Failure Indications

[jumbojohn]

I am writing an essay on "Failure Indications"

I believe that there is more to this than first apparent.

Professional pilots normally run through a safety briefing before flight, which might start something like - "Up to V1 we will stop for any fire, engine failure, take off config warning, windshear warning or monitor radar display caution" (possibly for predictive windshear warning equiped a/c)

But how is an engine failure detected by the crew? An engine failure might be indicated by a bang, swing, drop in EPR, drop in EGT or even a rise in EGT if engine damage has occured. (My words) Maybe worth discussion in briefing/simulator?

The take-off config warning is normally an audio warning, I think the last time I heard it was on a convertsion course 7 years ago.

On the a/c I fly (Boeing) EGT exceeding limits (not on start) requires the thrust lever to be closed then into the engine surge checklist not an engine fire, severe damage or seperation memory checklist.

Many simulators are not fully equiped for the variouse windshear scenarios to be practiced on takeoff so maybe pilots are not fully familiar with exactly what might be displayed and heards in the event.

In airline training we often discuss and practice various failures and emergencies but sometimes these problems are indicated quite subtly and sometimes lead us the wrong way.

Comments appreciated please Ladies and gentlemen.

[Lu Zuckerman]

During the initial design phase of the aircraft a Reliability engineer will perform a Failure Mode Effects Criticality Analysis (FMECA) on each system. It is imperative that this is done prior to the finalization of any part of the design. The FMECA will define the effects on the systems of an aircraft and if and how the effect can migrate up to the aircraft level. He will also show the method by which the pilots are made aware of the failure, which drives the design of the monitoring and display systems. This in turn drives the development of the operating and trouble shooting provided in the operations handbook as well as the troubleshooting guides for the maintainers..

Now, let’s enter the real world. Because of monetary constraints Reliability, Maintainability, and Systems Safety analyses are performed after the design has been frozen. So the FMECA will reflect the design and not influence the design. In this situation the pilots and maintainers are left holding the mucky end of the stick when a problem occurs. If the engineers did not think of it, it is not incorporated in the design until there is an accident.

:E

[GlueBall]

Even without checking the engine instruments, the most obvious and immediate engine failure or loss of thrust "indication" prior to V1 would be an adverse yaw towards the side of the failed engine; Unless in the case of a trijet where if the center engine fails, without asymetrical yaw, the pilots would be alerted by at least one or two bright "Eng-2 Fail" warning lights on the glareshield panel. The "Eng 2 Fail" warning light(s) would normally illuminate if N2 has decreased below something like 80% after having been advanced above something like 83% during the takeoff roll (in ground logic with flaps set, etc).

[jumbojohn]

Has anyone had any real subtle failures?:confused:

[lomapaseo]

Has anyone had any real subtle failures?

There have been numerous instances of uncommanded roll offs in thrust.

Some of these went undetected by the crew resulting in aicraft upset and accidents. Some were in climb and some were in cruise.

[Lu Zuckerman]

How about an uncommanded retraction of the flaps during takeoff?

Happened on an Air Canada A-320 several years ago around this time of the year.

The pilots almost lost it.

How about the uncommanded extension of a thrust reverser? In this case the pilots did lose it and all aboard were killed.

There are many more. Some reported and some unreported.



:E :E

[alf5071h]

The Propulsion System Malfunction and Inappropriate Crew Response (PSM+ICR) project report covers many issues of engine failure indication. The report was issued jointly by AIA and AECMA; you should find relevant information in vol 1 at: http://www1.faa.gov/certification/aircraft/engine_malf_intro.doc

The above may not link to the report. Try –
http://www.flightsafety.org/members/serveme.cfm/?path=/fsd/fsd_nov-dec99.pdf
http://www2.faa.gov/nsp/nsp/eng_mal.doc

[safetypee]

jumbojohn

Detection of information comes from the collection of individual pieces of data which must be combined form a picture – the information or indications. For an engine failure ‘820 EGT’ is just data, ‘low N1’ is data, ‘No.2 engine’ is just data, but ‘No 2 engine, 820 EGT rising, N1 80% decreasing rapidly’, begins to form a picture – information. Note the use of value, direction, and rate. If the problem is an engine failure more information is still required to form a valid judgment; comparison with expected data (is 820 EGT over-temp, or is 80%N1 below the expected takeoff value). Comparisons must also be made with adjacent engines, aircraft motion and other warnings; much of the comparing process can be achieved graphically through instrument pointers or swept area displays. There may also be a process of pattern matching; is this condition normally seen? Which engine is unusual? Is this failure similar to that seen in a simulator? Note sound is an important cue, but very loud bangs can be misleading – see PSM+ICR report.
A good training tip for engine monitoring during takeoff: – instead of calling ‘power set’ call the actual setting i.e. ‘power set 94%N1 (or xx EPR). This embeds the normal or expected values into your brain and thus speeds detection of any future abnormal condition. In addition its good CRM because your call may alert the other crew member to a condition that was not expected or intended e.g. unplanned de-rate takeoff.

