DC-8 and the 737 MAX

Bob Bogash

Home Safety Index Reader Comments Contact

Some thoughts on MCAS and the 737 MAX

In late 2018 and early 2019, the new Boeing 737 MAX suffered two bad accidents in which the airplane dove into the ground at high speed from a low altitude shortly after departure. The accidents were ascribed to a new automatic system called MCAS driving the horizontal stabilizer into the full airplane nosedown (AND) position. The pilots were unable to diagnose the problem - in part because they had not been informed of its existance on the airplane, nor were they trained on its behavior - and lost control of the airplane.

These two accidents had an eerie similarity with two or three accidents that had occurred about 55 years earlier, when the jet age was still in its infancy. The reactions to these two sets of accidents is demonstrative of the differences in communications, media, and public opinion over a half century of time.

On 29 Nov 1963, a Trans-Canada DC-8 took off from Montreal bound for Toronto. The climbout was conducted at night in IMC conditions. Five minutes after departure, it struck the ground at a speed of 470-485 kts and an angle of about 55 degrees. The aircraft had no FDR or CVR and so the Cause was determined via analysis and probabilities. The most likely cause was determined to be malfunction of the Pitch Trim Compensator System (PTC), which, combined with the aircraft’s CG, resulted in operation in an area of reduced or poor pitch stability. The aircraft crashed with the horizontal stab in the full (2 degree) AND position (Airplane Nose Down.)

The accident airplane, CF-TJN, a few months prior to the loss
Ken Fielding - via Creative Commons

Trans-Canada Accident Report here

During 1963, several other DC-8 incidents had occurred involving apparent failure of the pitch control and/or PTC system and consequent pitch instability. One of those occurred on an Eastern DC-8 (N8607) climbing out of Washington-Dulles 20 Aug 1963. In that incident, the stab was manually trimmed nearly full AND, causing an upset which resulted in a 40 degree nose down condition and required 13,000 ft to recover (which took place only a little more than 1000 ft above the ground.) That aircraft had encountered severe turbulence during climbout.

On 25 Feb 1964, an Eastern DC-8 (N8607 - the same airplane and the same pilot flying as Dulles incident), crashed about five minutes after departure from New Orleans. The airplane had no CVR and the FDR had ceased operating some time prior to the accident flight, but analysis showed that the aircraft had struck the surface of Lake Pontchartrain at a steep angle and high speed. The flight was in IMC and at night with convective activity in the area. The stab was found in the full 2 degree AND position. Investigation indicated continuing problems with the PTC system on this airplane.

These two accidents, just 3 months apart, and both involving high speed dives into the ground during early climbout, bear an uncanny resemblance to the two 737 MAX accidents. In all four accidents - although 55 years apart - the stab was found in the full AND position. Further, all four accidents involved automatic systems that were designed to resolve pitch control problems experienced in certain parts of the flight envelope (and in certain CG conditions.)

In the case of the DC-8 PTC, the system was designed to resolve the Mach Tuck problem experienced at high speed on large swept wing jet airplanes, caused by the Center of Lift moving aft as speed increases (called the Mach Trim system at Boeing.) In the case of the DC-8, a computer sensed dynamic air pressure and applied varying amounts of up elevator. The system is designed to operate from M 0.80 to M 0.88 applying 34 lbs of up elevator to the co-pilot’s control column at M 0.88.

The PTC system was problematic on DC-8s at the time, and had a history of extending incorrectly and remaining stuck in the extended position. Pilots unaware of the stuck PTC would apply stab AND trim to counteract the PTC’s Up elevator. This created the problem of a “jack-knifed” stab, with the stab Airplane Nose Down and the Elev Airplane Nose Up. In that contest, the larger stab always wins. See my 10 year old White paper on this subject here: http://www.rbogash.com/Safety/autopilots.html

The problem uncovered with the DC-8 was that, although intended for high speed operation, a stuck PTC would have previously unknown and unexplored serious impacts on the pitch and speed stability of the airplane during low speed operations. During flight tests conducted by FAA test pilot Richard Sliff, with the stab at 2 degrees AND and at 220 kts, sharp reversals occurred in the airplane’s pitch stability The instability was such that a pilot could pull 1.5g without feeling any resistance, and 2.5g with only 14 lbs pull force (on a system spec’d with a 6 lb friction.) (His report, and recommendations, were submitted to FAA Western Region and Douglas, whose comments were that “it was not a serious control problem.”) (See the Trans-Canada Accident Report.) This was after the DC-8 stab trim was restricted to 0.5 degree AND from its previous limit of 2.0 deg AND.

The following is quoted from Mr. Sliff's report:

"Because of the problems with airplane
N6571C another airplane was substituted to not only
check the validity of the findings on airplane N6561C
but to also continue with the testing . The problem
with this second airplane was that it was a modified
DC-8 called the 4% leading edge model powered b y
P& W JT3D engines . This airplane to be equivalent
to the standard wing airplane would have to be loaded
to a center of gravity 2% further aft . For maximum
rear C .G . this would require that the airplane be at
34% MAC.

In the interim while awaiting loading, a
flight was made on this airplane at a nominal C .G . of
26% MAC . The purpose of this flight was to check PTC
malfunctions against those experienced in the other airplanes .
The result was that there was not an appreciable difference,
however, a very interesting side benefit came out of this flight.

