Thursday, September 18, 2014

American Airlines Flight 191

The most deadly aircraft accident on U.S. soil was American Airlines flight 191.  This aircraft involved in this accident was a McDonnell Douglas DC-10.  This flight brought a lot of negative attention to the DC-10.

The accident occurred on 25 May 1979 while leaving O'Hare International Airport, Chicago, Illinois.  It happened only moments after takeoff.  The aircraft was accelerating down the runway fine, but at the moment of takeoff rotation the number one engine and pylon separated from the left wing and flew up and over the aircraft, landing behind on the runway.  The aircraft immediately lost hydraulic and electrical power supplied by the number one engine.  At the time all the crew knew is that they lost power in the first engine.  They immediately followed emergency procedures by reducing power on the other 2 engines and preparing for emergency landing.  A pilot on the ground was noted for saying that initially the aircraft seemed to climb normally.  Then shortly after the aircraft rolled rapidly to the left and pitched down into a field.

Ideally, completely losing one engine would not be enough to bring the aircraft down.  Unfortunately when the crew lost the number one engine they lost number one hydraulic power and number one electrical power.  Upon takeoff the aircraft had slats deployed as would have been normal for this stage of flight to create lift.  Each wing's set of slats where controlled by their respective hydraulic system.  When the aircraft lost number one hydraulic system the slats on the left wing lost hydraulic pressure on them so the force from the air moving across the left wing caused the left side slats to retract.  This effectively caused the left wing to go into an aerodynamic stall while the right wing maintained lift.  This effectively caused the aircraft to roll hard left (
NTSB, 1986, p. ii).  

When the aircraft lost number one electrical power the crew lost power to vital systems like the stick shaker and the slat disagree sensors (NTSB, 1986, p. 23).  The stick shaker is a system that literally shakes the pilots flight stick in order to let him know that the aircraft is going into a stall.  The slat disagree sensors are sensors that let the pilot know if the left and right wing slats are not in the same position.  This is obviously vital for a stage of flight as important as takeoff.  So as you can see, the crew had a major problem and no way to know about it.

At the time the DC-10 didn't have any sort of slat locking system to keep the slats deployed under air load like most modern aircraft have.  Because of this accident, slat relief valves were installed on all DC-10 aircraft (NTSB, 1986, p. 43).

All these things still don't answer the big question here in this accident and that is why did the number one engine separate from the wing in the first place?  For engine change procedures McDonnell-Douglas required that the engine be first removed from the pylon and then have the pylon removed from the aircraft.  To save time, American Airlines developed a technique where they removed the engine and the pylon whole while supported by a forklift (NTSB, 1986, p. 26).  This method saved a lot of time but it was difficult to keep the engine from moving.  Several weeks before the accident, the umber one engine was reported to have been changed.  The engine had shifted and damaged the mounts, but nothing was done when the damage was noticed.  After weeks of flying the mounts had developed growing fatigue cracks until the mount completely failed on takeoff.

This flight along with several other DC-10 crashes gave the aircraft a lot of negative attention.  So much so that after this flight the FAA pulled the flying certification and essentially ground the aircraft until further investigation was done.


NTSB. (1986). Aircraft accident report NTSB-AAR-79-17. Retrieved from http://libraryonline.erau.edu/online-full-text/ntsb/aircraft-accident-reports/AAR79-17.pdf


Friday, September 12, 2014

Another aircraft mishap caused by mechanical failure

Air France Flight 4590

Usually when we talk about mechanical failures causing aircraft mishaps we assume that we are talking about only one aircraft involved.  In the case of Air France Flight 4590, a mechanical failure from another aircraft ultimately caused this aircraft's mishap.  

Air France flight 4590 was one of the few Concorde aircraft that were designed for supersonic flight.  The Concorde had been flying for 27 years with no fatal mishaps up to this point.  The flight occurred on 25 July 2000.  The aircraft was flying from Charles de Gaulle International Airport near Paris to John F. Kennedy International Airport in New York City.  

Five minutes prior to take off, a Continental Airlines DC -10 was departing for New Jersey at the same airport.  Prior to its take off the aircraft had repairs done to one of the engine thrust-reversers.  The mechanic used a titanium alloy strip to replace the wear strip in the thrust reverser that wasn't authorized (Learmount, D., 2000).  The strip was a non-OEM part with an irregular rivet pattern that was not machined properly.  This wear strip is often replaced because it receives so much wear and tear.  When the DC-10 took off it dropped this wear strip on the runway.  Before takeoff of all Concorde aircraft the airport was supposed to conduct a visual runway inspection for debris, but this was never accomplished.

