FLCH Trap?

[BeechPapa 32]

In the year or so that I’ve owned the NGX, with hundreds of hours on it, I never crashed it. I might have had some scary go-arounds, and some spine-compressing landings, but the safety systems always prevented major disasters. Yesterday, I’m ashamed to admit, I crashed the 777. The incident unfolded in almost the exact same way as the Asiana 214 disaster –I was on short final for a visual approach, AT switches were both armed, AT on, yet the plane continued to lose airspeed on approach, to the point that it stalled right before the runway. In a panic, I finally noticed airspeed dropping, but it was too late to get the engines spooled up again and the plane hit the water at speeds freakishly similar to the Asiana flight.

 

According to my research, some pilots believe the Asiana disaster might have been caused by the behavior of the FLCH mode, what some pilots apparently refer to as the “flight level change trap”. 

http://www.aviationweek.com/Article.aspx?id=/article-xml/AW_07_22_2013_p25-597816.xml My incident is described almost word for word in this article. I was in FLCH mode when I crashed, and had selected 0 feet in the altitude window. The plane began to idle at 500ft or so, and with gear extended and flaps coming down, it diminished very quickly. Nothing I could do would make the AT ‘wake up’. Disengaging AP (which I thought would disengage FLCH) didn’t do anything to wake up the throttles. Despite the fact that AT was on, the plane continue to idle at 118kts at 400 feet, dropping quickly. The flight envelope protection systems never kicked on either. It’s as though FLCH completely disables flight envelope protection. I am not one to believe strongly in automated systems (see avatar) and don’t particularly think a pilot should rely on them to save the plane, yet I am a bit frustrated that flight envelope protection didn’t activate. To be honest, the 777’s systems can seem spooky sometimes. I am still having issues trying to understand how all these fly-by-wire systems interact and where the line between manual and automatic is truly drawn.

 

I read that airlines are working on the FLCH trap in training now, and I really wish I could get some of that training too. Has anyone ran across this FLCH issue or have an idea of what Boeing’s goal was with the FLCH mode from a theoretical standpoint? And what could I have done differently? Should FLCH never be used during approaches?

[scandinavian13 5,192]

 

 

Has anyone ran across this FLCH issue or have an idea of what Boeing’s goal was with the FLCH mode from a theoretical standpoint?

 

Pitch for speed based on a set thrust point - usually rated thrust for climb, idle for descent.

 

 

 

And what could I have done differently?

 

Not used FL CH.  Taken control of the throttles physically.  Remember AP is NOT AT.  They interact, but they don't control each other.

 

 

 

Should FLCH never be used during approaches?

 

FL CH should never be used if its concept is not understood for the reasons seen.

[MarkW 1,502]

Interesting, I have actually never used FLCH below about 2000ft, generally using it below 10,000ft and then at 3000ft, once I have intercepted I either use V/S to follow a VOR approach, or hand fly.  On and ILS I would either hand fly it or use Land1 or 3 depending on the situation.  For sure anything below 1000ft would be manual or Land3 if visability was terrible.  I am actually surprised that FLCH would be used below 2000ft or even below 1000ft.

[FLEX1978 373]

It has been a while since I looked into the Asiana crash. Has it been confirmed what altitude was set in The MCP? You should never have 0 set in the MCP and descend in FLCH at low altitude, on final approach you should have the missed approach altitude set.

 

Thrust will just go to HOLD and sit at IDLE until someone on the flight deck wakes up.

 

EDIT - just read your your link, why on earth would some airlines want to set 0 in the MCP for a NPA??

[tcas_climb 45]

I think some Airlines set 0 in the altitude mcp is to prevent unwanted altitude capture so they can continue the approach with vnav guidance to the runway. This may be one of the reasons but personally I understand what could happen when you do this and still it's not something I practice much. I usually set it to the DA.

 

Edit: it would appear that in Asiana case, they were pretty high and fast initially and possibly set 0 to get down, and in the rush of things forgot about it.

[FLEX1978 373]

Thanks Bryan

[BeechPapa 32]

It has been a while since I looked into the Asiana crash. Has it been confirmed what altitude was set in The MCP? You should never have 0 set in the MCP and descend in FLCH at low altitude, on final approach you should have the missed approach altitude set.

 

Thrust will just go to HOLD and sit at IDLE until someone on the flight deck wakes up.

 

EDIT - just read your your link, why on earth would some airlines want to set 0 in the MCP for a NPA??

I can't speak for anyone other than myself, but I set 0 ft in the MCP because I was rushed for approach, busy preparing for landing configuration, and needed to put the plane into some sort of descent mode in the meantime. In general terms my philosophy during this approach was "wam, bam, thank you maam" and it really really bit me this time. I avoided VS because I was forced onto a steep descent for final (FSX ATC) and didn't want to overspeed. So, for some odd reason, I thought FLCH was the safer bet. I am starting to realize the stupid move was setting the MCP to 0 because it never gave the plane a chance to hit an altitude hold point. But I am still surprised by the stubbornness of the AT to stay at a given thrust point despite the fact the airplane has dropped below VREF and the speed selected in the IAS/MACH window. I am fairly sure after this debacle I will remember to never go lower than missed approach in the altitude window.