This area was during manoeuvering with a fully extended PTC at a velocity
of approximately 220 knots and the airplane trimmed
to its previous extreme of full nose down (2 .0° AND) .
It was observed that any attempt at manoeuvering the
airplane with the elevator system resulted in sharp
reversals in the airplane's manoeuvering stability .
This would be another strong reason for limiting the
airplane nose down stabilizer travel . These findings
were conveyed to the Western Region for their further
testing with the,Douglas Company . They responded .at
a later date that they had confirmed my "findings and
felt that with the restricted stabilizer, it was not
a serious control problem."

In the case of the Montreal accident, the turbulence encountered would have been sufficient to displace the aircraft from its neutral pitch trim point leading to divergent oscillations, “eventually assuming a dive attitude.”

Among other findings uncovered during the flight tests were the stalling of the stab trim motors and the inability to hand trim the stab at high speeds (a situation also explored in the NWA 720B dive into the Everglades (12 Feb 1963) – a somewhat similar accident involving turbulence, convective activity, and pilot trimming to full AND. During the resulting dive, the crew were unable to trim the stab ANU due to the high air loads. This same situation occurred during the 737 MAX accident scenarios and is one of the items apparently causing delay to the MCAS fix adoption. (In fact, Boeing used to teach the procedure for unloading the stab allowing opposite manual trimming – a not-so-easy application of pitch in the down direction followed by quick cranking on the trim wheels. This procedure used to be taught but was dropped for some reason over the years.)

The EAL accident report makes very interesting reading, and although a bit long, is worthy of the time, given today’s hyper mode with regard to the MAX. One very interesting section deals with the shallow dive angles and considerable altitude needed to recover from a pitchover. Specifically, the report states that “recovery becomes problematic if a pushover to 30 degrees is initiated at any altitude below 5000 ft” (the Dulles airplane pushed over 40 degrees and required 13,000 ft altitude to recover.) The two MAX accident airplanes had pitchovers associated with the MCAS system driving the stab to full AND at altitudes in the 5000 ft range; their speeds increased dramatically and their stab/elev’s became quickly jack-knifed while the stab air loads precluded manual re-trimming. It didn't take long for the airplanes involved to become unrecoverable with the altitude available.

EAL Accident Report is here.

Also, of interest, is that despite these accidents, the DC-8 was not grounded (nor was the Electra after two wing-loss accidents in the same time period.) In fact, adding the NWA 720B accident, there were three mysterious jet transport high speed fatal dives in just a 1 year period. There were no FBI investigations, congressional hearings, or media frenzy. Nobody at Douglas got fired - least of all the CEO. People apparently had a different mindset a way back then.

Post Script

My readers are eagle-eyed tech types who peruse every word. One of them recalled a Pilot Report from 1961 - almost 2 years before these accidents - and was even able to retrieve it! That's a write-up almost 60 years old!!! This incident occurred in 1960 and was reported in January 1961.

“The captain was having some trouble trimming the aircraft pitch wise. because it appeared to ‘change balance' just after he had trimmed it out. As we had 103 passengers on board. we thought large groups of them were wandering about the cabin. At 16,000 feet the autopilot was engaged. At 18,000 feet feeble turbulence was encountered. At 19,000 feet the aircraft suddenly assumed a very nose high attitude.

The auto pilot was disengaged, and although maximum forward trim was applied. a firm forward positive pressure was necessary to keep the nose from climbing up. “First reaction was a runaway trim motor. However, as the trim was contrary to aircraft tendency change, the reaction changed to one of jammed stabilizer. Shortly thereafter we thought of a change of location of freight caused by turbulence, severely aggravating the 30.6% MAC with which we had started, in fact to such a degree that even maximum trim would not compensate for it.

“Later on the engineer noticed that two circuit breakers of the pitch trim compensator had popped out. The pitch trim compensator override was selected, and immediately the trim forces on the stabilizer returned to more or less normal. and a normal trim condition was selected. “An examination of the pitch trim compensator computer followed, and a large puddle of coffee was found on the floor by the captain's briefcase. “Apparently a lot of coffee had been spilt in this area, before the aircraft arrived at______. and had been slowly penetrating into the computer base. When the pitch trim compensator was checked before takeoff it was working properly, but about 10 minutes after takeoff it started working erratically for a short time and finally shorted completely in the fully extended position.”

I've always been of the opinion that "Incidents" are the most valuable gold nuggets out there - if we would just pick them up. Airplanes are always talking to us, and it's up to us to be listening. If we ignore their messages long enough, they will rear up and bite us in the ass. Prior to AF447, there had been about 36 prior pitot freezing incidents - some very serious - without anyone taking them very seriously. So, the airplane dove into the Atlantic, asking "Are you listening now???" It seems we only get serious about checking things out after a lot of people get killed. Aviation safety seems to need to be written in blood.

In this case, one can only wonder if we had followed up on the report with some sort of diligence whether a few horrific crashes could have been prevented. Then again, the real reason we don't pay attention to incidents (other than "they made it, didn't they?" - like the Lion Air MAX flight immediately prior to the crash flight) is that you can't count accidents that never happened.

Bob Bogash

Return to Home Page