PPRuNe Forums - View Single Post - AS332L2 Ditching off Shetland: 23rd August 2013
Old 6th Sep 2013, 11:26
  #1304 (permalink)  
HeliComparator
 
Join Date: Aug 2004
Location: Aberdeen
Age: 67
Posts: 2,090
Received 39 Likes on 21 Posts
Metcha, the sensors used are common to the screens and the autopilot. Pitch, roll, heading, slip and vertical speed all come from the two AHRS - normally configured so that AHRS 1 is displayed on the LH set of screens, AHRS2 on the right. They are constantly cross checked at various points and any discrepancies between them are flagged up by the appropriate screen zones going amber.

However, the autopilot is very clever and doesn't just blindly use the pitch, roll, HDG etc data. It looks at the raw pitch rate, roll rate, yaw rate and tri axial accelerating coming direct from the AHRSs fibre optic gyros and accelerometers and calculates its own values of pitch, roll and heading and vertical speed. Long term, it will "like" one of the two AHRS and slowly wash out any integration errors so that it values merge. However, if an AHRS throws a wobbly, it will ignore it and just look at the other one (and the standby AHRS) quite seamlessly.

Air data similarly is used by both the screens and the autopilot, LH Sensor for LH pilot etc. the autopilot will in general use the air data from the pilot flying's side unless it fails or is unreasonable, in which case it will seamlessly use the other side. Again, there is cross checking between the two air data sensors at several levels, and the any discrepancies are made very visible to the pilots.

Nav data is a little more simplistic - the selected Nav data is displayed by the screen and used by the autopilot according to which screen is coupled. But of course there is automatic cross checking and any discrepancy between say ILS 1 and ILS 2 is flagged up. The autopilot is very clever in its use of Nav data - it hybridises it with inertial data so that, for example with a brief loss of ILS signal, the flight path is unaffected.

There is a degree of hybridisation of the air data too, but since there isn't necessarily a direct correlation between air data (say airspeed) and inertial data (ie ground speed), due to gusts for example, the inertial component is not that great. Of course its greater for altitude data since sudden pressure changes don't happen in the way that sudden wind changes do.

Phew, a bit complicated as you can see, and its all very clever! The bottom line for the pilots though, is that it just works, and works very precisely and intuitively.
HeliComparator is offline