Autopilot .... cat's out of the bag now

[PilotKris]

Trim annunciate is, in my opinion, a must have if you've got electric trim (and most LSA and experimental planes do).

Here's the scenario:

While IMC on a coupled ILS approach, there is a trim-runaway and the trim stops at full up. The pilot is unaware as the AP is flying the plane. He clears the cloud base just before DH and disengages the AP to land the airplane and WHOOOOA! The plane pitches up and climbs back into the clouds, stalls, drops a wing and exits the base of the clouds 30 degrees nose down/45 degree bank at less than 200' AGL. Everybody dies...

You can see how full down trim can be just a deadly.

A friend of mine was on the team that investigated a run-away trim accident on an ILS approach. The panicked pilot pushed the PPT instead of the AP/Trim disconnect. My friend heard the ATC recording of the last 45 seconds of 4 people's life’s. It was the most horrific thing he'd ever experienced...

Autopilot + Electric Trim without annunciation is an accident waiting to happen...IMHO

PilotKris

 

[Bill_Denton]

In the "Duhh..." category...

Why not just a quick glance at the trim position indicator before disengaging the autopilot?

Trim annunciate is, in my opinion, a must have if you've got electric trim (and most LSA and experimental planes do).

Here's the scenario:

While IMC on a coupled ILS approach, there is a trim-runaway and the trim stops at full up. The pilot is unaware as the AP is flying the plane. He clears the cloud base just before DH and disengages the AP to land the airplane and WHOOOOA!  The plane pitches up and climbs back into the clouds, stalls, drops a wing and exits the base of the clouds 30 degrees nose down/45 degree bank at less than 200' AGL. Everybody dies...

You can see how full down trim can be just a deadly.

A friend of mine was on the team that investigated a run-away trim accident on an ILS approach. The panicked pilot pushed the PPT instead of the AP/Trim disconnect. My friend heard the ATC recording of the last 45 seconds of 4 people's life’s. It was the most horrific thing he'd ever experienced...

Autopilot + Electric Trim without annunciation is an accident waiting to happen...IMHO

PilotKris

 

[khorton]

A friend of mine was on the team that investigated a run-away trim accident on an ILS approach. The panicked pilot pushed the PPT instead of the AP/Trim disconnect. My friend heard the ATC recording of the last 45 seconds of 4 people's life’s. It was the most horrific thing he'd ever experienced...

Autopilot + Electric Trim without annunciation is an accident waiting to happen...IMHO

Do you know what type of aircraft that was?  I wonder what indications that aircraft had of trim motion, autopilot holding trim force, etc.  Certainly whatever annunciations were present didn't prevent the accident.

 

[PilotKris]

In the "Duhh..." category...

Why not just a quick glance at the trim position indicator before disengaging the autopilot?

In the "Things that you didn't think of that have actually killed people" category....

1. The failure that caused the runway trim could have caused the failure of the trim indicator (such as tripping the breaker).

2. Trim indicators are notoriously hard to see as the only time you need to look at them is prior to take off.

3. Adding another thing to do/look at (and divert your attention from flying the aircraft) in the 24 seconds between DH and hitting the ground is a bad idea.

There is no portion of a flight more demanding than the final portion of an IFR approach as we transition from the gages to visual flight (or execute a missed approach). That's why we have a checklist to accomplish prior to the FAF. Nothing is going to move your fuel selector while you're on the approach but something could be moving your elevator trim.

This isn't a "nice to have" item, it's a safety of flight item.

 

[Thomas_Schaad]

I agree, an out of trim indicator is something I would definitly miss as well.

Just as an idea: Once I flew for a while the IAI1125 Astra. This aircraft has a beep going on all the time the trim is moved. Since I transitioned to another type I really miss this "beep" going on. It was definitly the best way to control trim movements and I therefore wired the trim of my Pulsar to a sonalert which would now indicate an uncontrolled trim movement as well.

Kind regards

Thomas

 

[PilotKris]

It was a Barron and I don't know the details of what equipment was installed but obviously the pilot screwed up.

That's why hammer into my students (that have planes with electric trim) the various ways that disconnect the electric trim (up to and including turning off the master). They need to be able to do it with eyes closed.

