A logical Explanation for the TAM A320 Accident's Observed Runway Events at Congonhas

To be crystal clear, we'll start off with the conclusion (that's based upon the histories and observations/links provided below):

a.  It would appear that TAM Flt JJ3054 did land a little far in and a little hot.

b.  Once established with the left-hand engine in reverse and spoilers deployed, it then became apparent to the handling pilot that a (typical Ibiza style) A320 wheel-braking failure had occurred (i.e. the never totally resolved BSCU failure mode).

c.  Both recognizing and resolving this situation can be a very runway- and time-consuming affair (see below on 03 Aug 2003 accident to A320 regn C-FTDF at http://www.aaib.gov.uk/cms_resources/C-FTDF.pdf ), so the captain understandably aware of the lethal lack of an overrun, decided to apply power and go-round.

d.  An engine going from full rev to full forward thrust could possibly suffer a compressor stall while the reverser doors are closing. The A320 Thrust reverser can take up to 5 secs to re-stow, so the application of TOGA power caused a predictable compressor surge on the LH engine well after the RH engine had gained full power, thus causing the first flash seen on the video and the divergence  (due to the considerable thrust asymmetry) well left of the centerline.

e. However even if the pilots did NOT apply power in an attempt to go round, leaving the throttle of the engine (#1) with the locked-out reverser up at 22.5 degrees (i.e. not retarding it to idle at the "retard" call, would allow that engine to automatically go to TOGA once the auto-throttle disconnected. In addition the spoilers would not auto-deploy nor the autobrake operate. It's a nasty trap and a facet of Airbus operation that isn't often encountered and can easily be overlooked. See what one pilot said about his own personal experience:

f.  Thus probably not a lot to do with the lack of runway grooving, but possibly directly related to the runway length and abysmal lack of a usable RESA (Runway End Safety Area).... and the resulting pilot apprehension.

http://tinyurl.com/2dpkwa

The directional control problems in the America West A320 accident below is pertinent as it was the RH (#2) thrust reverser on the TAM A320 that was OUT of SERVICE. The TAM crew ended up well LEFT of centerline in impacting the TAM Cargo Building so that divergence to the LEFT would be easily explained by the RH engine being intentionally TOGA'd (i.e. the button-pressed for go-round power). The braking fault (see later) would have been the reason for them NEEDING to go round.

from N635AW accident (next below):

"In an effort at maintaining directional control, the captain then moved the #1 thrust lever out of reverse and inadvertently moved it to the Take-Off/Go-Around (TOGA) position, while leaving the #2 thrust lever in the full reverse position."

Not sure this throttle movement is possible on a 737 but it appears so on an AB. It may be relevant to the TAM accident - and perhaps it needs to be looked at by AirBus.

Read on.....

After an asymmetrical deployment of the thrust reversers during landing rollout deceleration, the captain failed to maintain directional control of the airplane and it veered off the runway, collapsing the nose gear and damaging the forward fuselage. Several days before the flight the #1 thrust reverser had been rendered inoperative and mechanically locked in the stowed position by maintenance personnel. In accordance with approved minimum equipment list (MEL) procedures, the airplane was allowed to continue in service with a conspicuous placard noting the inoperative status of the #1 reverser placed next to the engine's thrust lever. When this crew picked up the airplane at the departure airport, the inbound crew briefed the captain on the status of the #1 thrust reverser. The captain was the flying pilot for this leg of the flight and the airplane touched down on the centerline of the runway about 1,200 feet beyond its threshold. The captain moved both thrust levers into the reverse position and the airplane began yawing right. In an effort at maintaining directional control, the captain then moved the #1 thrust lever out of reverse and inadvertently moved it to the Take-Off/Go-Around (TOGA) position, while leaving the #2 thrust lever in the full reverse position. The thrust asymmetry created by the left engine at TOGA power with the right engine in full reverse greatly increased the right yaw forces, and they were not adequately compensated for by the crew's application of rudder and brake inputs. Upon veering off the side of the runway onto the dirt infield, the nose gear strut collapsed. The airplane slid to a stop in a nose down pitch attitude, about 7,650 feet from the threshold. There was no fire. Company procedures required the flying pilot (the captain) to give an approach and landing briefing to the nonflying pilot (first officer). The captain did not brief the first officer regarding the thrust reverser's MEL'd status, nor was he specifically required to do so by the company operations manual. Also, the first officer did not remind the captain of its status, nor was there a specific requirement to do so. The operations manual did state that the approach briefing should include, among other things, "the landing flap setting...target airspeed...autobrake level (if desired) consistent with runway length, desired stopping distance, and any special problems." The airline's crew resource management procedures tasked the nonflying pilot to be supportive of the flying pilot and backup his performance if pertinent items were omitted from the approach briefing. The maintenance, repair history, and functionality of various components associated with the airplane's directional control systems were evaluated, including the brake system, the nose landing gear strut and wheels, the brakes, the antiskid system, the thrust levers and reversers, and the throttle control unit. No discrepancies were found regarding these components.

