In the blink of an eye, Jimmy Leeward’s P-51 Galloping Ghost went from rounding the last turn at National Championship Air Races in Reno to sharply pitching upward, rolling over, and then diving straight down into box seats full of spectators. Strange as it is to say, there are reasons to be grateful — had his airplane hit the grandstands, there would have been hundreds, not dozens, of injuries.

Air shows are a big deal in America — attendance is around 17 million visitors to the 400 or so air shows each year, roughly around the same attendance as the NFL — and, apart from the causal attendees, there’s a sense of community around types of planes, types of shows, and locations. They know the history of the sport; the Ramstein disaster, for one, still lingers in the minds of many in the air show community, and the frequency of fatal and near-fatal crashes is not lost on anyone. The casualty numbers are lower, but they still look more like a major air disaster than a simple crash; Reno will likely take a similar place to Ramstein in the minds of the air show community, and may end the National Championship Air Races, at least in their current form.

As always, when a crash happens, the media attention shifts quickly to the National Transportation Safety Board’s “Go Team” investigation. Maybe it’s something about the allure of governmental rapid response teams, or maybe it’s the idea that, with an investigation and findings will come some sort of closure. The NTSB is a good organization with talented and dedicated personnel, and it’s no stretch to say that NTSB Aviation Accident Reports and other recommendations have saved countless lives, but one thing needs to be understood about the NTSB.

Coincidentally, a few hours before the crash I conducted the deposition of the former fleet operations manager for a company involved in a fatal maritime accident. The NTSB Marine Accident Report recommended that her company “review existing safety management program and develop improved means to ensure that your company’s safety and emergency procedures are understood and adhered to by employees in safety-critical positions.” The fleet operations manager argued that the NTSB’s finding that her company had at all contributed to the accident was merely “political.”

In one sense, she’s right: the NTSB’s findings are “political,” in that they are made by the government for the benefit of everyone, rather than made for accountability among those involved in the crash. The NTSB reviews accidents primarily for the purpose of making recommendations for the future and secondarily for determining fault. In contrast, civil litigation exists to determine who should pay for the losses arising from an injury, and thus cases are reviewed by the judicial system primarily for determining fault.

This difference in focus isn’t just a matter of word choice. There’s an entire field of ‘root cause analysis’ that assesses the way in which accidents and other failures are investigated. Its lessons have been applied to aircraft safety as well, including in the federal regulations governing military aircraft safety, which direct audits towards the cause, not just the symptom, of safety deficiencies.

That’s not to say the NTSB’s process is flawed or that their conclusions are wrong (although it’s always disturbing to me how the “party system” always gives the likely culpable parties a seat at the table but never gives any voice to the victims). It’s just important to understand that they answer a different question — what can we, as a government agency, recommend to prevent this in the future? — from the question asked in a lawsuit: who, if anyone, was responsible?

Tim OBrien Photo - P51 Loss Of Trim TabInitial reports have focused on the trim tab of the Galloping Ghost. A remarkable photo just before the crash taken by Tim O’Brien, himself an air show organizer, shows the plane missing one of its left side trim tab entirely. Still images from video taken of the crash show the trim tab in the process of falling off.

It wouldn’t be surprising if the flutter caused the trim tab to break off. (For those unfamiliar with flutter, Mike Danko dug up an old NASA video of trim tab flutter). That’s a known problem with P-51s; consider this report regarding the P-51 Voodoo Chile at the Reno National Championship Air Races just a couple years ago:

… Voodoo very abruptly pulled up; however, Hannah didn’t radio a distress call. … Steve Hinton flew over to take a look Voodoo. “You OK Bob?” called Hinton. “Yea, this thing just popped big time,” replied Hannah. What Hannah didn’t mention is that the g-load from the quick pull-up had caused him to black out. He finally managed to reach the throttle and reduced Voodoo’s power. At that point Hannah radioed that he “(wasn’t) out of it yet,” but he wasn’t thinking clearly. Later, he declared a mayday and made a perfect landing. … On the ground one could see what cause Voodoo’s problems during the race. The left elevator torque tube failed when the elevator trim fluttered and departed the plane. Fortunately, Bob Hannah’s skill and coolness in the cockpit saved day.

