Tuesday, March 23, 1999
Attorney Specializing in Aviation: NTSB Investigation of the Boeing 737 Took Longer Than it Did to Conceive, Design, Certify, Produce And Field the Aircraft
In response to the NTSB's March 23, 1999, hearings on the USAir 427 accident, Attorney Arthur Alan Wolk, who has focused his efforts in the field of air crash litigation, issued the following opinion:
The National Transportation Safety Board, which is about to announce its long awaited conclusions on the causes of the crash of USAir 427, which occurred on September 8, 1994, and United 585, which occurred on March 3, 1991, took longer to complete its investigation of these two accidents than it did for Boeing to conceive, design, certify, produce, and field the Boeing 737.
Not only did it take too long, but from statements issued by both the NTSB and the Federal Aviation Administration, it is clear that both of those agencies of government still don't have a clue either about why the rudder of the Boeing 737 has a mind of its own or the means to fix it. In short, eight years after 25 people were killed in United 585, five years after 132 people were killed in USAir 427, and six years after more than 100 people were killed in Copa 201, the Boeing 737 still is not fixed.
The FAA, in the face of pending recommendations that a dual rudder actuating system be installed in all existing Boeing 737s to prevent the single actuator causing a crash, has said, "There is no data to suggest planes with dual power control units are less prone to an in-flight upset than the 737." It also pointed out that a 737 "is equipped with a standby rudder system that serves a similar purpose." This statement demonstrates that the Federal Aviation Administration, the agency of government responsible for ensuring the safety of flight and for certifying the Boeing 737, still doesn't understand how the rudder control system of that aircraft works. The standby rudder actuator on a Boeing 737 has absolutely no role to play in serving to prevent a rudder hardover caused by an errant rudder power control unit. In fact, because of defects in the standby rudder actuator, it can actually make the situation worse. These statements, then, from the FAA are even more frightening in the face of this lengthy investigation and at least three accidents taking hundreds of lives. It means that the FAA still doesn't have a clue about the fundamental and basic operation of an aircraft that it certified.
The problem is very simple. The Boeing 737 is the only transport category airplane that has a single actuator for the rudder, which is in violation of the federal regulations which require redundancy. Boeing got around the redundancy requirement by performing an analysis, without flight tests, that established that the chance of failure was so remote, it didn't have to have a redundant design. The FAA went along with it, and it was wrong. In fact, the FAA and Boeing admit that the Failure Modes and Effects Analysis performed to obtain that certification was wrong. Therefore, certification was wrong. Therefore, the airplane should have been grounded or fixed. It is neither.
In the face of overwhelming evidence that the rudder was the cause of at least three unsolved Boeing 737 rollover accidents in the last decade, the FAA recently granted certification to three new 737 models, the -600, -700 and -800 Series. In a stunning abdication of its responsibility to ensure safety, the FAA granted that certification with, yes, a single rudder actuator in each of those aircraft. In short, the advanced model Boeing 737 being produced at the rate of 24 a month continue to have the same fundamental flaw that the 3,000 existing 737s have -- no redundancy in the rudder control system.
The FAA also claims that now pilots are being trained to handle the problem of a runaway rudder. This also demonstrates that the FAA is out of the loop and out of touch with the reality of airline pilot training. The advanced maneuvers training (or upset training, as it has been called by the pilots) presumes that the rudder malfunction no longer exists. In other words, the rudder, which may have gone to full travel and caused the airplane to be upset in the first place, has now returned to its normal position. Unfortunately, when the rudder fails in a Boeing 737, it stays there, so all the upset training and all the advanced maneuvers training in the world won't protect the crew and passengers of a Boeing 737 with an errant rudder.
The FAA and the NTSB also laud the changes that were recommended by the NTSB and implemented in Boeing 737 rudder control systems. These included an inspection to determine that adverse tolerances in the manufacture of the servo valve in the rudder could not stack up and cause a malfunction, changes in the design of the servo valve purportedly to prevent rudder malfunction, and the addition of a rudder limiter that would keep the rudder from going full travel and, therefore, allow the pilots to regain control using the ailerons on the wings, which could overpower the rudder. Unfortunately, none of those changes either has or will work to prevent an accident under the most critical phases of flight -- landing and takeoff.
