Maintenance Human Factors incidents
 

I am doing a bit of research accidents and incidents that had causal factors associated with maintenance failures, particularly the human factors aspect.
Causal factors such as errors induced by organisational failures, fatigue related issues, shift handover, time pressures etc.
I am looking for accident or incident reports that specify or allude to some of these issues. Any suggestions or links would be much appreciated :ok:

Here are two that I studied during my Masters coursework:

Convair 580 and Embraer Brasilia

In the first, the pitch actuating mechanism failed due to heavy wear. Although a special oil analysis inspection was in place to detect the wear, it failed to do so. The failure occurred during high IAS, leading to very rapid overspeed and blade failure, cutting the fuselage in two.

In the second, the pitch feedback leadscrew failed due to wear. A preflight runup check failed to detect the incipient failure. The failure occurred during approach, causing high yawing moment and wing airflow disruption on one side, with uncontrollable roll.

Both accidents had common factors:

1) The physical characteristic of all variable-pitch propeller blades to migrate to fine (low) pitch when control forces are absent.

2) maintenance/inspection programs which proved inadequate.

The first point is a fact that Hamilton Standard, which has built controllable props since the early 30s, and the FAA certification branch seemed to have jointly forgotten by the 80s; The Failure Modes, Effects, and Criticality Analysis of the HS prop on the Brasilia was remarkably "lightweight".

Interesting speculation: IF the Brasilia failure had occurred at high IAS, the overspeed might have caused the prop blades to fail, but their lightweight composite design might have prevented them from penetrating the fuselage, which had occurred in the CV580 accident.

Thanks Barit1, I think I might use one of those and the Air Transat incident in the Azores. Any other suggestions are welcome :ok:

Look back through the AAIB website - you'll find plently to get your teeth into

http://www.aaib.gov.uk/publications/index.cfm

check out the reports ref G-CPER (BA 757) and G-YMME (BA 777) on the AAIB website.....both very near misses (although BA are of the opinion that safety was not compromised....yeh, whatever!)
If you download CAP 716 off the CAA website you will see the key accidents in HF history mentioned....BAC 1-11, Excalibur A320, BMI 737 etc etc etc.
The best one to read however w.r.t maintenance system failure is Alaska 261 available on the NTSB website. Bit of haul to read but......pure maintenance accident and scary! All maintenance staff should know about this one......who hasn't skimped on the lubrication cos' 'it's only grease'....

oh...and wait for the helios 737 accident report, due out soon and the ATR-72 which ditched off the coast of Sicily last year due to fuel starvation as a result of the incorrect fuel gauge being installed!!

Excellent, thanks for that the BA 757 one is perfect for my paper, I will also use the Air Transat dual engine failure as both this and the 757 are both fairly new and have official reports.

FAA Human factors site is

http://www.hf.faa.gov/hfmaint/Default.aspx?tabid=275

Its is a huge data base, training coarse etc so ensure lots of paper in the printer available :)

Also a source of good articles relating to maintenance incidents monthly (including thise caused by human factors) reports is the Aviation Maintenance magazine
as available also online at www.aviationmx.com


Just completed putting together our Human factors coarse for our engineers (awaiting it approval)

Sea King
 

Check out the Australian Navy Sea King heli crash. Very sad indeed.

Saddly, here is the freshest one on Spectrum 33 (VLBJ) test article:

http://www.ntsb.gov/ntsb/brief.asp?e...31X01059&key=1

NTSB Identification: SEA06FA146
14 CFR Part 91: General Aviation
Accident occurred Tuesday, July 25, 2006 in Spanish Fork, UT
Aircraft: Spectrum Aeronautical LLC 33, registration: N322LA
Injuries: 2 Fatal.

This is preliminary information, subject to change, and may contain errors. Any errors in this report will be corrected when the final report has been completed.

On July 25, 2006, approximately 1606 mountain daylight time, a Spectrum 33 experimental twin-engine jet airplane, N322LA, collided with terrain following a loss of control during the initial climb after takeoff from runway 30 at Spanish Fork-Springville Airport, Spanish Fork, Utah. The airplane, which was registered to and operated by Spectrum Aeronautical LLC, was destroyed by impact forces. The two commercial pilots aboard received fatal injuries. Visual meteorological conditions prevailed and no flight plan was filed for the 14 CFR Part 91 local maintenance test flight. The flight was originating when the accident occurred.

Witness observations indicate that the airplane entered a right roll almost immediately after takeoff. The roll continued to about 90 degrees right wing down when the right wingtip impacted the ground.