Some airplanes, like my Aztec, are impossible to fly if the trim runs away (a single pilot does not have to strength to overpower a full nose up or nose down trim). The Inner Marker is not the time/place to learn that.

A "trim in motion" or "trim annunciator" may or may not have prevented that particular accident (additional training certainly would have), but there is no reason we should be designing systems that increase pilot workload and increase the potential for more pilot errors...

PilotKris

 

[Ken_Kopp]

How many conditions exist that can put a pilot in jeopardy during transition from IMC to VMC at low altitude and is it possible or even reasonable to design, equip, fund and train for ALL of them?

Runaway trim would be really nasty so having an indication would be nice i agree..but so would the following:

An engine failure is bad too..should I install a back-up?

What about a complete electrical failure just prior to breaking out at DH? Need a standby alternator or B/U battery.

What if the aileron attach bolts fail due to turbulance...better install secondary attach points..

How about that gascolator? What happens if that spring loaded retaining pin slips? Lost all my gas on short final...better have an alternate fuel feed system and tank just in case...

Didn't expect my prop governor to fail at low RPM either...electric prop pitch controller for a back-up? Or switch to fixed pitch prop?

Not trying to be a smart A** seriously..my point is simple...there are thousands of POTENTIAL threats to continued safe flying. How you prioritize them is a matter of experience, training, and luck. How you mitigate the impact should they occur is also a matter of experience, training, luck and funding. It would be nice to have systems in place to mitigate all of the potential failure modes in an airplane but it is simply not possible to do so..funding, weight, technology are all barriers to this end.

What we can do is conduct a very thorough and honest Operational Risk Assesment (ORM we call it in Naval Aviation) using the following steps:

1. Identify the hazards

2. Assess these hazards and assign a risk category based upon probability of occurance and level of severity should it occur. What is the prob of a trim runaway (low in my opinion), what is the severity (high during coupled IMC approach maybe). So risk is Low-High= Medium

3. Make risk decisions: is Medium acceptable for today's flight? If yes, then mitigate each risk through change in procedure, additional training, change of mission, or cancel the flight all together. Possible mitigation could be to cross check trim indicator prior to FAF and again before DH and positively identify circuit breakers, switches prior to approach initiation - Cmon - your workload shouldn't be that high if the autopilot is doing its job!! What else can you do to minimize the severity of a trim runaway? What about the probability? Was it checked during last annual? What is the history of your particular trim system? Has it been giving your trouble lately? Be honest with yourself and you'll conduct at much better ORM.

4. Accept no uneccesary risk - if you really feel the runaway trim issue is that risky (probablity high, severity high) then don't buy the AP76 and fly approaches manaully as you do now. Or only fly coupled approaches in VMC (which defeats the intent of the autopilot really).

We've used ORM very successfully to reduce our collective mishap rates significantly over the past 10 years. It works well - far better than adding systems for all the possible shortcomings of our aircraft - there are many and always will be. Very very few mishaps, military or civilian occur due to mechanical failures. Most are pilot induced in some fashion (fuel, weather, CFIT).

Just my humble opionion

CDR Kenneth Kopp, US Navy
XO HSM-70

 

[khorton]

One thing to keep in mind is that if an autopilot has the ability to run the electric trim, this adds yet one more failure mode that can cause a trim runaway.  I agree that from a pure pilot workload perspective it is very nice to have an autopilot that adjusts the pitch trim if the aircraft becomes out of trim.  But, this feature would need to be very carefully designed so that single failures, or single failure + dormant failures, cannot cause a trim runaway.

Engineering and test time costs money.  How much extra are buyers prepared to pay to have an autopilot that can control pitch trim to keep the aircraft in trim?

 

[Louiscfi]

geez..didnt know I would spark a whole big debate. For my plane (zenith 601) the trim indicator is more of a convenience than a necessity since even at full up/down trim i can still over-power it. (and trust me....im no strong man)

I suppose if you plan on installing this autopilot in your baron/aztec/mooney/etc you may want to have a more complex barrier between a runaway trim condition and your autopilot...but, for my plane, and I assume most experimental planes that this is intended for...it just doesn't really matter.

"just my 2 cents...plz dont sue" .....something about a turnip and blood, I dont know.

 

[PhantomPholly]

Maybe someone said it and I just missed it, but...

