looking and learning
IMHO the incident report does not review the crew’s activities in context, and the analysis and recommendations warrant wider consideration.
A single IRS fault/failure tripped the Flight Guidance Computer (FGC). This computer provides AP, FD, AT, but also, altitude alerting, auto pitch trim, flap trim compensation, and electric pitch trim.
When selecting flap between 0 and 18 (and vice versa) there is a significant change in pitch trim, which with an unpowered control system is felt directly on the control column. Normally flap trim compensation alleviates this force; the crew can assist / override this with electric trim. With a FGC failure, manual trim may be required, which for the lower flap angle selections requires quick and ‘extensive’ trim wheel movement; even with electric trim (but no FTC) timely action is important.
Thus inadequate/late trim application could have contributed to the altitude deviations / pitch attitudes, and together with the primary instrument and other failures, IMC, and coincident ILS join, then the flight path might be as expected. This failure situation is unlikely to have been practiced, particularly with disorienting/distracting factors (attitude mismatch/failure), bank attitude warnings (warranted/unwarranted), loss of altitude alerts, and P2 flying.
The apparent knowledge weaknesses of ATT/HDG caution (yellow) and cross side compass display might be understandable in the stress of the situation, which together with potentially disorientating bank angle calls contributed to P2’s doubts.
With an unusual technical failure combined with aircraft reconfiguration and manoeuvring for the approach, the crew could have been unprepared for the trim change and perhaps were situational disorientated by the flight deck indications. Situation awareness suffered, logical assumptions were made, which in hindsight have been questioned; but why should a crew seek deeper analysis if their mental model associated the latter incident with the first, particularly where the main indications were the same.
Unsatisfactory CRM is a glib comment without qualification. With hindsight, workload management might have been better. However, with P2 unsure of his displays, the Captain’s choice is to either fly and communicate, with P2 diagnosing, or put the workload on P2 with diagnosis and communication; however in this incident communication was a relatively high priority – inability to navigate and that ATC were very helpful; this is a judgement call and it would be unfair to criticise without experiencing the entirety of the situation.
Why recommend a technical change to the stby attitude instrument when it meets the certification requirements and has not been faulted over 20+ years in this and other aircraft types.
What parallax (#34); the instrument is satisfactory for emergency use (abnormal crosscheck #36). We should not expect pilots to pass an IRT with it and the resultant extended scan pattern.
Why should instruments fail more often with age of design; the current rate, although less than electronic versions, has been judged satisfactory.
The training recommendations will probably result in more practice with stby instruments, but this overlooks the significant point of the aircraft handling with abnormal trim operation - FTC failure. The crew need to have a ‘feel’ for the aircraft in this unusual configuration.
Not every situation can be simulated; realism in surprise and stress is very difficult.
We might teach ‘CRM’, workload management, and surprise management, in the class room, but who can assure that the correct behaviour will be recalled in actual situations. This begs the question if the operational scenarios in combination with rare events are extending human performance to the limit – perhaps it’s unrealistic to ask any more from the individual. The industry might be reaching a balance where cost of restricting operational workload matches additional training costs.
This was a serious incident when judged against the high standards of today’s excellent safety record.
The outcome was safe; it may not have been a tidy execution, but it was just within the limits of human performance (those particular individuals, in that particular situation).
The industry might learn from this; what went right, why. Also note the similarities with recent accidents – the role of trim and aircraft ‘feel’, the need to fly attitude not computations or feel/stick position, and difficulties of situation awareness / assumptions – both by the crew and the organisational system.
Is the industry assuming too much about the effectiveness of training, about crew behaviour, knowledge retention/recall, or the extent of / ability to gain experience of unusual combinations of events.
This was not a blame and train accident; it’s for looking and learning.
Last edited by alf5071h; 18th Dec 2012 at 22:35.
Reason: typo