By Sherman E Hall
Although this incident took place almost forty years ago, it will resonate with pilots everywhere, and particularly with the intrepid owners and operators of the classic aircraft that continues to feature prominently on the Australian aircraft register.
As a lucky pilot who has survived a Piper Seneca in-flight break-up I would like to offer some thoughts that might help make aircraft owners aware of the potentially serious problems of trim tab free play.
The flutter testing flights for the Piper PA-34 N33589 were not felt to be very dangerous because of our previous success with the Seneca. In fact, I had just completed a series of flutter flights on a Seneca II equipped with Robertson spoilers and tip tanks. The accelerometer traces were good, but the camera that recorded the speeds and altitude malfunctioned. The owner was impatient and wanted his aircraft, the FAA wouldn’t accept our data without speed verifications, and by the time the bad film had been developed the aircraft was reconfigured and ready for delivery. The FAA allowed us to deliver the aircraft with some restrictions, including that flutter testing still needed to be completed.
The intended test was to climb to 25,000 feet, with camera and recorders on and to pulse each control at several predetermined speeds and at Vd (10 per cent past red line). The first flights (on December 24th 1976) were aborted because low temperature (-45 degrees F) at 25,000 feet froze the elevator jackscrew and only low-speed data could be collected. The heavy grease was removed from the jackscrew and a light silicone spray was used instead.
Flight tests on December 27th appeared to be going well. I climbed to 25,000 feet on an IFR clearance above a 4000 to 6000 foot broken cloud cover. The test equipment turned on and trim function was good. Intermediate speeds looked good and at Vd (250 IAS), at approximately 20,000 feet, I completed the following control pulses: aileron (spoiler-equipped) left and right, rudder left and right, elevator down, and then I reached to start the up-elevator pulse. The aircraft was in trim and my left hand was holding the wheel lightly. The air was smooth and there had been no vibration.
It felt like an explosion; the elevator failure resulted in an extremely violent nose-down tumble. Both wings tore off, firstly outboard of the engine nacelle, then again at the fuselage. The fuselage was ruptured, parts of the tail section and rear door were gone, the fibreglass nose section was gone and the windshields were blown out. Negative Gs had me pinned against the cabin ceiling, the violent shaking had torn off my helmet, and I had to protect my head with my hands to keep from being knocked out.
The survival instinct had taken over. I knew I had to get out of there to stay alive, but I couldn’t move and couldn’t do anything except try to protect myself against the beating I was taking. As abruptly as it started, the violence stopped. It was smooth and calm, with a sense of slow rolling or tumbling. The cold air (-40 degrees F) was blowing through the hulk of the fuselage. Seat belt off and out through the cabin door. I opened my chute immediately, which could have been fatal. As my chute opened, an engine with a section of wing dropped past me. I can still clearly see that engine in my mind. The prop was feathered and the engine was running. The total sequence of failure and getting out only took seconds, but it felt like minutes.
I was awed by the destructive disintegration of that aircraft and my miraculous survival. I was now floating down at 20,000 feet or so, with small light pieces of aluminum, insulation and fabric all around me. In whatever direction I looked I could see confetti.
This was December and I was over a cloud layer somewhere above the Cascade Mountains of Washington. I really wasn’t dressed appropriately and again my thoughts turned to survival. Luck truly was on my side that day. Wear some torn clothes, a parachute under your left arm, put up your thumb and you get a ride every time.
This was the first aircraft lost by Robertson in flutter testing. The FAA took a harder stand for more testing and complete evaluation by Robertson for flutter clearance. Prior to this loss, Robertson inspected and conformed only its modification and assumed that an airworthy aircraft would have all controls balanced and free play within limits. N33589 did not have the elevator tab free play measured before that test, but I firmly believe it was past its limits. On the walk-around inspection I commented to the company inspector about the apparent amount of free play and we agreed it was a common thing on Senecas. Hindsight is 20/20, and I know where I should have stopped that flight test.
Some time later, Robertson elected to continue the certification of the tip tanks on the Seneca II. The FAA insisted that we place the trim tabs at the maximum limit. Robertson was not evaluating the tail, but was looking for differences in wing response with the tip tanks. They lost the argument, and we instrumented the tail and went to maximum free play on the tab. But wait: when the inspector checked the original tab free play, he found it considerably past the limit and we had to build new bushings to tighten up the free play to the maximum allowed. The flight testing was conducted as cautiously as possible, with several flights of increasing speed and the data reviewed between them. I had declined to do this test flight so a consultant pilot was hired.
On one flight near Vd a problem was encountered that nearly caused the loss of this aircraft. The cabin door popped open at 225 KIAS and the pilot thought the aircraft was coming apart. The accelerometer couldn’t be read because it went off scale. The door problem couldn’t be solved satisfactorily and the pilot insisted that the front door should be taped shut with aluminum tape, leaving his only exit through the rear cabin. The test was successfully completed and proved the aircraft is flutter-free when within limits. The consulting pilot still believes the door problem could have caused the tail failure of N33589. Interestingly, Piper redesigned the door latch on later models.
I still believe the Seneca is one of the best light twins on the market. Most pilots and maintenance facilities do not understand the consequences of improperly balanced controls and excessive tab free play. Robertson Corporation corrected many of these problems during R/Stol modifications, but the owners were usually unhappy at the added cost. An aircraft with controls out of limits cannot be returned to service and must be grounded until corrected.
Pilots and mechanics may think of flutter as a continuing vibration, but divergent flutter should be thought of as an explosion. If the conditions are correct, it is as dangerous as a lighted stick of dynamite.
My recommendation for the owner of any aircraft is to ensure that your aircraft is maintained properly. All control surfaces can be checked on a simple pre-flight. Free play limits are design limits, with no margin of safety if they are exceeded. Replace bushings or install high-tolerance bolts to stay within these limits. An AD is not issued unless the airworthiness is in question, and this means an unsafe condition exists. You might be the first person to discover this unsafe condition, and the consequences could be dire.
The requirement to inspect the rudder servo trim tab is found in AD/PA349 and FAA AD 73-13-01.
Reprinted with permission from Aviation Safety, April 1982