When the Air France Concorde took off it was over max gross takeoff weight.  Partly this was because of all the fuel needed to make the international flight.  Reports also showed that the aircraft had a center of gravity that was way too far aft for safe flight.  The wing tank had also been overfilled.  Despite this, none of these factors were directly linked to have caused the crash.  

Upon takeoff the Concorde's main landing gear struck the metal strip left on the runway from the DC-10.  The tires exploded and the air blast from the explosion caused the wing tank to rupture.  The fuel rupture naturally caught fire and the aircraft lost power to one of the four engines that were desperately needed during takeoff.  The aircraft failed to maintain altitude and stalled, striking a nearby hotel.  Altogether 113 people perished including 4 people on the ground.



Though the main landing gear striking the metal strip was determined to be the major cause of the accident, there were still several other factors and theories in play here.  In order to save weight the Concorde was designed to fly without flaps and slats.  This obviously did save weight, but dramatically increased the takeoff and landing speed of the Concorde.  This increase of speed also increased the speed of the main landing gear wheels, making the aircraft more prone to tire failure. This increase in speed made the tires explode more violently.  Long before this accident in November of 1981 the NTSB sent a letter to the French Aviation safety board, BEA (Bureau d'Enquêtes et d'Analyses pour la Sécurité de l'Aviation Civile), stating their concerns over tire safety for the Concorde (Cosgrove, M., 2010).  

Also there are theories that because the Concorde was over max takeoff weight that the pilots needed excess runway to takeoff and that the strip of metal was past the point of normal takeoff for a Concorde (Cosgrove, M., 2010).  It was also found that there was a crucial spacer missing in the main landing gear that kept the landing gear from pivoting.  There were tire marks on the runway from where the tires were pulling to the left, possibly slowing the aircraft and making it harder to get liftoff.  The BEA found that these factors were negligible and not a factor in the accident.

Regardless the reasons for this mishap, it proved the end of the flying service of the Concorde.




References

Cosgrove, M., (2010). Op-Ed: Was the Concorde disaster trial a French cover-up?, Digitaljournal.com Retrieved September 12, 2014.

Learmount, D., (2000).  Poor repair' to DC-10 was cause of Concorde crash, Flightglobal.com. Retrieved September 12, 2014.

Wednesday, September 10, 2014

One example of a mishap caused by mechanical failure.

Chapter 10: ASA Flight 529

Most aircraft have several backup systems for when mechanical failure happens in flight in order to keep the failure from proving fatal.  Sometimes despite these backup systems, an aircraft can have a mechanical failure that is so awful the mishap still proves fatal.  This is what happened on American Southeast Airlines flight 529.  

American Southeast Airlines specialized in being a regional airline that flew short routes from large airports to smaller ones (Walters, J., & Sumwalt, R., 2000, page 215).  Flight 529 was an EMB-120 type aircraft.  This plane is a small 29 passenger turbo-propeller aircraft that has an aircrew that consists of a captain, a first officer, and a flight attendant.  




This flight happened on August 21, 1995.  That day there was rain and fog that limited visibility.  Immediately after takeoff the aircraft experienced a fair amount of turbulence due to weather, but later it stabilized out.  About twenty minutes after takeoff the crew heard several loud thuds.  The pilots had autopilot disconnect and got warnings for engine control and engine oil.  The captain then feathered the left propeller and cut off fuel to the left engine.

The EMB-120 became uncontrollable, but the pilots couldn't figure out why.  The propellers are designed to do what is called feathering.  This process moves the propeller blades flat so they are inline with the air stream to help prevent excess drag from the engine.  Normally if you feather the propeller and yaw trimmed the aircraft can be managed.  Unfortunately this engine failed so badly that it got knocked askew in its mount and three of the blades were wedged into the wing causing large amounts of drag.  The pilots had to throttle down the right engine a considerable amount to be able to maintain some sort of control of the aircraft.  Because of this they were descending rapidly.  

The crew tried to coordinate with ATC to get the nearest runway to land on.  Unfortunately the ATC agent wasn't able to give them clearance to the nearest airport due to lack of information from that airport.  Their aircraft was four miles out from the runway that they were coordinated from when they started receiving altitude warnings from the Ground Proximity Warning System (Walters, J., & Sumwalt, R., 2000, page 214). The pilots requested that ATC have the fire department waiting, but they were not called in time.

