Radalt / TAWS requirements

[Genghis the Engineer]

I've been wading through the books trying to get my head around TAWS / EGPWS system requirements.

So far as I can tell, the regs require a TAWS, at A or B level functionality for aeroplanes or helicopters over a given size. It is normal to include a radalt feed into the system, but not strictly required - although so far as I can tell all aeroplane systems and most(?) helicopter systems do include a radalt feed.

Also so far as I can tell, there's no requirement for duplication of any part of a TAWS system, including of the radalt?

Do I read this correctly, and if I'm wrong, is anybody able to point at the regulations that prove me wrong?

[Genghis the Engineer]

Thanks IGh, that's clear enough about how it works - do you happen to know if there's any regulatory requirement for anything to be duplicated within that system?

[safetypee]

G, from a regulatory aspect (ICAO Annex 6) and National Authorities Ops Reqs; all that they required is to have the appropriate level of TAWS fitted.
Technical requirements are in AC25-23 which outlines the manner in which TAWS should work, and also in TSO's (TSO 151b?).
My knowledge is dated, but I would be very surprised if any TAWS ‘A’ would be certifiable without Rad Alt.
TAWS ‘B’ might be with the use of GPS, see CAA Letter of Consultation 6 Nov 2002, but most documentation suggest that Geometric Altitude is not accurate enough.

[WeekendFlyer]

If I recall correctly there are some TAWS alerts that can trigger on approach, e.g. high sink rate or too much bank angle below a certain height. These may depend on a Radalt input to ensure they are accurate and, perhaps more importantly, don't get triggered falsely. Baro/Inertial/GPS height may not be good enough by itself, particularly when descending in the landing configuration (as you probably know, on some aircraft types with high throttle settings + full flap/slat and gear down are not usually as amenable to static pressure error correction as they are in the cruise configuration). The addition of Radalt data in to the mix improves the accuracy in the height solution for the final phase of flight.

Having said that, given that TAWS almost certainly uses some form of Kalman filtering to fuse the data from the different height sources, I would suspect loss of the Radalt would not be a major problem. It would just reduce the accuracy slightly when operating close to the ground, particularly on approach, so the system should respond by increasing the alert thresholds a bit. It might increase the false alarm rate slightly but I doubt it would be unsafe in any way. Therefore I suspect one Radalt would be sufficient.

As to specific regulations and requirements, I would have a good look through the EUROCAE and RTCA websites and FAA ACs to see if they have a specific MASPS document or similar (if you haven't already!).

[Genghis the Engineer]

Thanks all, that's really helpful.

Just when you think you understand aviation technology, something new jumps out at your and you need to ask some dumb questions just to get a basic understanding of which stones to look under. That really is where Pprune is at its best !

[aterpster]

WeekendFlyer:

If I recall correctly there are some TAWS alerts that can trigger on approach, e.g. high sink rate or too much bank angle below a certain height. These may depend on a Radalt input to ensure they are accurate and, perhaps more importantly, don't get triggered falsely. Baro/Inertial/GPS height may not be good enough by itself, particularly when descending in the landing configuration (as you probably know, on some aircraft types with high throttle settings + full flap/slat and gear down are not usually as amenable to static pressure error correction as they are in the cruise configuration). The addition of Radalt data in to the mix improves the accuracy in the height solution for the final phase of flight.

The original ground proximity warning system that the FAA mandated as a result of the 1974 TWA 514 CFIT crash, used only a radar altimeter and aircraft configuration. That remained unchanged until the AAL Cali, Colombia CFIT accident in 1995.