On the TruTrack, at least, and I suspect on several other popular autopilots which use stepper servos, you can set the max AP servo force rather low and if the airplane is the teensiest bit out of trim you will know it immediately - there will be a rather loud "clicking" from the servo, and the nose of the airplane will jerk repetitively each time the servo "slips a notch."

IMHO, that behavior IS an out of trim annunciation - and much more noticeable than any annunciator light. No extra feature is needed.

 

[PilotKris]

Hey Kenneth,

If you actually saw how the automatic trim sensor worked, I'd think that you'd rate the failure probability much higher.

For automatic trim autopilots, operational error is even more likely as all that needs to happen to is for pressure to be applied to the elevator (perhaps from a passenger in the other seat) to cause a serious out of trim condition. Many a pilot has caused an out of trim condition by trying to "help" the autopilot (again a training issue).

With mechanical trim, again it takes no mechanical failure at all to cause a serious out of trim situation. If the AP was engaged during cruise (or worse decent) and then transitioned to approach, a large out of trim would develop. Without a trim-prompt indicator, there would be no way to correctly trim the elevator without disconnecting the autopilot (high-probability).

Another very-likely scenario is someone for electric trim is for someone (like a passenger) to inadvertently presses the trim button. On many aircraft it’s easy to mistake the PPT for a trim button. The out of trim condition again would be masked by the AP until disconnected (medium-probability).

As others on this forum have pointed out, if the autopilot had only an altitude hold function, trim-prompt would not be necessary (low-severity). But a true multi-mode autopilot (VS, Alt pre-select, Airspeed select, GS coupling) absolutely needs a trim-prompt (high-severity).

Once you have a trim-prompt, it's very easy to add automatic trim function (a relay or two and a few diodes) if the plane already has electric trim (as most LSA and Experimentals have).

This isn't just a nice-to-have feature (although it would be really nice to have) but a safety enhancement. Since we're talking to a company who is dynamic in its product development, I suggest that we encourage (rather discourage) safety features to be added. Most of us will not need them but if it saves some poor sub-standard pilot (and his passengers) from death, wouldn't it be worth it?

This isn't some "pie-in-the-sky" "happen once in a blue moon" problem. Real People are REAL DEAD. Shouldn't we learn from their mistakes?

PilotKris

 

[dodsond]

Setting the clutch breakout low will prevent large out of trim disconnects, but then the autopilot won't be able to fly several useful profiles. It might not even be able to hold altitude in common thermal producing weather patterns. If I must contantly trim, the then I might as well hand fly the aircraft.

Settng the clutch a bit higher (there is a limit based on safety) allows the autopilot to maintian the commanded profile during transitions. An annunicator will inform the pilot that the aircraft is out of trim. This should be checked before an intentional disconnect, and periodically during normal operations when workload permits. It should be checked more oftne during intentional state changes affecting pitch (such as begining a climb, just after level off or level flight speed changes).

The discussion on electric trim is a bit off point. Runaway trim is a valid safety issue, but only indirectly related to autopilot function. If runaway trim alarms and procedures are proper, then it doesn't matter if a shorted swith, meddling passenger or malfunctioning autopilot is causing it.

The annunciator is necessary regardless of manual or automatic electric trim.

-Doug Dodson
Glasair II-S FT
Flight Test Engineer, CFI-A/S&ME,I,G

 

[PhantomPholly]

Setting the clutch breakout low will prevent large out of trim disconnects, but then the autopilot won't be able to fly several useful profiles. It might not even be able to hold altitude in common thermal producing weather patterns. If I must contantly trim, the then I might as well hand fly the aircraft.

Settng the clutch a bit higher (there is a limit based on safety) allows the autopilot to maintian the commanded profile during transitions. An annunicator will inform the pilot that the aircraft is out of trim.

Yes, exactly - and the click, click, click of the autopilot as the nose goes pop, pop, pop is about the most noticeable "annunciator" I can imagine. Much better than a light or even a nondescript beeper - you can "feel" that the plane is out of trim and the autopilot is "working too hard." Too, you can't be fooled by a "failed annunciator" - because if that happens you are liable to notice that the AP is no longer flying the aircraft... This solution is almost as good, but not quite, as your passenger smacking you in the head and saying, "Fix it Dummy!"

