Frontal assault

March 1, 1995

Weather moves in mysterious ways and sometimes it moves fast. Fortunately, airplanes have almost always moved faster, but the pilot must have the will to retreat. When there is an accident, the aircraft is frequently suspect and the legal system gets involved. The process of resolution seldom produces a clear victory. What follows is the National Transportation Safety Board's analysis of a 1988 accident and an opposing viewpoint raised at a resulting trial.

The private pilot had accumulated 256 hours in about five years while progressing through the ranks of Cessna singles to the Cessna 210. This was his first flight in the 210 without an instructor although he had accumulated more than 30 hours of dual 210 time. He was not instrument rated.

At 8:45 a.m., the flight departed Raleigh, North Carolina, for a VFR trip to Washington, D.C. Shortly after takeoff, the pilot obtained VFR flight following from Washington Center. According to Raleigh Flight Watch, the weather did not appear threatening when the pilot called about 45 minutes before the accident. Although the flight was on top of a broken layer of clouds at 9,500 feet, there was some anxiety about avoiding the cloud deck for descent into Washington National. The specialist seemed unconcerned: "...clear skies into the Richmond area with that scattered layer at 9,000, remains scattered through Washington at that altitude, so you should be okay."

The pilot replied that he was getting close to the cloud tops. The specialist then provided the Washington National terminal forecast of 4,000 to 5,000 scattered, with a possible 9,000 broken layer, and suggested that the pilot contact Washington Flight Watch in a few minutes. Just after the 210 returned to the Center frequency another aircraft requested weather for Harrisburg, Pennsylvania, which is about 80 miles north of Washington. The specialist responded with 1,900 scattered, estimated 5,000 broken, and towering cumulus to the east. A front had passed very quickly, dropping visibility to only two miles in blowing snow, followed by rapid clearing. Had the pilot remained on the Flight Watch frequency to hear that information, it might have alerted him that the front was about to change the D.C. area weather drastically.

Everything progressed routinely with a handoff to Washington Approach at 10:02 a.m. EST (1502 Coordinated Universal Time). Descending to 3,500 feet, the pilot requested 2,500 and was cleared down. The following transcript was provided by Washington National Tower. The times of the transmissions are shown in UTC. Abbreviations are as follows: 01N is a Cessna 210, N6401N; Approach is a Washington Approach radar controller; and Dulles is Dulles Tower.

