily available. Specifically designed to prevent such injuries, the device was in general use but was not mandated. “Ironhead” – perhaps the only driver on the circuit with an open face helmet – rejected it and became perhaps the fifth such fatality in under a year.

      NASCAR made the device mandatory for its three top series several month later. The total WhatNot Cup driver crash fatality count to date since then – 21 years later – is zero.

      • That’s weird, I seem to remember it being called a freak thing. Of course, it is a weird thing from a common sense rather than statistical perspective, so that may be the trick.
        I recall the reports about the pilot in the DA20 suffering the same result.
        Thanks for the specifics.

        • Freakish because it looked so mild, but was the perfect combo of angles and speeds. Again, sad ‘cuz it was a recognized and mitigatable danger.

          I’d guess that the DA 20 pancaking involved an abrupt deceleration in forward speed, as well as in the vertical. I’m no expert, but I think the occupants would have been subjected to the same head whipping action that seems to have a 67% mortality rate (based upon a sampling of three).

  13. Learning from other people’s experiences. The accident appears to have been caused by an unfortunate sequence of worsening weather conditions aggravated by what may have been, a human behind-the-eight ball “perception of environmental elements and events concerning time or space, the comprehension of their meaning, and the projection of their future status.”
    I wonder if I would have been able to prevent it.

  14. Nice bird. I flew the twin diesel at a dealership in Bozeman about ten years ago. The airplane has a lot going for it, like just about everything. This accident reminds me of the Bearcat that Charlie Hilliard perished in at Lakeland when it too tipped over.

    God bless.

  15. Paul – you have a good bead on reality and write exceedingly well. I see this as a freak accident. I’m curious about how a fatality would occur, but doubtful I’ll get useful information that will make my PA30 safer.

    I can imagine my wife and I in our Twin Comanche at the hold short line. Not terribly surprised that the local storm created all this wind suddenly, but willing to wait it out and depart into clear air. There but for the grace of God go I…….. Frank

  16. re: microbursts. Four summers ago, a friend and I were flying locally in his Lancair 235 to check out some instrumentation problems. It was a typical Virginia summer day, with pop-up thunderstorms developing here and there. I kept an eye on a small cell to our west that seemed to be slowly making its way toward our home airport, and we decided to call it a day.

    After we got the bird secured in the hangar, sure enough it began to rain. The day was so hot and sticky that we left the rolling doors partly open to enjoy the cool air. Without warning, the wind began to pick up, whipping the rain inside the hangar at full force. Within seconds it must have been sixty miles an hour or more. We struggled to roll the doors closed; the wind whipped one door off the track and down toward the airplane. Steve stood there under the end of the derailed door; Atlas, holding it off his Lancair. Almost immediately, a huge gust came into the hangar and stripped the entire roof off the frame – all of it – every piece. I thought this might be all for us.

    But as quick as it came, it was over. The cell passed and we were standing in brilliant sunlight, soaked to the skin, but the airplane was unharmed. A 150 on the field was not so lucky; it had broken its tethers and tumbled end over end about a hundred yards.

    We checked with the National Weather Service, which told us we had almost certainly got an up close look at a microburst. I hope it’s my last.

  17. I’m wondering if the lap belt inertia reel failed to lock in response to a vertical load, allowing the pilot’s head to strike something overhead.