 

 

 

I have actually never used FLCH below about 2000ft, generally using it below 10,000ft and then at 3000ft

 

I'd stick to that rule if I were you. I might be sticking to it as well.  

[DAL1850 51]

It seems counter-intuitive to use FLCH to fly the last leg of an approach given that FLCH is an idle-thrust pitch mode.

If I'm not in a LAND3 config, I'll usually be in SPD HOLD (commanding Vref+5 or half-gust speed if winds are gusting) and manually align my throttles to sync with the currently commanded thrust. At DH, I'll disconnect and land her manually; throttles will be right where they need to be and responsive. MCP is typically set to missed approach altitude after ILS G/S capture. Alternatively, I'll set to my DA if flying an RNAV approach under VNAV guidance (even with VNAV I always use speed intervention at low altitude). 0 on the MCP (or anything at or below the field elevation) seems plain nutty to me and it surprises me this is an "allowable practice" in the real world. I can see the hassle with inadvertent altitude capture, but I'd rather fly a missed approach than a dead one.

[xkoote 57]

Hi,

 

The FL CH mode of the 777 is not a trap. I do not fly the 777 but the MD-80 and in there it's exactly the same. If you fly in IAS HOLD in descent, the throttle will idle. If you then disengage the autopilot and follow the flight directors you will hold 137 kts (in this scenario) until 0 feet (Not too smart) and I am SURE an incorrect SOP. 

I don't have problems with setting 0 in VNAV. But then you HAVE to fly an RNAV approach as your SOP stipulates. What they did was disengage the A/P and took over manually to arrest the descent rate. The flight directors were probably giving them nose down as they slowed below the 137 in the speed window.

By setting 0 in the window and arming FL CH they already painted themselves into a corner.

 

Now if all factors mentioned on different sites and repeated above is true, then the stall speed would have been 137 (Vref+5) - 5(Vref) / 1.3(Vref is 1.3 Vs more or less) = Close to 100 knots. Now here is where in my opinion it all came undone.

In the 777 manuals in the Flight Controls System description page it states that even if the A/T is only armed - at near stick shaker activation - the A/T will engage and accelerate the aircraft to Vmin.man....BUT

 

"Note: When the pitch mode is FLCH or TOGA, or the airplane is below 400 feet above the airport on takeoff, or below 100 feet radio altitude on approach, the autothrottle will not automatically engage."

 

They were in FL CH and could not get the A/T speed protection. I read that they tried commanding various MCP functions. If by then they were already below 100 feet (judging by the video, they were below 100 feet for longer than usual) they would also have been unable to get the speed protection system as by then the A/T were set up for the retard command.

It was an unfortunate series of events that led to the accident if all speculations turn out to be the case.

But setting 0 in the altitude window with FL CH mode armed, below 2000 feet and the autopilot disengaged is a BIG no no. Too bad they found all corners where the speed protection system was disabled. It is also too bad that they didn't seem to have taken advantage of a Boeing philosophy. When all else fails, shove those two sticks to their stops and hold them there !!!

 

 

Cheers,

 

Xander

[FLEX1978 373]

Very nice post Xander, but unless I am mistaken you are confusing IAS HOLD on the MD80 with THRUST HOLD on the 777 and some other Boeing types.

 

Does the MD80 not have an IDLE CLAMP like the MD11? This is what HOLD is on the 777, it will not Hold 137 knots or whatever speed you have in the MCP if you are hand flying with the FD's on and in FLCH ignoring FD commands, Thrust will just remain at IDLE and happily stall the aircraft in some situations. This is why it is called the FLCH trap.

 

Apologies if I have misunderstood you, it sounds like you are talking about IAS hold, something completely differant.

 

Kind regards

[bjratchf 80]

We ran the accident scenario at my yearly recurrent sim training in the 767. I really don't see how they crashed it- the situation starts to "not look right" and that instinctual pilot urge (that almost all, if not ALL) US-trained pilots have to take over manually. FLCH basically makes the pitch-power control a little backwards, the philosophy being "if I add power, the plane will lessen the rate of descent." However, in a large jet, power changes take a long time to cause a change in speed, and trajectory (which is why stabilized approach criteria is SO important), so I don't see how one can successfully use this method to maintain an constant 3 degree angle of descent with an autopilot on- ESPECIALLY since autopilot masks developing errors/threats and kind of "numbs" us as pilots.

[Fabo 213]

 

 

Apologies if I have misunderstood you, it sounds like you are talking about IAS hold, something completely differant.

 

IAS/MACH hold in MD-80 works the same way as FLCH in the 777. What you are probably thinking of is known as Speed Select in the dog.

[FLEX1978 373]

IAS/MACH hold in MD-80 works the same way as FLCH in the 777. What you are probably thinking of is known as Speed Select in the dog.

Thanks for the clarification really isn't any excuse for me not to own the Mad dog! time permitting I will definitely purchase and start studying that sim.

 

Regards

[WR269 367]

Thanks for the clarification really isn't any excuse for me not to own the Mad dog! time permitting I will definitely purchase and start studying that sim.

 

Regards

Good luck! Welcome to the burnt engine club, you are likely to roast a few!

 

Sent from my Mobile thing

 

 

[Jetstream96 20]

Why did you use FLCH for visual approach? FLCH won't do the job. No offence, but I don't see the reason to use FLCH after you intercepted the glideslope.