The rudder limiter is designed not to function at altitudes 700' above the ground and below on landing and 1,000' above the ground and below on takeoff. So at a point in time when the crew has the least possible altitude within which to regain control of the aircraft, the potential for rudder malfunction is uncorrected. Even with the current modifications, the United 585 crash in Colorado Springs would still likely have occurred.
There is always a risk when interpretation of safety regulations is stretched, strained, winked at or disregarded, and that is, an accident. How many more accidents? How many more people will die before those who should be held accountable are held accountable?
According to the NTSB, there may be recommendations to split the rudder of the 737, like the 727 rudder, and use two actuators to provide a level of redundancy required by the regulations, and certainly called for by the accident history. The FAA and Boeing decry such a recommendation, claiming that it would harm the systems in the aircraft and be expensive. Neither the agency responsible for safety nor the manufacturer of the aircraft ultimately responsible for safety address the harm to the other set of systems which will result from the failure to implement these necessary modifications, and that is the systems of the human beings aboard these aircraft. It is those systems that need to be protected at all cost.
But what is perhaps most indefensible in this flurry of sound bites and competition among federal agencies to gain the political upper hand of looking like they are doing something to help the public in the face of an inexcusable eight year delay is the abject failure to appreciate how modern technology can make correction of the 737's flaws far less expensive and less complicated than envisaged or complained of.
We live in a fly-by-wire technology era, where wires instead of hydraulic lines and electric actuators instead of hydraulic actuators are the way aircraft are designed and built. In short, the Boeing 737 can have a second actuator electrically operated and driven and equipped with a comparator, such that if the hydraulic actuator and the electrical actuator don't agree, the rudder will not move. It's cheap, it's simple, it can be implemented relatively quickly, and it can solve the problem.
With regard to the 737 Advanced model aircraft and the certification without rudder activator redundancy, while the NTSB sat back and let it happen, shame on the U.S. Government, shame on the FAA, and shame on the NTSB.
These new airplanes should never have been certificated without a dual rudder actuator, and had the NTSB applied enough pressure, they wouldn't have been. The NTSB has demonstrated a lack of accuracy and thoroughness in accident investigation historically, but now delay can be added to its list of inadequacies.
It is stunning that the first scanning electron micrographs of the United 585 rudder control servo valve that were obtained by the NTSB, were obtained not as a result of its own metallurgical analysis of this suspect component, but of photographs that were supplied by none other than an expert hired by Arthur Alan Wolk. It is stunning that the NTSB didn't know until weeks ago that the servo valve of both United 585 and USAir 427 had metal burrs left in them during manufacture that were not in compliance with the drawings; in other words, manufacturing defects, and that those burrs could cause jamming of the slides of the rudder actuator. It is frightening to think that eight years after initiating an investigation into the crashes of three airplanes full of people that the NTSB is just getting around to having a hearing to tell people what everybody in the industry knew within hours of the accidents -- they were all caused by the rudder, and if you have an airplane with a single rudder actuator, when the regulations require two, that's probably why the rudder caused the accident.
The FAA, Boeing and the NTSB say that the necessary changes in the 737 already in the process of being implemented make a safe airplane even safer. Statistically, that's probably true. After all, 90,000,000 hours and only a few crashes that we know were caused by the rudder. That's a pretty good record. So, I guess then, we have abandoned the principle that when we actually know of a reason for an aircraft crash, it must be fixed so that others will not lose their lives.
So, really instead of having a little machine that checks your boarding pass when you get on a Boeing 737, each passenger should simply spin a roulette wheel, because that is exactly the gamble every passenger and every crew member takes when he or she gets aboard a Boeing 737 with a single rudder actuator. I suggest your carry-on bag should be a parachute!