This should be checked before an intentional disconnect, and periodically during normal operations when workload permits. It should be checked more oftne during intentional state changes affecting pitch (such as begining a climb, just after level off or level flight speed changes).

Agreed. In fact, the easiest way to check this "annunciator" would be to intentionally run / move the trim out of whack while the autopilot is engaged. This exercise, followed by intentionally disengaging the AP once max out-of-trim is reached, will have the additional benefit of providing the pilot with the opportunity to experience what the "worst out of trim" condition might be expected when they disconnect the autopilot intentionally and discover that the aircraft is out of trim.

The discussion on electric trim is a bit off point. Runaway trim is a valid safety issue, but only indirectly related to autopilot function. If runaway trim alarms and procedures are proper, then it doesn't matter if a shorted swith, meddling passenger or malfunctioning autopilot is causing it.

The annunciator is necessary regardless of manual or automatic electric trim.

-Doug Dodson
Glasair II-S FT
Flight Test Engineer, CFI-A/S&ME,I,G

All good points. To conclude, then, the two best "out of trim" indicators you could ask for are a) your hands, if you are hand-flying, and b) the bobbing / jerking nose of the aircraft that occurs when the autopilot is engaged and overloaded.

Overall, Doug, you make a terrific argument for replacing all existing "trim annunciators" with an autopilot with configurable torque limits!

Bill Reister
Not a high-tech test engineer - Just a dumb-ole ex fighter jock.

 

[dodsond]

Bill,

Please do not count my position as favoring low autopilot control authority (torque limits) to replace proper status annunciation.

I would infer that having the aircraft jerk in pitch and being OK with it means you fly solo most of the time.   I would prefer the autopilot fly the profile I assign it while I monitor its status.  Auto trim with the proper failure annunciation is nice to have and safe.  Without electric trim, then a simple up/down light is common, allows the autopilot to function properly across the envelope, and keeps pilot workload reasonable.  The clutch limits should be enountered only occasionally in normal operation.  Their purpose should be to allow the pilot to overpower the autopilot if necessary.

-Doug

 

[PilotKris]

Now that we're all pretty much in agreement (except Bill who apparently likes flying a bucking-bronco) that we really want a Trim-Prompt annunciator. Let's put this topic to bed.

To summarize:

1. Manual Non-electric Trim + altitude hold only AP = Trim-Prompt is just nice to have

2. Manual Electric Trim + altitude hold only AP + Trim-Prompt is important

3. Manual Non-electric Trim + multimode AP = Trim-Prompt is VERY important

4. Manual Electric Trim + multimode AP = Trim-Prompt is a MUST HAVE!

5. Automatic Trim + multimode AP = Trim-in-motion/annunciate is a MUST HAVE!

OK with all (except Bill)?

PilotKris

 

[dodsond]

To summarize:

1. Manual Non-electric Trim + altitude hold only AP = Trim-Prompt is just nice to have

2. Manual Electric Trim + altitude hold only AP + Trim-Prompt is important

3. Manual Non-electric Trim + multimode AP = Trim-Prompt is VERY important

4. Manual Electric Trim + multimode AP = Trim-Prompt is a MUST HAVE!

5. Automatic Trim + multimode AP = Trim-in-motion/annunciate is a MUST HAVE!

I don't see any difference in electric vs. non-electric trim. I would revise as follows:

1. Manual Trim + altitude hold only AP = Trim-Prompt is just nice to have

2. Manual Trim + multimode AP = Trim-Prompt is a MUST HAVE!

3. Automatic Trim + multimode AP = Trim-in-motion/annunciate is a MUST HAVE!

The basic prompt can be a silent up/down visual indicator on the EFIS or autopilot control panel, tone or voice would be nice to have. The trim in motion should have a tone or voice prompt.

-Doug

 

[PhantomPholly]

Now that we're all pretty much in agreement (except Bill who apparently likes flying a bucking-bronco) that we really want a Trim-Prompt annunciator. Let's put this topic to bed.

To summarize:

1. Manual Non-electric Trim + altitude hold only AP = Trim-Prompt is just nice to have

2. Manual Electric Trim + altitude hold only AP + Trim-Prompt is important

3. Manual Non-electric Trim + multimode AP = Trim-Prompt is VERY important

4. Manual Electric Trim + multimode AP = Trim-Prompt is a MUST HAVE!

5. Automatic Trim + multimode AP = Trim-in-motion/annunciate is a MUST HAVE!

OK with all (except Bill)?

