AF 447 Thread No. 11

[Chris Scott]

A340 AIRPROX and AF447

Lonewolf_50 and respondents,

You discuss a period of 17 seconds before the first sidestick intervention. Re the AAIB Bulletin into this incident, I still find the description on Page 11 ("A340 Data") confusing, and the time scale on its DFDR trace virtually unreadable on the version I have. However, it states clearly enough (my underscoring):
"The first recorded sidestick input was made at 14:22:08 which was some 28 seconds after the commencement of the Master Warning."

The Master Warning and CRC (continuous repetitive chime) coincided with the AP disengagement (due to MMO exceedance, mainly caused by a sudden fall in temperature), but the aural warning (cavalry charge) was delayed because the CRC and a TCAS warning took priority. 10 seconds after the Master Warning, an up-gust briefly increased the AoA towards alpha-prot. That resulted in a phase-advanced engagement of AoA (protection) Law, after which the EFCS was tasked to reach and maintain alpha-prot, a much higher AoA than desired. That started the pitch-up, and after (roughly) a further 8 seconds the aircraft started climbing in earnest. After that, it was about 10 seconds until the first sidestick input.

So I presume your 17 seconds is the crucial period between engagement of AoA Law (+ Pitch-up) and the first (rather tentative) sidestick movement? This was less than half down, and it is unclear if it was sustained for the period of more than one second that was required to disengage AoA Law. The first application of more than half down-stick (which would disengage AoA Law instantly) started about 5 seconds later. It is only then that the a/c starts to pitch down rapidly, i.e., 22 seconds after the engagement of AoA Law.


RetiredF4 and DonH have identified the likely reason for the lack of sidestick action by either pilot. To be a bit more more specific, one or both pilots were probably trying to re-engage one of the two APs, possibly trying them both alternately. Many of us have seen this kind of behaviour: priority given to re-engaging the automatics, rather than flying the aeroplane.

This zoom climb involved an error of omission by the PF; AF447's an error of commission.

As I wrote a few days ago:
"The common factor this incident had with the much later AF447 accident is that of the unexpected transition from automatics to "manual": specifically from AP to FBW, in a pitch law other than Normal. In this AIRPROX case, the FBW system forced the a/c into an undesired zoom-climb which needed positive crew intervention to counter. In AF447, the FBW made no such input; the PF forcing the a/c into a zoom-climb by inappropriate sidestick inputs."

In both cases, there were extenuating circumstances - if only minor ones.

Most of us are in agreement that, had the AF447 PF sat on his hands for a while, the A330 would have maintained a good pitch attitude at least. In the A340 AIRPROX case, however, we are all (myself included) chiding him/her for failing to prevent the EFCS (FBW) from pitching the aircraft into a zoom-climb which - in the absence of any collision - would have become an embarrassing phugoid. So although the initial parts of the "zoom climbs" are rather similar, their causes are very different.

In this case, at least two unusual met phenomena and the logic of the EFCS conspired in such a way that PF inactivity was not an option - if flight separation was to be assured. The AAIB avoids censorship, as follows:
"Such was the vigour of the A340’s climb in AoA law, the aircraft could well have climbed through FL 363 (thus provoking a TCAS RA with revised software version 7.0) in a very short time, even if the crew had applied nose-down sidestick as soon as they heard the (delayed) autopilot disconnect warning."