The National Transportation Safety Board determines the probable cause(s) of this accident as follows:

Date:	09 JUL 2006
Time:	ca 07:50
Type:	Airbus A.310-324
Operator:	S7 Airlines
Registration:	F-OGYP
C/n / msn:	442
First flight:	1987-04-03
Engines:	2 Pratt & Whitney PW4152
Crew:	Fatalities: 5 / Occupants: 8
Passengers:	Fatalities: 120 / Occupants: 195
Total:	Fatalities: 125 / Occupants: 203
Airplane damage:	Written off
Location:	Irkutsk Airport (IKT) (Russia)
Phase:	Landing
Nature:	Domestic Scheduled Passenger
Departure airport:	Moskva-Domodedovo Airport (DME/UUDD), Russia
Destination airport:	Irkutsk Airport (IKT/UIII), Russia
Flightnumber:	778

Weather around the time of the accident (23:00 UTC / 08:00 local) was: UIII 082300Z 28005MPS 3500 -SHRA OVC006CB 11/09 Q1002 NOSIG RMK QBB190 QFE707/0943 30290250= (Wind 280 degrees at 5m/sec visibility 3500m, light rain showers, 8 oktas overcast cloud at 600ft with thunder clouds, temperature 11C dewpoint 9C, QNH 1002hPa no significant weather)

"When he pressed the switch for the reversing system, located between the pilots' seats, with one finger of his right hand, Shibanov most likely bumped the handle that controlled the left, deactivated engine, located only centimeters away, with his other fingers. As a result, he simultaneously turned on the right reversing system and left takeoff system and the plane picked up speed, turning to the right, hitting garages and bursting into flames.

Pilots questioned by Kommersant say that the cause of the crash should be attributed to the error of the dispatcher and the unfortunate design of the plane's cockpit: the controls for the throttle and the reversing system can be pressed simultaneously when reaching for only one of them, especially in an emergency situation. On an aircraft of that class, it would be possible electronically to prevent the activation of the forward throttle while turning on the reversing system."

http://www.kommersant.com/p724140/Irkutsk_A310_Crash/

Status: Final Date: 28 AUG 2002 Time: 18:44 Type: Airbus A.320-231 Operator: America West Registration: N635AW C/n / msn: 092 First flight: 1990 Total airframe hrs: 40084.0 Cycles: 18530.0 Engines: 2 IAE V2500-A1 Crew: Fatalities: 0 / Occupants: 5 Passengers: Fatalities: 0 / Occupants: 154 Total: Fatalities: 0 / Occupants: 159 Airplane damage: Written off Location: Phoenix-Sky Harbor International Airport, AZ (PHX) (United States of America) Phase: Landing Nature: Domestic Scheduled Passenger Departure airport: Houston-George Bush Intercontinental Airport, TX (IAH/KIAH), United States of America Destination airport: Phoenix-Sky Harbor International Airport, AZ (PHX/KPHX), United States of America Flightnumber: 794 Narrative: Flight 794 departed Houston with a an inoperative nr. 1 thrust reverser. On August 20, 2002, the number one thrust reverser had been deactivated by maintenance personnel. The airplane touched down at Phoenix on the centerline of runway 08 about 1,200 feet beyond its threshold. During rollout the captain positioned both thrust levers into reverse but then took the number one thrust lever out of the reverse position and inadvertently moved it to the Take-Off/Go-Around (TOGA) position, while leaving the #2 thrust lever in the full reverse position. Full left rudder and full left brake application did not compensate for the yaw. The airplane continued swerving to the right until exiting the right side of the runway. It crossed the apron east of intersection B8, and experienced the collapse and partial separation of its nose gear strut assembly upon traversing the dirt infield area south of the runway between intersections B9 and B10. PROBABLE CAUSE: "The captain's failure to maintain directional control and his inadvertent application of asymmetrical engine thrust while attempting to move the #1 thrust lever out of reverse. A factor in the accident was the crew's inadequate coordination and crew resource management."