1509:51 01N: "Washington Approach, Zero-One-November; I'm getting into some snow here and low clouds."
1509:55 Approach: "Zero-One-November, roger."
1509:58 Approach: "Zero-One-November would you like an IFR...are you IFR rated?"
1510:00 01N: "Negative."
1510:04 Approach: "Zero-One-November, roger; maintain VFR."
1511:44 01N: "Zero-One-November, I'm reversing my course." [The frontal weather was approaching from the northwest. Once the pilot was inside the clouds a 180-degree turn would have turned the aircraft to the southwest, and carried the flight deeper into the fast-moving clouds. A turn to the east or southeast would probably have been the shortest route to VFR.]
1511:49 Approach: "Zero-One-November, roger. Would you like me to check around to other airports and see what the weather is for you?"
1511:53 01N: "Roger; I need to get out of this."
1511:55 Approach: "All right, Zero-One-November."
1511:59 Approach (to Dulles Tower, landline transmission): "Dulles, Wash 62 line, VFR or IFR."
1511:59 01N: "Zero-One-November, is it clear over your airport?"
1512:08 Dulles (to Washington Approach): "VFR specialist here; go ahead."
1512:10 Approach (to Dulles): "Thank you. I'm just checking. I got a guy stuck in the soup — I love that talk."
1512:13 Approach: "Zero-One-November, Washington weather 5,000." To Dulles: "I'll call you back, Jerry." [It appears that the controller misspoke the ceiling here but corrects it shortly.]
1512:22 Dulles: "OK."
1512:24 Approach: "The Washington weather: 500 scattered, estimated ceiling, 1,500 broken, 3,500 overcast, visibility four, light rain and snow showers."
1512:38 01N: "Roger, clear me to on top at 3,000 — ah — 5,500 I'm going on top."
1512:41 Approach: "OK, Six-Four-Zero-One-November, fly heading 270. Climb and maintain VFR conditions on top. If not on top by 5,500, maintain 5,500 and advise." [The controller apparently did not have a good picture of the weather, since the 270-degree recommended heading was putting the flight right into the front.]
1512:50 01N: "Roger; going up."
1513:30 Approach: "Zero-One-November, what would you like to do now, sir?"
1513:33 01N: "Roger; let me get on top and clear me to the nearest airport that's clear."
1513:43 Approach: "All right, Dulles says they're VFR now, sir."
1513:42 01N: "Roger. Do you have anything private, and I'll wait till this gets over, then come on in to National?"
1513:47 Approach: "I can't really say what the weather is at PG or College Park [general aviation airports in the vicinity], sir."
1513:53 01N: "Roger; let me clear up on top, then I'll check back with you."
1514:00 Approach: "OK. Don't leave the frequency, though. I've got traffic all around you down there."
1514:02 01N: "Roger."
1514:49 Approach (to Dulles Approach): Ten south of Davidson [another local airport], Zero-One-November is climbing VFR to 5,500 feet. He got caught IFR and he's not rated."
1514:51 Dulles Approach (to Washington Approach): "My guy Two-Four-Lima [apparently another VFR aircraft whose pilot chose a loweraltitude] can't stay VFR. He's been going down."
1518:18 01N: "Zero-One-November needs a turn."
1518:20 Approach: "Zero-One-November, roger. Where would you like to go now?"
1518:23 01N: "Just turn me away from this stuff."
1518:24 Approach: "All right, turn right heading 360, sir."
1518:37 Approach (to Dulles): "What's the weather like at Manassas? You have any idea?"
1518:56 Dulles: "It's pretty good, Joe. It's cleared out over here. They can probably get over there VFR if they're at 2,500 or below."
1519:03 Approach (to Dulles): "OK; thanks, Jerry. I'll let you know."
1519:07 Approach: "Zero-One-November, Dulles says it's clear around Manassas, if you'd like to go over toward Manassas."
1519:56 Approach: "Zero-One-November, Washington."

There was no further contact with the airplane, despite repeated attempts.

Recorded radar information showed that at 1507 UTC the aircraft was about 26 miles southwest of Washington National when it descended from 3,500. At about 1512, the aircraft was 16 miles southwest and had descended to 1,300 feet. This was well below the assigned altitude of 2,500. The aircraft went another mile, and at 1513 it reversed course and began climbing. It headed southwest until it was about 25 miles from the airport. The last radar hit showed 4,600 feet near the Marine Corps Air Station at Quantico, Virginia.

The 1520 weather observation at Quantico reported 500 scattered, 1,500 broken, and five miles in light rain and snow, with winds out of the northwest at 20 gusting to 29 knots. From 1517 to 1520, the visibility dropped from 12 to five miles and then to three miles at 1527.

National Weather Service radar records showed a 90-mile-long area of snow showers (oriented southwest to northeast) centered near Quantico. The line was moving southwest at 40 knots, with tops to 13,000 feet.

An airline pilot flying a Boeing 737 had encountered "severe weather conditions" while inbound to Dulles about an hour before the accident. The pilot said he was unable to stay out of the clouds and requested an ILS. Once in the clouds, the air was very turbulent and airspeed varied as much as 15 knots. He broke out over the approach lights at 300 feet agl and reported very heavy snow. "At no time during our approach and landing did approach control, tower, or other aircraft make any mention of these conditions or even the possibility of them," the pilot said. "Within 30 minutes, this storm was gone and Dulles was VFR with scattered clouds. In 32 years of flying, with the exception of thunderstorms, I have never seen such a rapidly developing, moving, and severe weather condition move through an area."

There were at least five witnesses who saw the 210 descend from the base of the clouds in pieces. A flight instructor had just landed at Quantico in a Cessna 172RG. He stated in a witness report, "I elected not to go far from the air base, due to some very ominous-looking clouds that appeared to be approaching from the west northwest...with a base of 1,000 to 2,000 msl (roughly our altitude) with a layer of 'scud' underneath, perhaps 200 to 500 agl. The clouds were very dense, which I think caused their dark color.... The cloud cover to the east of the ominous clouds was scattered to broken at a fairly high altitude (estimated 3,000 to 4,000 or above)."

