Flight International, 27 March - 2 Paril 1996, page 13, contains some comments by David Learmount on the CVR transcript additional to those contained in the DGAC press release of 18 March 1996. There are two air-data computers (ADC) on the B757. The captain's instruments use the Left ADC and the First Officer's instruments the Right ADC. According to the B757 manual description of the Air Data System, the instruments may be manually switched to use the other ADC as 'alternate source'. The phrase 'alternate source' has operational significance to pilots of all stripes - in case of pitot-static system malfunction, one is trained to use the 'alternate source' of pitot-static information.

Learmount notes that the center autopilot is slaved to the captain's (Left) ADC. In Boeing document C22-10-711-02 (the B757 systems description), on page 46 of Section Autopilot General, the section FCC/Air Data Computer Interface, Paragraph 1 reads (in full):

1. General

There is a captain's and a first officer's alternate select switch relay which are controlled by their associated air data instrument source select switch. These relays control the digital bus inputs to the left, center and right flight control computers (FCC).

The left and center FCCs receive normal data from the left air data computer (ADC) bus No. 1 and alternate data from the right ADC bus No. 4. The right FCC receives normal data from the right ADC bus No. 1 and alternate data from the left ADC bus No. 4.

Some salient points concerning crew behavior are:

• the captain engaged the center autopilot, which (we may surmise he should have known) is slaved to the left air-data computer which feeds his own ASI, even though he had already decided that his ASI was providing faulty information;
• he didn't use the alternate source of pitot-static information after he had decided that his ASI information was faulty;
• the crew seemed to be confused by two advisory messages, when no response was required for safety;
• when the copilot said that his ASI showed that speed was reducing, the captain decided that the co-pilot's ASI was also malfunctioning;
• at no time does it appear they consulted checklists or manuals, or compared pilot's and co-pilot's ASIs with the standby ASI, or used the alternate pitot-static source for the pilot flying (the captain).

Specifically, Learmount says:

The Cockpit-Voice recorder [...] reveals a picture of pilots who, faced with minor problems, became confused and lost control of a flyable aircraft [...].

[After the crew had remarked that the captain's ASI was not working and elected to continue the takeoff, they] did not consult checklists or manuals and, points out the report, neither did they compare the main instruments with the standby ASI. [...]

The captain flew the take-off. Almost 2min later, the flaps had been retracted and the captain ordered the center autopilot engaged. This is normal procedure - but the center autopilot is slaved to the air-data computer which senses airspeed from the captain's ASI. The captain's ASI, as height increased, began to read a speed of 350kt which was far faster than the real airspeed. As a result the autopilot/autothrottle acted to increase pitch-up attitutde and reduce power to lower the speed.

At this point two "advisory" messages [rudder ratio and Mach/airspeed] appeared on the engine-indicating and crew-alerting system. 757 pilots say that no response was essential for safety, but the crew's reaction indicated confusion.

Meanwhile, the co-pilot stated that his ASI was reading 200kt "decreasing". The captain's response indicated that he thought the co-pilot's ASI was now also wrong. Stall speed at that configuration is close to 160kt. The overspeed warning, slaved to the captain's faulty ASI, sounded, but, within 22s, the stall-warning alert started as the captain brought the speed back, and the sound of the stickshaker [...] continued until impact. [...]

Back to top

Back to `Incidents and Accidents'.

Peter B. Ladkin, 1999-02-08
	Last modification on 1999-06-15 by Michael Blume