At this point the flight attendant was preparing the cabin for the crash and instructing the passengers on how to brace for the impact.  While demonstrating what to do she looked out the window and saw the tops of trees passing by so she quickly strapped in. (Walters, J., & Sumwalt, R., 2000, page 215)  The pilots somehow managed to keep the aircraft from striking a bunch of trees and crash landed the plane into a field.  The plane slid five hundred feet and broke in half.  It had only nine-and-a-half minutes from when the crew first started hearing thuds in the cockpit at 18,000 feet to when the aircraft struck the ground (Walters, J., & Sumwalt, R., 2000, page 215).




Most of the passengers were able to escape immediately through the large opening in the fuselage.  Others were not so lucky.  In less than a minute some cut electrical lines started a fire and smoke was engulfing the area.  Passengers were catching fire as they were leaving the scene.  Though seriously injured the flight attendant was doing her best to put out the fires.  All together there were nine fatalities including the captain.

The NTSB determined the cause of the accident was "the in-flight fracture and separation of a propeller blade resulting in distortion of the left engine nacelle, causing excessive drag, loss of wing lift and reduced directional control of the airplane" (Walters, J., & Sumwalt, R., 2000, page 223).  The propeler blade that broke off was a Hamilton Standard 14RF-9 design.  These blades are made of aluminum with a conical shaped hole called a taper bore drilled into them that is used for weight reduction and balancing.  The mechanics place different amounts of lead wool inside to help balance the blade.

The NTSB found that the blade failed because of fatigue cracking.  This was likely caused by corrosion and mechanical damage from tools that were used to insert the lead wool.  Though mechanics had noticed these flaws before, they never removed the blade from service.  It was also determined that ultrasonic inspections should have been required that would have found the cracks in this blade and taken it out of service.

It's sad that nine people had to die to learn the lessons from this accident, but I think a lot was learned.  Altogether the aviation industry is a little safer because of the findings from the NTSB, and I think that is the point of aircraft safety investigations.



References

Walters, J., & Sumwalt, R. (2000). Aircraft accident analysis final reports. New York: McGraw-Hill.

Thursday, September 4, 2014

Air Detective 14, Fire

This weeks Air Detective is about fire.  A fire in flight is can be catastrophic.  Sometimes what is just as bad as an actual fire in flight is a false fire indication.  This was the case for a
Russian Airline called Aeroflot.  On July 6, 1982, Aeroflot flight 411 took off from Moscow.  The aircraft was a four engine Ilyushin-62.  The Ilyushin-62 has the engines sit directly next to each other, directly next to the fuselage.  With this design, if the pilots have an engine fire in any one engine they had to shut down both engines in the pair in order to avoid further damage to the other engine and the fuselage.  Shortly after take off the pilots got an engine fire indication.  Because of this they shut down both engines on that nacelle.  The aircraft tried to return to the airport, but because half the engines were off, the aircraft was very difficult to control.  Due to lose of control they crashed six miles from the airport.  Afterwards, investigators found that the fire indication was false and caused by a faulty fire warning system in the powerplant.   Soviet officials would not disclose any information about the passangers, but western sources believed about 90 passangers were on board and none survived.




Fire was also a very serious problem on an aircraft crash that was very close to home for me.  I'm talking about United Airlines flight 232.  I moved to Sioux City, Iowa in early 1989.  I was pretty young, but I still remember it.  United Airlines flight 232 was a DC-10 that had the center engine that sits in the vertical stab fail and break apart so violently that it sent shrapnel through three hydraulics systems.  This made it so that the aircrew lost all control of their flight controls.  The crew had to use the remaining engines to guide the aircraft to Sioux City Gateway Airport.  


Above is a picture of the failed fan assembly.

When the aircraft crash landed, it rolled and caught fire.  There were thirty five people that were sitting in the middle of the fuselage directly above the fuel tanks.  They died from smoke inhalation.  In most fires the victims die from smoke inhalation and not from the actual fire.  Many innocent people lost their lives, but many also survived because of the efforts of the crew, emergency personnel, and 285 Air National Guardsmen that were stationed and on duty at the airport.