from this link

Date: 22 MAR 1998 Time: 19:41 Type: Airbus A.320-214 Operator: Philippine Air Lines Registration: RP-C3222 C/n / msn: 708 First flight: 1997 Total airframe hrs: 1224.0 Cycles: 1070.0 Engines: 2 CFMI CFM56-5B4 Crew: Fatalities: 0 / Occupants: 6 Passengers: Fatalities: 0 / Occupants: 124 Total: Fatalities: 0 / Occupants: 130 Ground casualties: Fatalities: 3 Airplane damage: Written off Location: Bacolod Airport (BCD) (Philippines) Phase: Landing Nature: Domestic Scheduled Passenger Departure airport: Manila International Airport (MNL/RPLL), Philippines Destination airport: Bacolod Airport (BCD/RPVB), Philippines Flightnumber: 137 Narrative: Flight PR 137 was a regular scheduled passenger flight and departed Manila for Bacolod at 18:40. The airplane departed with the thrust reverser of engine nr.1 inoperative. At 19:20, PR137 called Bacolod Approach Control and reported passing FL260 and 55 DME to Bacolod . The crew then requested landing instructions and was instructed to descend to FL90 after passing Iloilo and descend to 3,000 ft for a VOR runway 04 approach. Wind was 030° at 08 kts, altimeter 1014 mbs, transition level at FL60 and temperature at 28°C .At 19:28, the flight requested to intercept the final approach to runway 04 and Approach Control replied "PR 137 visual approach on final" . At 19:37, Bacolod Tower cleared the flight to land at runway 04 and the clearance was acknowledged by the pilot. The approach was flown with the Autothrust system engaged in SPEED mode. The thrust lever of engine no.1 was left in Climb detent. Upon touchdown the first officer called out "no spoilers, no reverse, no decel". Engine no.2 was set to full reverse thrust after touchdown, but the engine no .1 thrust lever was not retarded to idle and remained in the climb power position. Consequently, the spoilers did not deploy. Because one engine was set to reverse, the autothrust system automatically disengaged. With the autothrust disengaged, nr.1 engine thrust increased to climb thrust. Due to the asymmetrical thrust condition, the A320 ran off the right side of the runway. At this speed, rudder and nosewheel steering are ineffective. Engine no.2 was moved out of reverse up to more than 70 percent N1 and the airplane swerved back onto the runway. The A320 continued past the runway end. The aircraft hit the airport perimeter fence and then jumped over a small river. It continued to slice through a hallow block fence where it went through several clusters of shanties and trees. No fire ensued after the crash. PROBABLE CAUSE: "The probable cause of this accident was the inability of the pilot flying to assess properly the situational condition of the aircraft immediately upon touch down with No. 1 engine reverse inoperative, thereby causing an adverse flight condition of extreme differential power application during the landing roll resulting in runway excursion and finally an overshoot. Contributory to this accident is the apparent lack of technical systems knowledge and lack of appreciation of the disastrous effects of misinterpreting provisions and requirements of a Minimum Equipment List (MEL).

from this link

Reference the thrust levers on the Bus...
it's not a single piece, it consists of 2 things.
1. The thrust lever
2. Reverse latching levers.

To actuate thrust reverse:
Thrust lever to idle
pull the reverse latching levers... this actuates idle reverse
Pull thrust lever to the rear for more reverse thrust.

Requires both main landing gears to be compressed for a full On The Ground detection, AND
A thrust reverse signal fed through at least one of the Spoiler Elevator Computer. The signal from the SEC opens the reversers hydraulic shut off valve to prevent its deployment without this signal.

Actuation of reversers will have FADEC commanding IDLE. Once IDLE is attained, it will go to idle reverse, complete the reverser sleeves deployment before going beyond idle reverse.

If we go with a one reverser inop scenario, it's actually pretty simple.

Sequence:
1. Pilot selects Reversers on 1 or both engines
2. Pilot then selects a go-around.

On the reverser actuation, refer to the above. BUT, 1 engine remains on forward idle.
On go-round selection:
the engine with the reverse inop will go straight to TOGA
the engine with the operating reverser will:
   1. Select thrust to TOGA and stow the reverser sleeves.
   2. Idle Thrust Protection will immediately kick in because no reverse thrust is selected and the sleeves are still in transition to stow.
   3. Idle thrust will remain on the engine despite TOGA power selection until no HYD pressure is detected downstream of the Hyd Control unit on the reverser.