PilotKris

Kris,

I'm not sure where you get that I'm disagreeing with your statements that trim annunciation is a "must have" for flying IFR - I'm not. What I'm disagreeing with is that only a separate, dedicated trim annunciation system can fill that role. Not only is that incorrect, I believe that it is more dangerous than using the autopilot. While my previous reply was definitely a bit tongue-in-cheek, it was also serious - you guys are inventing a problem that needn't exist in our typical (collectively) environment.

The safety problem we are trying to solve is to ensure that, should an out of trim situation occur while the autopilot is engaged, it does not get so bad that either intentional or unexpected autopilot disengagement causes an immediate unsafe condition.

As was pointed out by another poster, trim annunciators can also fail. If you ALREADY have an autopilot (such as a TruTrack) that allows configuring torque limits AND DON'T set them to something reasonable, then you are just as foolish as the person who flies a GPSS approach to minimums on the AP without ANY sort of trim annunciator. Whether you personally are comfortable with that as your PRIMARY annunciator (passenger comfort can be a factor on larger planes, but really shouldn't be an issue in a 2-seater) is a personal pilot decision - certainly it would be unacceptable in a 747, but for most of us flying 2 seat experimentals it would be an equally foolish expense to add a separate system to pre-warn you that you forgot to adjust your trim (as you should automatically do when configuring for an approach). By the way - if you guys are really focusing this conversation on 747s, then I must be on the wrong forum....

In my aircraft I can position my trim in all phases of flight too within a nit of perfect without any feel-feedback - just because I know my airplane. With electric trim you must have an indicator to provide a visual reference - but you should still be able to perform the same feat under most circumstances, and if you are flying the plane often enough to consider yourself IFR capable it should be "automatic." If you are regularly forgetting to do this as you change phases of flight / configuration, you probably NEED someone to smack you "up side the head."

In conclusion, I don't consider it an OPINION that a SEPARATE trim annunciator is anything more than a "nice to have" for those who already have a torque-limited autopilot. What's more - if you are relying on an instrument that can fail (a separate annunciator system) JUST so that you can set your AP torque limits SO HIGH that actually reaching them constitutes an emergency - I would submit that you are programming for disaster and should reconsider your dangerous theory.

Mileage may vary...

 

[PilotKris]

I don't see any difference in electric vs. non-electric trim.  I would revise as follows:

1. Manual Trim + altitude hold only AP = Trim-Prompt is just nice to have

2.  Manual Trim + multimode AP = Trim-Prompt is a MUST HAVE!

3. Automatic Trim + multimode AP = Trim-in-motion/annunciate is a MUST HAVE!

The basic prompt can be a silent up/down visual indicator on the EFIS or autopilot control panel, tone or voice would be nice to have.  The trim in motion should have a tone or voice prompt.

-Doug

Hey Doug,

The only real difference is that with manual electric trim, someone (passenger, or pilot) or something (broken switch) could cause an out-of-trim condition that would be masked by the autopilot thereby increasing the importantance of having a trim-prompt.

It's not likely that a passenger would move a the trim wheel by hand but it's very likely that they might mistake the trim button for the PPT.

PilotKris

 

[PilotKris]

Bill Bill Bill...How did I know you'd be the only one to disagree...

No one is suggesting setting the torque limits to unsafe levels
(I have no idea where you got that impression)

The limit needs to be set to the level recommended by the AP manufacturer. I don't know about TruTrack, but I've never seen a manufacture recommend using the torque-limiting mechanism as a trim indicator. I've only heard of it referred to as a safety mechanism so that the pilot can override the autopilot and to prevent damage to the aircraft if the control hits its stop.

You're the only one who is comfortable using low torque limits as an out-of trim indicator. Let's not forget that the torque-limiting mechanism can also fail...

We're giving suggestions to Dynon on ways to build a better product. While you've already got a TruTrack autopilot, the rest of us would like to see Dynon make a better unit than TruTrack.

PilotKris

 

[PhantomPholly]

Bill Bill Bill...How did I know you'd be the only one to disagree...

I think it is closer to "violent agreement"...