Belsie, L. (1989) Piecing Together the Reasons for United Flight 232 Crash. Christian Science Monitor Boston, MA, Retrieved from www.lexisnexis.com/hottopics/lnacademic

NTSB. (1990). Aircraft accident report NTSB-AAR-90/06. Retrieved from http://www.airdisaster.com/reports/ntsb/AAR90-06.pdf

Press, A. (1982, Jul 06). Soviet Jetliner en Route to West Africa Crashes In Moscow; 90 Believed; Killed. Boston Globe (Pre-1997 Fulltext) Retrieved from http://search.proquest.com.ezproxy.libproxy.db.erau.edu/docview/294171178?accountid=27203

Saturday, August 30, 2014

Chapter 7 Post

This chapter talks about a MD-88 aircraft striking the approach end of a runway  with its main landing gear and narrowly escaping a catastrophe.  The NTSB found that one of the major contributing factors for this accident was the fact that the captain had contact lenses that weren't approved by the FAA for pilots.  These lenses created an optical illusion for the captain where he couldn't quite tell how fast he was descending.

Though mechanical failure was in no way to blame for this accident, there was an aircraft factor I found particularly interesting.  The crew had a vertical speed indicator that was labeled as "noninstantaneous."  This meant that the indicator could have a lag of up to 4 seconds before displaying accurate climb or descent rate.  The indicator could have been updated to show instantaneous rate if it was integrated with the inertial reference unit.  The NTSB determined that the lag time in the VSI also contributed to the accident.

Walters, J., & Sumwalt, R. (2000). Aircraft accident analysis final reports. New York: McGraw-Hill.

Air Detective Tips

The two major concepts that are talked about in the Air Detective Tips are the 5M and the Shell concepts. In this post I will go over them and how they are applicable to mechanical failures in aircraft.
The 5M concept is applied to accidents to help identify what causes an accident. The 5M's are Man, Machine, Mission, Medium, and Management (Lawin, R). With this model an investigator can bring these factors together to see which things may have worked together to cause an accident. Obviously in my paper I will be focusing on the machine part of this article.
Investigators can also use the SHELL model to better understand aircraft accidents. This model shows the interface between the human element or "liveware" and the surrounding elements (Lawin, R). The first example is liveware-software. This could mean training manuals or instructions. The next is liveware-hardware. This is how the human factor interacts with the aircraft and the mechanical parts of the aircraft. This is what I will focus on the most. The next one is liveware-environment and this encompasses environmental factors like weather or wind shear. The last is liveware-liveware. This happens a lot when there is a conflict amongst aircrew. Things like proper CRM could fall under this category.

https://erau.blackboard.com/bbcswebdav/pid-15077961-dt-content-rid-38727694_4/institution/Worldwide_Online/SFTY_UG_Courses/SFTY_330/Air_Detective_Tips/AirDetectiveTips9to12_0712.pdf

AD-13

This Air Detective tip talks about all the details that are scrutinized in an accident investigation.  Basically its a brief outline of each one of the factors that is looked at.  I have now been part of a couple major aircraft mishaps and I can tell you that investigators are thorough in looking at everything involved even if they already know the cause of the mishap.  Everything from aircraft forms to training records are seized and locked away and looked over.


Monday, August 18, 2014

Aircraft Accidents Caused by Mechanical Failure A brief overview

I chose my topic because it was something that really interested  me.  For me it is easier to read about an accident when no one is necessarily at fault and it is simple mechanical failure.  I also chose mechanical failure because it's one of the most common causes of aircraft crashes.  I picked this topic because as an aircraft mechanic myself, it is easy for me to understand and translate some of terms associated with this topic.




The truth of the matter is that about half of aircraft crashes occur from pilot error (Rohrer, F.,  Castella, T. 2014).  Unfortunately I believe that whenever an aircraft incident happens mechanics or mechanical failure is one of the first things blamed.  Throughout time it seems that pilots are always exalted.  Even I have been guilty of feeling the same way.  Even more unfortunately I have had two airplanes I have worked on go down.  One was because of a commonly know mechanical issue that was neglected and the other was just pilot error.  The instant I heard about each crash all I could think about is what could I have done wrong to endanger that aircraft.


Of course the whole purpose of investigating aircraft incidents is to prevent them.  If I get nothing else out of this class I will take what I learn here to make my workplace safer. 


Aircraft mechanical failure can be defined as a failure or break down in a mechanical aircraft component due to that component reaching the extent of its "age" or improper install or application. 


Pilot error is a decision, action or inaction by a pilot of an aircraft determined to be a cause or contributing factor in an accident or incident.


Here is a relevant link

http://www.bbc.com/news/magazine-26563806

Rohrer, F., & Castella, T. (2014). Mechanical v human: Why do planes crash? Retrieved August 30, 2014.