When the trim tab fell off Voodoo, the plane shot upwards and the 10G deceleration force caused Bob Hannah to black out entirely. That’s just as you would expect: the faster you go, the more the plane points upwards on its own, and the more you need to point the nose down to trim the airplane. Thus, at speed and level, the trim tab points up relative to the airflow over the elevator, causing the elevator to be deflected slightly down to maintain level flight.

At over 500 miles per hour, there are enormous airloads on the elevator trim tab to keep the elevator in a position that allows the pilot to maintain control, making damage to the trim tab more likely. Remove the trim tab and the non-trimmed elevator settings immediately deflect up, just like when pulling the stick back hard. That’s what causes the abrupt climb (and corresponding loss of consciousness) when the trim tab falls off.

Hannah regained consciousness at 9,000 feet and, as you can tell from the above, took some time to come back to his senses. You can see pictures of the damage here. It was even the same trim tab. The difference between Voodoo’s close call and Galloping Ghost’s tragedy may have been pure, dumb luck: Voodoo didn’t roll after losing the trim tab while Galloping Ghost did.

But that doesn’t necessarily mean flutter caused the trim tab to dislodge, or that the trim tab was the cause of the accident, or that the trim tab was the only cause of the accident. It’s quite possible something else caused the Galloping Ghost to climb rapidly, and in that process the flutter developed or the trim tab was damaged. As has been reported, some members of the crowd noticed “a strange gurgling engine noise” before Galloping Ghost pitched upwards. Further, as discussed below, it’s possible the trim tab failure could have been avoided, and more could have been done — such as ensuring the pilot was harnessed properly and plotting the race further from the stands — to prevent this tragedy.

So, where might liability fall? I often say that fatal maritime and aviation accidents rarely happen as the result of a single, unlikely event. Usually, they’re caused by a cascade of failure.

First, pilot error. There was initially some chatter about the propriety of allowing an “80 year old man” (Leeward was really 74, but “80” somehow ended up being widely repeated) to fly in an air race, but he was among the more highly qualified pilots in the country — including having a third class medical certification certification as of March 2010, which was still valid as of the accident. The question, then, is not if the organizers or other third parties appropriately evaluated his health — he was certified — but rather if he or anyone involved with his flight preparation recognized any physical disabilities that arose after March 2010 and which could have impaired his ability to fly. Of course, his health is irrelevant if he never became disabled and there’s some other explanation for why his aircraft suddenly climbed, rolled, and dove.

There might be other pilot error, though. Planes aren’t like cars; improper control settings can cause damage to the plane even in the absence of a defect or a collision. Leeward could have been using excessive trim inputs to assist with the turns. It also appears that Leeward didn’t lock his shoulder harness to stop him from slumping forward, which is why you can barely see him in the plane. After the abrupt turn upwards and corresponding deceleration, his torso probably contacted the stick causing the plane to roll (unlike Bob Hannah, who had a locked harness [Update: See Brad Haskin’s comment about this below.]). The videos show him hitting the ground at full power, so he was not conscious.

Second, defective parts, an unsafe design, or inadequate maintenance. A broken trim tab is obviously a problem, and it’s a known weakness in P-51s. Was the plane properly inspected and maintained? Here, there are questions of method and timing. FAR 43.13 requires:

Each person performing maintenance, alteration, or preventive maintenance on an aircraft, engine, propeller, or appliance shall use the methods, techniques, and practices prescribed in the current manufacturer’s maintenance manual or Instructions for Continued Airworthiness prepared by its manufacturer, or other methods, techniques, and practices acceptable to the Administrator, except as noted in §43.16. He shall use the tools, equipment, and test apparatus necessary to assure completion of the work in accordance with accepted industry practices. If special equipment or test apparatus is recommended by the manufacturer involved, he must use that equipment or apparatus or its equivalent acceptable to the Administrator.