No one is suggesting setting the torque limits to unsafe levels
(I have no idea where you got that impression)

I got that impression from the implication that torque limits on the autopilot are NOT in and of themselves an annunciator, and because a different response suggested that in some aircraft (don't recall which, but it was much heavier than our typical experimental) the writer felt that they "needed" very high torque limits to prevent autopilot disengagement in a normal flight profile. For the first, I think we agree that in the absence of any other indication torque limits certainly DO provide annunciation - we just disagree whether or not they fill the bill for your "must have" evaluation. For the second, again referring to the experimental world I would be deeply surprised if control forces vary enough from cruise to climb (all flaps up) to result in "dangerous out of trim stick forces." Thus - again in our community and for the types of aircraft likely to be equipped with a Dynon - torque limits should be perfectly adequate warning that runaway trim has occurred.

The limit needs to be set to the level recommended by the AP manufacturer.

Absolutely! I don't see this as contradictory to what I am talking about.

I don't know about TruTrack, but I've never seen a manufacture recommend using the torque-limiting mechanism as a trim indicator.

Perhaps not, but it is certainly a logical corollary. Consider this:

(from the TruTrack Installation Manual) The reason for setting LAT TORQUE to a setting less than its maximum (12) is to reduce the current draw of the servo and to
make it easier to override the autopilot should the need arise. Manual override is not normally required, ..., but it is best to have a setting of torque which can be comfortably overridden if necessary.

(Bold mine) It is a logical corollary that any trim condition which overrides "a setting of torque which can be comfortably overridden if necessary" is necessarily approaching the level the pilot chose as a "comfortable to safely override the autopilot" - and thus any force GREATER than that is "uncomfortable" and therefore is approaching a dangerous situation. I'd call that a tremendous "trim annunciation"...

Speaking of TruTrack, they built trim annunciation into their earlier autopilots. Incidentally, I had one of their earlier ones on my previous plane - it indicated the opposite of the necessary correction and there was no way to reverse it for my installation - annoying! Also annoying was that, at neutral trim, it indicated that a trim change was needed - broken. Dynon has a great opportunity here to do better...

I've only heard of it referred to as a safety mechanism so that the pilot can override the autopilot and to prevent damage to the aircraft if the control hits its stop.

You're the only one who is comfortable using low torque limits as an out-of trim indicator.

I'd hardly call this single board a representative sample of who is and is not comfortable with this - and some writers expressed that they didn't feel it was necessary at all. My whole reason for posting originally was that I didn't want some newbies to think that an unnecessary piece of equipment was actually necessary. We're all on budgets, and it is important to understand what is REALLY necessary from what is nice to have.

Let's not forget that the torque-limiting mechanism can also fail...

Can it? I wonder - it certainly might be possible with older style mechanism, but with modern stepper motors? If you size your servos right, even MAX torque should be possible to overcome by the pilot (although certainly more difficult and therefore lending a greater risk potential). Thus, the two possible failure modes would be "MAX Torque" (still providing some level of safety if you did not oversize your servos) and "NO Torque" (autopilot failure - back to hand flying and the Mk I Tactile Trim Sensors). Oh, I like that - the "TTS Annunciators!"

We're giving suggestions to Dynon on ways to build a better product. While you've already got a TruTrack autopilot, the rest of us would like to see Dynon make a better unit than TruTrack.

PilotKris

Ok - in that vein I will agree that it is a terrific idea for Dynon to include some sort of out-of-trim annunciation. On the screen; verbally through the HS34. You could use it to insure that trim is PERFECT before you disengage the AP. If their servos are already smart then they are already generating DSAB-data concerning "Torque Required" which can be averaged over a few seconds to determine if there is an out-of-trim situation.

BUT PLEASE NOTE THAT THAT'S NOT AN INDEPENDENT ANNUNCIATOR! It is just presenting the same data in a different way to the user. Or, giving a finer granularity of information if you will. But, NOT separate.

In conclusion, an average-force indicator showing precise out-of-trim error would be nice to have, but NOT NECESSARY, in addition to a properly sized torque-limited autopilot in the experimental world. Airliners is a whole different story (I've flown the 737 sim) - an autopilot can compensate for an out of trim condition the pilot could not control. But in our world (the world of the recreational experimental flyer) it just isn't possible to create dangerously out of trim stick forces unnoticed unless your autopilot is incorrectly sized / installed.