According to the NTSB, the wreckage was scattered over a one- square-mile area. The fuselage was generally intact and highly compressed. None of the three occupants survived. The rudder, elevator, upper vertical stabilizer, and horizontal stabilizer had separated. Both wings separated from the fuselage. The engine appeared to have been developing power, and the tachometer indicated 2,500 rpm.

The probable cause, determined by the NTSB, is unfortunately a standard litany for VFR pilots who press on into instrument meteorological conditions: "Pilot-in-command — poor in-flight planning/decision, VFR flight into IMC — inadvertent. Aircraft control not maintained, spatial disorientation, exceeded design stress limits of the aircraft."

This was the NTSB's official determination, but the story was far from played out. The pilot's widow sued both Cessna and the FAA. Cessna was charged with building a defective aircraft; and the government, with issuing a defective forecast. The litigation that followed the accident showed a considerable difference of opinion. This is a fascinating, and to many a disturbing, aspect of our system. I believe it ultimately affects the safety of pilots and exerts much influence in the decision making of manufacturers and the government. It's also troubling that this case did not come to trial until 1994, six years after the accident.

When John S. Yodice, AOPA's general counsel, and I visited the court clerk to look at the transcript of the trial, it filled more than 30 volumes and physically stood more than two feet thick. The trial had lasted more than a month and involved multiple expert witnesses on both sides. The closing arguments of the attorneys have been summarized, but I have tried to maintain context.

The plaintiff, the pilot's widow, contended that the Cessna 210 was improperly designed and that the aircraft structure had failed although the aircraft was operated in the flight envelope. Her attorney stated: "...sometime slightly after 1518:43, the Cessna 210 [the pilot] was flying began to come apart in the air. It began to bind at the elevator trim tab assembly. It began to flutter...and it threw itself into a violent pitch because he lost the ability to exert pitch control on the airplane." The plaintiffs had introduced into evidence a list of at least 19 Cessna 210s that had suffered in-flight breakups from October 1970 to December 1986.

The plaintiff's case is built on their theory that flutter is caused by distortion or failure of the trim tab actuator, which allows the tab to vibrate. The distortion is caused by the actuator's coming in contact with a spar in the elevator. The hole through which the actuator passes provides plenty of clearance until the aircraft is in flight. Under certain conditions, where the entire empennage assembly twists slightly, the angle of the actuator rod passing through the hole does not provide adequate clearance, the plaintiffs contended. It doesn't happen in all flight conditions — only under the combination of certain loads, centers of gravity, and airspeeds. The contention was that Cessna had improperly tested the aircraft and had known about the problem as far back as 1963.

Cessna's attorney responded: "...In terms of trim tabs and pushrods, there is some compelling evidence...that got this airplane certified by the FAA as airworthy, and that [Cessna's expert witness] had physically conducted tests, not only static tests...but...[flight] tested the very conditions that are...[contested] today....

"[Cessna's expert] took that condition...and flew the airplane to V NE and showed you in a film that was like watching the grass grow...nothing happened to that trim tab or that elevator or the pushrod.... If the plaintiffs are right, [Cessna's expert] and [another engineer], who flew the airplane on the film you saw yesterday, would not be here. They would be victims of flutter."

The plaintiff had to discount spatial disorientation to absolve the pilot of responsibility for the in-flight breakup.

The plaintiff's attorney: "We heard the most extraordinary testimony in this case because what Cessna needs is [a Cessna expert witness] recipe — low-time VFR pilot plus cloud equals spatial disorientation.... But darn, don't we have a problem in this case because we look at the track and the track is perfectly straight [referring to the radar track by ATC].... We want it to look like that pilot can't fly in the clouds...and he's weaving around, he's out of control and then he crashes. That's what they [Cessna] want the track to look like...."

The plaintiffs had an expert witness who stated that a typical loss-of-control accident produces a radar plot that wanders in many directions before the final spiral develops, which is not what this track showed.

At the conclusion of the trial the jury found Cessna guilty of improper design and awarded the plaintiff $5 million. The FAA and NTSB have not issued any airworthiness directives relating to Cessna 210 trim tab flutter to date. The case is being appealed now. What would your verdict have been?

If anything is obvious, it is that the emotional aspects of this case are intertwined with the technical. The financial concerns of the family and the manufacturer are a separate issue from the realities of aircraft design and certification. Over the past two decades we have seen juries with little or no aeronautical expertise pass negative judgment on an aircraft certification system that has delivered some of the finest flying machines ever designed.