You should also consider reviewing the relevant sections of JAR 25. See what the certification standards require as opposed to what the operational side of the industry assumes; there is often a wide gap. Details as to what are appropriate indications and the expected levels of alerting effectiveness are given in supplementary documents (ACs) i.e steady lights vs flashing lights, audio, tactile etc. Aircraft certification is based on probability of failure and failure rate. For critical systems a higher level of integrity is required, with appropriate levels of warning (usually red); the crew should be immediately aware, take note, or take action. Failures in less critical systems are alerted at an amber level – crew awareness where action may be required. Failures in systems of low criticality may be annunciated or treated as information. There are also failures or circumstances for which there are no specific warnings or procedures.

Crews should be trained to deal with these levels of failure: First routine actions following low attention getting alerts (lights or instrument indications). Second, failures where alerts or audio / visual warnings are given. Crews should then follow appropriate procedures. And third, rare and sometimes un-annunciated or ambiguous events, and probably without any specific procedure or drill. Where these rare events occur (massive engine failure outside of certification assumptions) then crews have to use airmanship to solve the problem. Some problem solving is explained by having a deeper understanding of situation awareness – detecting (perception) of information and understanding (comprehension) and then taking an appropriate course of action. See the rather lengthy ESSAI scientific report: www.essai.net/introduction.htm

[alexban]

You reminded me of one simulator training session,when we simulated engine failure 4-5 kt before v1.We knew this will happen ,but it was very dificult to actually see it ,think and made the call before v1.
In real life is even harder.If you have a master caution on,you react faster.But I think,as a nonflying pilot ,you'll look not only in the cockpit.
At start ,for eg. I have a flow,like check oil pressure-N1-vib-EGT-speed,then one glance out,then again EGt,speed,etc.
There are many things to check in a short time,and I think it's possible to overlook some subtle failure.For eg oil pressure decreasing.Untill you'll get a master caution,or you'll see it,you can be well over v1.
' During takeoff,if EGT exceeds limits,after 80kt,do not retard thrus lever,untill 400'AGT '( FCTM-boeing)
Take-off config warning it's possible to appear later on take off (actually did- speed brakes lever moved slightly up due to vibrations)
indeed ' these problems are indicated quite subtly and sometimes lead us the wrong way.'sometimes your reflexes can give you wrong messages.On one simulator session we forgot to check the atis before take-off (after 3 hrs of ldg,goaround,ldg.....)
So the instructor gave us a crosswind of 40 kts right at liftoff.
We flew the plane,no problem,but our first reaction was :Engine failure ! what..ah....no!For few seconds we were sure we had an engine failure,despite no indications on the instruments.The CRM saved our
:mad:
So,yes,I think failure indications is something to talk about.

[Ignition Override]

My (US) airline has the world's largest fleet of a certain old technology plane. They have no lights that say "eng. failure" or "eng. flameout". You don't want to abort at very high speed for an EPR gauge fluctuating, if the N1, N2 and EGT are stable-these are self-powered gauges, whereas the EPR is powered by an AC bus. But if you are in icing conditions, you might be having trouble with the P2-P7 probes etc.

My company recently decided to reduce the abort/go speed to 80 knots, regarding annunciator lights, abnormal sounds etc which do NOT indicate a problem with the airplane's ability to takeoff and safely climb out.:uhoh:

[alf5071h]

Ignition Override.

You company’s arbitrary revision of the “abort/go speed to 80 knots” is potentially hazardous. Not only have they introduced a new decision / check – to assess if the indications / cues are true / false or engine related / otherwise, they have created a unsafe operation in the event of loss of power and a go decision. There appears to be many opportunities for human error due to flawed judgment, misleading indications, etc. Have they / you considered the minimum control speed on the ground? You may decide to go at 80 kts – the aircraft appears capable of flight, but with asymmetric thrust you will only go off the side of the runway.
Please send a copy of the PSM+ICR report to your management. Every time that an additional human activity is introduced (action or thought) there is opportunity for error. KISS and FSOPs.