Examination of the accident site revealed that the initial impact point was located about 150 feet right of the runway 30 centerline. A ground scar oriented on a magnetic heading of about 330 degrees extended from the initial impact point to a barbed wire fence about 120 feet away. Various pieces of right wing debris were found along the ground scar. The wreckage path veered about 20 degrees right at the fence and then remained essentially straight to the main wreckage site on about a 350 degree magnetic heading. The main wreckage was located about 750 feet from the initial impact point and included the forward fuselage, aft fuselage and a majority of the wing structure. All major components of the airplane were accounted for in the wreckage path or with the main wreckage. There was no evidence found of any pre-existing failures of the airplane's structure.

Roll control on the airplane was from the pilots' side sticks to the ailerons through a mechanical system of torque tubes and push-pull tubes. The left side stick was primary, and the right side stick was slaved to the left side stick. The roll control motion of the left side stick was linked through a quadrant below the cockpit floor to the lower torque tube. The lower torque tube ran from the quadrant to the aft pressure bulkhead. The translation linkage, the linkages and bell cranks that translated the rotational motion of the lower torque tube to a linear motion of the aileron push-pull tubes, was located on the aft side of the pressure bulkhead in the main landing gear (MLG) gearbox area.

During examination of the wreckage, aileron control continuity could not be established from the cockpit to the aft pressure bulkhead due to fragmentation of the airplane, however, all of the lower torque tube was accounted for. Control continuity was established from the torque tube input on the aft pressure bulkhead to the aileron bellcrank on the right wing and to the torque tube about 50 inches inboard of the aileron bellcrank on the left wing. Examination of the translation linkage on the aft side of the aft pressure bulkhead revealed that it was connected in a manner that reversed the roll control. Specifically, the linkage was connected such that left roll input from the side sticks would have deflected the ailerons to produce right roll of the airplane, and right roll input from the side sticks would have deflected the ailerons to produce left roll of the airplane.

According to information provided by the operator, the airplane had accumulated about 44 hours total flight time since its first flight on January 7, 2006. Prior to the accident flight, the airplane's most recent flight, flight number 46, had taken place on June 30, 2006. During the time between flight 46 and the accident flight, the airplane had been undergoing maintenance. The maintenance included removal of the MLG in order to stiffen the MLG struts. Upon reinstallation of the MLG, it was found that inadequate clearance now existed between the left MLG strut and the aileron upper torque tube V-bracket. The V-bracket was removed and redesigned to allow proper clearance of the MLG. Removal of the V-bracket required disconnection and removal of a portion of the translation linkage

A similar crossed aileron control brought down a L-188 at ORD in 1961.

There are probably many cases of throttle rigging errors or potential errors. Here's one of the worst:

The CT64 turboprop engine in the DHC-5 Buffalo had some positive-engagement features on the throttle and fuel/idle select shafts of the FCU, but DHC never had the mating features to make it foolproof.

Consequently there were several ways to get it wrong. In one case at the flight line, during a zero-torque run (no prop installed) the engine oversped but the o/s trip circuit kept it from destroying itself.

But the oscillating speed was misinterpreted by the techs, and to fix the annoyance they started pulling c/b's. At this point the overspeed continued to well over 150%, the power turbine discs came apart and cut the fuselage in two.

:ugh:

Avro's Chief Designer, Roy Chadwick, Chief Test Pilot Bill Thorn and two other crew were killed in the 1947 prototype Tudor II accident, which was also caused by crossed aileron controls.

report I have just received this week of a crash this week


Reversed Aileron Linkage Eyed in Spectrum 33 Crash


According to Spectrum Aeronautical LLC founder and CEO Linden Blue, the proof of concept aircraft that crashed last Tuesday in Spanish Fork, Utah, killing test pilots Glenn Maben and Nathan Forrest, had recently undergone reinforcement of its main landing gear attachment structure, a process that required disconnecting and reconnecting the aileron control linkages. During reassembly, Blue told B&CA Senior Editor Fred George, the aileron control linkages were reversed. Safety checks by engineers, technicians, ground service personnel and, ultimately, the pilots failed to detect the flaw, Blue said. NTSB is investigating the accident; its findings are not expected for several months.

Blue said most of the scheduled test flying of the first aircraft had been completed before the crash, which occurred on its 46th flight, and that construction of a second, ’company conforming‘ aircraft, is well underway with first flight slated early to mid 2007.

The second aircraft will have aerodynamic and structural commonality with production versions, including a conventional elevator and trimmable horizontal stabilizer on the tail instead of the stabilator fitted to the original POC aircraft. He said a third, fully-conforming production aircraft will be built for FAA / EASA flight testing. Production versions will be designed so that the aileron linkages cannot be reversed during initial assembly or routine maintenance, Blue asserted.

Language Error in Aviation Maintenance
 

This is a link to a study done on language problems in aviation maintenance. Not 100% what you are looking for but certainly an interesting read for you. Best,PE
http://www.hf.faa.gov/docs/508/docs/...anguageYr2.pdf

Not exactly PE, but an interesting read none the less and validates much of my last decades experience with multicultural engineering personnel