It sounds reasonable to have a jury hear the case, with experts from both sides providing evidence to prove their case. The reality is not that pure. The government investigation is either vindicated or vilified, depending on the point of view. Both sides hire experts who are paid a lot to support a plausible mechanical failure theory or to get the manufacturer off the hook. Objectivity on both sides has occasionally been swayed in such circumstances.

Most juries are ignorant about aviation; and I question whether they would qualify as peers, in the strict definition of the word, when a pilot's judgment is in question. They are subject to emotional persuasion, swayed by personality and circumstance, and confronted with highly confusing technical testimony. Widows and children generally fare well against large corporations, which is society's way of taking care of the weak and the wronged.

It is high time that all sides agreed on responsibility. The FAA is responsible for the certification and safety in aircraft design, not the courts. I really have to question why the FAA hasn't challenged this design-by-jury process. If there is a problem with an aircraft, then it should be promptly dealt with through the appropriate recall systems. We shouldn't be waiting six years to find out that there's a defect — if there is one.

And we, as pilots, are responsible for operating within our limits and those of our aircraft. The penalties of gravity relating to ignorance or incompetence are severe. You may not consider that fair, but anyone who flies should have no illusions.

Our present system is billed as an unbiased analysis of the facts. It just happens that millions of dollars are hanging in the balance. F. Lee Bailey, the trial attorney, recently put this into perspective when he said, "The God of common law is not justice but consistency. If we shaft you, we will shaft your neighbor." Profound thought.

We are using the legal system to deal with both a social and technical problem. It is the wrong tool. The insurance and punitive damage aspect must be separated from the technical if we are to have unprejudiced reviews, and ultimately, safer aircraft.

Before the opposition unlimbers its weapons, let me say that the present accident investigation system is far from perfect. It is up to the manufacturers and the government authorities to carefully analyze every serious accident where there is any specter of improper design. While most people in the business have the highest integrity, there is at least a temptation on the part of the manufacturer and its certifying authority, the FAA, to protect past judgments and designs. Occasionally a defect sneaks through; and until it surfaces enough times, innocent pilots are unfairly charged with responsibility.

It is also true that the FAA and NTSB do not always have the time, money, expertise, equipment — and occasionally, the inclination — to do a thorough investigation. This is more applicable to general aviation than airline accidents. We should admit that, once in a while, there are poorly designed aircraft and equipment. Plaintiffs are entitled to recover their losses. Manufacturers are responsible for fixing the problem for the entire fleet, not just a single plaintiff.

Special certification reviews were conducted for the Piper Malibu, V-tail Beech Bonanza, Learjet 20 series, and Mitsubishi MU-2. Enough accidents occurred under questionable circumstances to persuade the FAA and NTSB to scrutinize the aircraft more closely, to be sure that nothing had been overlooked. If the certification process fails, the FAA is accountable to fix it. In most reviews the aircraft met or exceeded the certification standards, but sometimes a modification was required. That is how the system should work.

To aid investigations, we might equip aircraft with inexpensive cockpit voice recorders that replay the last 10 minutes of flight. Technically, that would be relatively easy. Perhaps a basic flight data recorder would also help to limit the gargantuan expense and creative speculation that surround these cases. That's more feasible in new aircraft, where electronic sensors could be wired in from the beginning. These devices, which won't solve all accidents, are a two-edged sword that could point to the pilot or the manufacturer.

Here's one suggestion on how to deal with the technicalities: The FAA and NTSB should devote the proper time, money, and personnel to the job. A non-government oversight board of aeronautical personnel — pilots, engineers, maintenance technicians, air traffic control experts, and possibly attorneys might review the process. Some retirees might be just the ticket. No one on the committee should have ties to any of the players. They would determine if the investigation was properly conducted. No punches pulled and no emotion — just the facts.

How much would all this cost? Possibly a good bit less than what we're paying now. We need a more equitable way of distributing this inevitable cost. Once a thorough and neutral investigation has been completed, then allow the legal system to redress grievances. Many of these cases may still go to trial, but I suspect that a lot more of them would be resolved with settlements — and at less cost than the current design-by-jury method.


See also the index of "Safety Pilot" articles, organized by subject. Bruce Landsberg is executive director of the AOPA Air Safety Foundation.