Monday, December 22, 2008

Continental 737 Crash at Denver–Better Look at the Rudder

Yet another Boeing 737 crashes, but this time no one was killed. The flight crew masterfully rejected a takeoff that went wrong. Loud noises were heard that were reminiscent of the sounds identified just before domestic flights on United 585 and USAir 427 rolled over and dived to the ground, killing a total of 152 people in 1991 and 1994 respectively, and overseas airlines COPA 201 and SilkAir 185 crashed, taking the lives of another 151 people in 1992 and 1997.

If I were the NTSB investigator in charge, I would pull the rudder actuator and take some SEM photographs to see if the actuator bears a resemblance to the three other actuators that showed witness marks of jamming.

In my opinion, the Boeing 737 still does not have a reliably redundant rudder control system, and even after hundreds of deaths, the FAA allowed Boeing to build an entirely new generation of B-737’s with a single rudder actuator when all of its other aircraft have at least two.

Noises heard on earlier cockpit voice recorders were the death sounds of an aircraft about to go out of control. These sounds are generated by the hydraulic system telegraphing its agonizing inability to control the rudder. At speeds below 190 knots, the rudder will cause a rapid roll of the aircraft that cannot be stopped before tragedy occurs.

While redesigned after the accidents of the 1990’s, the rudder control system still has no true redundancy. If the flight crew of this aircraft sensed that they were about to lose directional control, they saved themselves and all their passengers from certain death.

The airplane is trashed and some people were hurt, but everyone will ultimately go home to their families this Christmas. Congratulations to a “heads up” Continental crew.

Arthur Alan Wolk
December 22, 2008

Thursday, September 18, 2008

Madrid MD-82 Crash, Déjà Vu

Two decades ago in Detroit Michigan, Northwest Airlines Flight 255, an MD-82, crashed on takeoff, killing all aboard except for a toddler. The crew had failed to extend the wing flaps and the takeoff configuration warning was disabled due to lack of electrical power to the device, so no warning was sounded.

Now it appears that first witness reports about an engine explosion on Spanair MD-82 upon its takeoff in Madrid, Spain on August 20 were in error. Instead, investigators have found that the plane did not have its wing flaps deployed when it stalled and crashed to the runway killing 153 of its 175 passengers and crew. Once again, it appears that the crew failed to extend the wing flaps, thus ignoring that item on the pre-takeoff check list. The cockpit voice recorder should confirm or deny whether the crew announced the need to set flaps for takeoff.

Typically, takeoff configuration warnings do not sound because they have been disabled due to frequent false warnings. A warning system is useless if it frequently malfunctions because flight crews will just ignore the warnings as unreliable. On the other hand, pre-takeoff check lists, which include challenge and response by the flight crew working together, should have resulted in proper flap extension. It has not yet been determined why the takeoff warning on the Spanair aircraft didn’t work and it was never determined why it didn’t work on the Northwest aircraft more than 20 years ago.

The flight path of both the Northwest and Spanair aircraft are eerily similar, with the nose seen coming up to takeoff altitude, followed by an aerodynamic stall resulting in a rapid descent to the ground with a large loss of life.

The fact that Spanish investigators heard no takeoff configuration warning on the cockpit voice recorder is just a “same-old, same-old” repeat of the well-known adage that aircraft always telegraph their intention to fail long before an accident. This problem has been around for at least 20 years and obviously a fix has not been ordered by the FAA, the agency responsible for ensuring aircraft safety.

It is hideous that the manufacturer hasn’t fixed this known fatal flaw that has now taken hundreds of lives.

Arthur Alan Wolk
September 18, 2008

Monday, September 8, 2008

Criminalization of Air Disasters

Nothing good comes of criminal prosecutions following air disasters. While such proceedings may satisfy the public’s zeal to punish those responsible, the result is that the flow of information necessary to correct aviation problems dries up over the long term because of the fear that such information will be used for criminal prosecution in the event of accidents.

It is bad enough that manufacturers and airlines now hide what they do, or more importantly what they don’t do, in an effort to escape civil liability for accidents. Criminalization has always been fraught with the specter of witnesses using their Fifth Amendment rights not to incriminate themselves (which has the effect of impeding investigations that might result in safety improvements).

Moreover, public authorities, whether prosecutors or public investigators, do a terrible job at investigating aircraft accidents and are too often the tools of manufacturers and airlines. Plaintiffs’ lawyers do the majority of aircraft accident investigations in the United States and spend far more, examine more intensively and extensively, and take sworn testimony more often to get to the bottom of these accidents. Criminalization will impede, not enhance, these efforts. What we need is more zealous sanctions when airlines and manufacturers hide information from the certifying authorities, distort warnings received from the field and flat out lie during civil proceedings. We need fewer judges who are selected for their promise to deter plaintiffs’ lawsuits; we should go back to hiring judges based on their demonstrated lack of bias and predilection.

This issue has been around for years and is most often discussed in countries where civil litigation does not exist the way it does in the United States. Where there is no suitable vehicle to get to the truth civilly, criminalization is the fall-back position taken out of frustration. The real solution is to expand civil litigation systems in countries that don’t currently have them so that safety is enhanced rather than deterred by the regressive effects of criminalization.

Arthur Alan Wolk
September 9, 2008

Friday, August 22, 2008

Crash of Spanair MD-82 Could Have Been Prevented

One hundred fifty-three people were killed when a fully-loaded Spanair MD-82 crashed on takeoff from the Madrid airport on August 20, 2008. Witnesses describe an explosion in the left engine and the aircraft falling to the runway, veering off and then exploding into an inferno.

