Menu Sign In Contact FAQ
Banner
Welcome to our forums

Trusting instruments

Peter wrote:

It is easier to hold an attitude when it is a big display, with a brown bottom half and a blue top half, for sure. But if you get a failure, the plane will be very hard to fly using the backup instrument(s).

The popular backup seems to be a G5, which is quite independent. But as you say a G1000 (or other integrated system) fault grounds the aircraft, because although you might continue and land on the G5 you aren’t going to take off on it.

My RV-7, when I get as far as the panel, will have some glass (perhaps a G5 or whatever the similar product is then) but no integrated system where a single failure grounds the aircraft. Each instrument will be swappable by me, because being at the mercy of a dealer or manufacturer is not part of my philosophy.

EGLM & EGTN

I should also add that when flying in IMC, the autopilot is used 100% of the time. So you are sitting there, monitoring everything, and should spot an AI failure in good time.

The time the autopilot isn’t used is when practicing IMC hand flying, or when practicing or training for an IMCR or IR revalidation In the latter cases, switch to manual flight at least 10 mins before flying an approach.

Some people have installed an electric AI right next to the vacuum AI. It looks a little odd but should make any problem immediately obvious.

Does each G5 have its own internal AHRS?

Administrator
Shoreham EGKA, United Kingdom

Peter wrote:

Does each G5 have its own internal AHRS?

Yes.

ESKC (Uppsala/Sundbro), Sweden

This is why the old Stark 1-2-3-method of recovery should be studied and kept fresh.

1. Center turn indicator with rudder only.
2. Center ball with ailerons only.
3. Center climb indicator (VSI) with elevator only.

@AdamFrisch have not heard of this method. LOC in the training world is in a bit of a muddle. On the IR you have techniques for recovery of unusual attitudes on a limited panel, which are different to the techniques taught on A-UPRT. A-UPRT is strictly an introductory module to a swept wing jet type rating, although schools shoe horn the course somewhat randomly, either during CPL phase or before the IR.

One of the CRM case studies on A-UPRT is AA587, which would advise you not to use the step 1 of this Stark technique. Even on a straight wing aircraft using a turn coordinator, once you are past neutral stability, using rudder to get wings level will not restore straight and level, you need to roll opposite rate one, and then roll level. Emphasis on roll. The rudder, outside of a snap roll, is not going to get you from inverted to upright straight and level. I realise there are differences between a T&B and a TC, and while the T&B is superior in spin recovery, and I expect this method may be linked to spin recovery (although step 3 is pro spin!), both instruments are most useful in a less than rate one turn environment.

In addition the VSI is a lagging instrument (even a IVSI will be prone to some lag in an unusual attitude). The vector change on the IAS being a more reliable secondary indicator of pitch attitude, so am not sure I would use step 3.

I also don’t understand step 2, unless the type is uniquely immune to adverse yaw. Would this approach be correct if you are asymmetric, for example?

Last Edited by RobertL18C at 19 Oct 11:33
Oxford (EGTK)

Yes, a Turn and Bank indicator is a better instrument than a TC for this purpose, where the two get muddled together. Howard Stark based his original Instrument Flying on that instrument.

When you eliminate acceleration forces, the ball is just a spirit level.

I think there is a speed (x*VS) & angle of bank (beta) where rudder can roll the wings but away it won’t: near VS, rudder is effective but dangerous, near VNE, rudder is useless and may snap

When flying gliders in clouds (legal in UK and there is a formal rating), we used to trim for near 1.3*VS keep wing level with rudder, power off with full open air-breaks and hand off the yoke except to correct speed excursions with pitch only, this assumes one has access to a working ASI (can get frozen) and working turn coordinator (battery can go flat but you have HDG/TRK or AHRS depending on the toy on your screen, some have even an electric AI)

I went trying that in VMC and IMC, things may work and they also go badly, the above works fine in Astir but go nuts in Nimbus, it also works fine in C172/Arrow with gear down (happy to bet my life on it eyes closed), it does not work satisfactorily in DA40 (will not bet my life on it), it goes wild in SR20/M20J

Whatever one thinks on how they will deal with LOC in IMC from reading or hearing from instructors, the stuff has to be tested first in VMC & IMC before getting classed as serious “wild card”, this includes: flying techniques, any plans to fly on backups, partial panels…for instance, I am 100% sure I can’t fly hand fly an ILS down to 400ft and go-around on Turn Coordinator, I went practicing once and had dead AI after takeoff in 500ft ceiling and gusty day, being current as hell those days I still took the opportunity to try it, conclusion: it’s too wiggly for me and I don’t think I will be up to that challenge again, a better plan would be: go find VMC, or drop aircraft on the beach using GPS, or just fly timed headings to long 4km runways and hope for the best…

Last Edited by Ibra at 19 Oct 12:43
Paris/Essex, United Kingdom

AdamFrisch wrote:

This is why the old Stark 1-2-3-method of recovery should be studied and kept fresh.

1. Center turn indicator with rudder only.
2. Center ball with ailerons only.
3. Center climb indicator (VSI) with elevator only.

I have heard of this technique but it seems very strange to me. How does it work with a TC rather than a T&B?

ESKC (Uppsala/Sundbro), Sweden

Peter wrote:

But it is fairly obvious it has gone

This is my concern with mechanical AIs – it isn’t always obvious. The failure mode I wrote about in my OP would have caused serious confusion for anybody if one were taking off into a low cloud base.

Peter wrote:

“the big cross”

That’s a very obvious failure mode, if you see that cross you’ll know immediately to disregard the PFD and go to your standby instruments.

United Kingdom

if one were taking off into a low cloud base.

Then one must be very careful and watching two gyro instruments concurrently. Also the transition from outside to instruments must be done very fast. In say OVC001 one has only a few seconds.

But I think that if an AI remains upright through turns, it is working OK for flight. An AI with a stationary flywheel will flop about in turns.

In my TB20 there are up to five “AI” instruments; 3 on the left and 2 on the right. Two share the SG102 AHRS data; the rest are separate. The only ones which never gave trouble are the electric TC and the electric AI… The KI256 has also never failed in flight; it has done this and nobody knows the reason.

if you see that cross you’ll know immediately to disregard the PFD and go to your standby instruments

What you don’t know is how far the software has permitted the attitude presentation to degrade before replacing it with the cross. A PFD normally uses an off the shelf AHRS module or at least off the shelf accelerometers / rate gyros (nobody makes their own ones; they normally come from Analog Devices or some such – see e.g. here and an AHRS will have 3 of each) and these probably won’t give a ready clue when they fail. So the software has to try to work out when the data is no longer consistent. I don’t think this can be done instantly; it has to use something like too big a heading rate of change (too fast a yaw) for a physical aircraft, the gravity vector being too far away from +1G in the absence of pitch/roll/yaw changes, etc.

Administrator
Shoreham EGKA, United Kingdom
Sign in to add your message

Back to Top