This theoretically should only take moments, but ideally, one should select idle, check the reversers have been stowed and no pressure is detected downstream of the reverser HCU (Reversers stowed indicated on the ECAM), then go to TOGA. However, in emergencies, this should not be a problem because the FADEC and reverser mechanism should take care of it... however, the delay even if only for a second, could result in asymmetrical thrust.

This is why on Boeings (and Buses), it is not recommended to go-around after reversers have been deployed because, in the haste of things, the pilot could select TOGA before the reversers are stowed and HYD Press gone to zero, which puts a risk on the engine at TOGA on accidental reverse deployment due to hyd pressure being present downstream of the reverser control unit. Should it be necessary, the safest way is to wait and confirm on the engine instruments that the reversers have been stowed prior to going to TOGA, both on Bus or Boeing. This is why people and companies have tried to keep it simple, i.e. no go-around after reverser deployment. It's just too risky.

Up to five seconds are required for a reverser to close in the forward thrust position. Add to that another 5-8 secs for a spool up to max thrust.

And this makes interesting reading...
http://www.smartcockpit.com/data/pdf...yingtechnique/Slippery_Runways.pdf

now......on the Subject of the Wheelbraking Anomalies Experienced by the Airbus family

(The Cause of the Go-Round Decision??)

On the A320 there is a recall drill (not on ECAM) for 'Loss of Braking'? It resulted from the A320 overrun accident at Ibiza (link). It relates to no-notice BSCU failure (Brakes, Steering and Control Unit) and requires prompt recall action by the crew..... as there's nil braking available.....

It requires quickly switching the A/Skid & N/W Strg switch off and then manually braking with max 1000psi pressure. (see page 50/105 at the above link). However, as the toe-brakes on the rudder pedals don't go flat to the floor in this case (under manual braking) and the deceleration due to two-engine reverse approximates the retardation from a low autobrake setting, it's not always clear, early on, that braking has failed.

If that A/Skid & N/W Strg switch cycling proved unsuccessful, then you were supposed to use short and successive applications of the park brake.

It became a published procedure, so I presume that it has happened to the extent that Airbus had to devise a new FCOM procedure.

"As our BSCU components continue to age, we are seeing a higher frequency of BSCU Channel Faults (X) almost exclusively on the older style Conventional BSCU PN C202163382D32. In most cases 86% of the time, these faults are simply corrected by cycling the A/SKD and N/W STRG switches off, then on, or with a computer/circuit breaker reset."

"Analysis showed that it took 10 to 13 seconds for the commander to recognise the lack of pedal

One hypothesis: Reverse thrust was selected on Eng 1 (eng 2 rev thrust was deactivated). Pilots realized they would not be able to stop the aircraft and, contrary to the recommendations of the FCOM (Flight Crew Operating Manual) to always execute a full stop once reverse thrust (TR) is selected, initiated a go-around by pushing the thrust levers to TOGA. An engine going from full rev to full forward thrust could possibly suffer a compressor stall while the reverser doors are closing, causing the flash we see in the video.

Of course, they may have considered it was safer to try to get airborne again... if they had experienced a BSCU-induced wheel-braking failure. The problem may then have become the same thrust asymmetry of  the accidents of registrations: N635AW and F-OGYP (the two incidents in cells 2 and 3 above).

- 17:03 LT - GOL B 737-800 lands at CGH and reports "slippery runway"
- 17:04 LT - INFRAERO, local airport manager, closes the airport for landing and takeoff ops. INFRAERO engineers measures the standing water: 0,6 mm.
- 17:30 LT - Airport returns to normal ops.
- 18:50 LT - PR-MBK lands at CGH, overruns and crashes.

Remark: Between 1730 and 1850, 40 aircraft (the majority of them were airliners - B 737 and A319/320) landed at CGH and no problem was reported. This increases the likelihood of it having been attributable to a technical malfunction on TAM's Flt JJ3054. Judging by the water spray seen on the video, it would appear that its #1 Rev was deployed and probably at full thrust. If a go around was attempted after TR deployment on #1, which is highly not recommended of course, the right engine would spool immediately and the left would have to come out of reverse and then start to spool up well after the right engine. The results of the thrust asymmetry would probably follow the tracking on the Google earth picture posted below which is a faithful illustration of what the security cameras showed.

Alternatively The flash at the end of the video might have been a tire burst, LH engine ingesting fence bits or, more likely, a compressor stall due to TOGA power being applied rapidly while the LH engine's reversers was still stowing. Part of the aircraft did graze a car on the road below.

TAM has now put its A320 MEL on their public website, probably in response to the press coverage of the T/R issue

http://www.taminforma.com.br/noticia.aspx?id=1497