About one hour before the crash, the crew was reported to have aborted a takeoff because of a high temperature warning, known as an overtemp, from the left engine. Troubleshooting took place and the aircraft was again dispatched for a three-hour flight to the Canary Islands. The aircraft was full of fuel and the outside temperature was a warm 86 degrees Fahrenheit.

It is inexplicable and horrendous that an aircraft such as this was cleared for service without an engine change when they couldn’t possibly have known what damage resulted from the overtemp. This situation in a jet engine is a serious matter requiring extensive investigation, not merely an hour’s worth of troubleshooting. It does not happen without good, and usually serious, cause including imminent failure of hot section components, failure of compressor and fan assemblies or likely turbine disintegration.

Jet engines nearly always telegraph their imminent failure and this one surely did. It was ignored. The Pratt and Whitney JT8D engine has had a long and satisfactory service history starting with the Boeing 727, 737, DC-9 and all the MD-80's. While generally reliable, it has exploded all too frequently and is the subject of numerous service bulletins and airworthiness directives by the Federal Aviation Administration as well as aviation safety agencies in other countries. Some have exploded so violently that the containment rings that are designed to prevent penetration of debris into other critical aircraft components have proved to be insufficient.

When the MD-82 was built and certified, it was supposed to be able to safely fly with one engine if the other failed after reaching V1 (the speed at which it becomes safe to continue the takeoff in spite of the failure of an engine). However, the claim was substantiated by test pilots who know the engine will be simulated to fail, and not by flight crews who are totally surprised by the event on a hot day while fully loaded.

It is not surprising that the airplane didn't fly because expecting humans to perform to perfection is unreasonable, and the temperature and weight were likely well beyond the test parameters. At the controls were Spanair test pilots who were also victims of a cascade of events that came together to create this tragedy.

At this stage of aviation history, there is no reason for this disaster and both the horror and pain it has caused to so many families. Airline travel can and ought to be 100 percent safe if people do their jobs. This was an unnecessary accident.

Arthur Alan Wolk
August 22, 2008

Friday, August 1, 2008

Hawker 800 Crash In Minnesota–Some Preliminary Thoughts for All Pilots

A Hawker 800 executive jet aircraft crashed at a small airport in Minnesota, south of the Twin Cities on July 31, 2008. Witnesses describe a touchdown on a wet runway, a roll-out, then the application of power and a crash just beyond the end of the runway. All aboard were killed, either from impact or the ensuing fire.

The crew had successfully diverted around heavy weather and was landing in an area of moderate rain showers. Just before the crash, a wind shear alert was given (a rapid change in wind direction and velocity). Normally, a flight crew will add as much as 10 knots to the landing reference speed in the event of wind shear. The aircraft made it to the airport and the runway, and presumably the lift dump system was deployed, which should have allowed the aircraft to stop in 4,000 feet on a dry runway, according to the flight manual. It is typical for jet aircraft to touch down in the first 1,000 feet but, due to the wind shear, the first 1,500 feet may have been used. Therefore, except for the wet runway, 5,500 feet should have been sufficient for landing to a full stop, leaving about 750 to 1,000 feet remaining, if all went as expected. However, because of reported tail winds and standing water on the runway, it was questionable whether stopping in the available distance was possible. Hydroplaning, further increasing stopping distance, was also likely.

For all, or some of these reasons, the landing didn't go as expected. The crew elected to “go-around,” meaning they would have had to stow the left dump spoilers and flaps, and trim and apply takeoff power. The engines, likely idle by this time, would have had to spool up and the aircraft would have needed to accelerate again to take-off speed. The time to make the decision to go-around, stow the lift dump and achieve take-off thrust could have taken 10 to 15 seconds and used up another 1,000 to 1,500 feet of runway. That left precious little, or no, runway for the aircraft to accelerate to lift-off speed.

The aircraft did not leave the ground. Wheel tracks show that it left the pavement, traveled in the runway safety area–still on the ground–struck an antenna array, and then fell into a culvert where it burned. While there is no doubt the crew tried to save the aircraft, the decision to go-around on that size runway may have doomed the flight.

Investigators will review the cockpit voice recorder and, if equipped, the flight data recorder for clues and will look at the wreckage to determine the aircraft configuration at time of impact.

This information should confirm that unless there was a mechanical malfunction, this accident was preventable, like so many others of a similar type.
  • First, a go-around from idle thrust and with lift dump deployed on a 5,500 foot runway with this aircraft is nearly impossible under normal circumstances.
  • Secondly, aircraft performance charts are prepared to sell airplanes and bear little resemblance to actual performance achieved by average pilots in the field.
  • Thirdly, a pilot must always plan for contingencies and on this approach there was wind shear, lightning in the distance in all quadrants, heavy weather nearby and a short wet runway.

Even the best flight crew can find itself without options under circumstances like these. The tragedy of this accident reminds us all that aviation safety means no accident, whether it’s a mechanical malfunction, the combination of foul weather and a short runway, a faulty decision to go-around based upon inadequate aircraft performance information or just a mistake.

No conclusions can or should be drawn about this accident as the investigation has just begun, but these are some thoughts that bear consideration regardless of the ultimate findings.

Arthur Alan Wolk
August 1, 2008