As Mike Busch summed it up, “The key to understanding FAR 43.13 is the phrase ‘methods, techniques and practices.’ That phrase refers to how to do something, not when to do something.” There’s thus not as much federal law for when maintenance should be done, but industry standards — and the general duty of reasonable under negligence law — demand it be done with some frequency. I’d be surprised if someone of Leeward’s stature didn’t have his planes inspected very frequently, but, in investigating these claims, you often never know what you find. We’ve seen aircraft parts sold as “new” that turned out to be patched up parts from the WWII-era.

Moreover, it bears mention that Galloping Ghost had been heavily modified to make it faster. Can a P-51 be modified to the point that it is so fast, and creates such force and stress, that it becomes inevitable that the elevator trim tab breaks off? Absolutely, and it would be negligent on the part of the owner, designer, and builder to fly such a plane at racing speeds above a dense crowd of spectators. [Update: See the comment about experimental planes below.]

Third, flight planning. Whenever a plane crashes into a populated area, the choice of route comes under scrutiny, as does the possibility of airport controller error. All low-flying planes create a danger of impact (consider, for example, the Ramstein disaster) and the National Championship Air Races has had more than its fair share of fatal crashes and near-misses like Voodoo. It is no stretch to say that it was only a matter of time before one of those accidents ended up happening in the stands. Sure, it’s thrilling for the crowd to be up close to the planes as they race, and unsurprising that the organizers would set up the race that way [Update: see TCinLA’s comment below, it wasn’t always that way, spectators used to be prohibited from even stepping foot on the tarmac, much less having seats there], but the companies making money off the event are charged by law with ensuring the safety of those spectators, including by toning down the thrills just a bit if it makes everyone safer.

So now what? Nine people are dead and dozens are injured, many seriously, many permanently. Lives lost, contributions to families lost, wages lost, and medical care needed. Some tort reformers have complained that, even though most accidents are simple pilot error, every airplane crash results in a flurry of litigation against airplane manufacturers, event organizers, and plane owners and mechanics.

There’s a reason for that: airplane crashes cause a lot of damage and are rarely a fluke or understandable mistake. They are one of the classic types of situations that our civil litigation system was designed to address.

A rule of thumb in aircraft disaster litigation is: name all involved parties in the complaint. It’s unfortunate, but in general plaintiffs need to sue everyone — from the City that hosted the event, to the organizers of the show, to the owners of the field, to the owner of the plane, the mechanics for the plane, and the pilot of the plane — to ensure that a plaintiff doesn’t learn, after the statute of limitations has expired, that they sued the wrong party. Defendants and insurance companies are more than happy to play games with shell entities and business arrangements, and there’s no way to know, pre-suit, what the contracts between the parties look like or how a court will rule on their various relationships. (Here’s one air show crash case, from California, in which the plaintiffs won against the company and City that managed the airport in front of the trial court only to have that entire part of their case thrown out on appeal.)

There’s no reason why an injured party can’t file a lawsuit before the NTSB investigation concludes — indeed, they usually have to, given how long the investigation takes — but typically those cases take a back seat to the initial interviews and damage assessment. Whatever the NTSB and legal outcome, this crash may spell the end of the National Championship Air Races, at least in their current form. Maybe that’s for the best; if they’re not run in a safe manner, they shouldn’t be run at all. Or maybe the solution is something as simple as ensuring planes conform to safe designs, mandating the use of proper harnesses, and moving the race a little bit further from the crowd.

[Note: I already allowed through a couple comments of the lawyers-are-scum variety, and I replied in turn. If you’re about to leave a comment that plaintiffs’ lawyers are ruining America or the like, then don’t bother. We have enough already. Words are cheap; lifelong medical care or the death of a spouse or parent, not so much.]