Author Topic: Now that the BUF thread has been locked...  (Read 47504 times)

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #25 on: February 19, 2009, 06:41:13 PM »
Exactly right. Its still very good information. Somthing else to toss into the mix you know?

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #26 on: February 19, 2009, 06:49:09 PM »
I can see how this would cause a stall thou. Im not familure with the -8 and its autopilot system. If the glideslope was deflected, and now the airplane thinks its "up there" and tries to go to it, and if the autothrottle (if equipped) is not engaged, the airplane pitches up, 31 deg, without power being added...stall.

But like you said, an accident is a series of smaller events. We just need to figure out what those are.

kea001

  • Guest
Re: Now that the BUF thread has been locked...
« Reply #27 on: February 19, 2009, 08:23:26 PM »
SInce I posted that, the following has been added to the article:


The National Transportation Safety Board told CNN the agency was "aware" of the Southwest Airlines alert, but would not comment further.

The issue is caused by a geographic feature at the airport, a valley, "something we can't do anything about," said FAA spokeswoman Laura Brown. She said the "altitude reading makes it look like you're a lot higher than you are, because there is a valley there."

The feature has been noted on FAA charts for years, she said.

"As far as we can tell, there is no way this had any role in the accident," Brown told CNN.

"It's not a navigation aid that would have applied to the approach."

The alert from Southwest Airlines advises pilots that the problem could cause the planes navigational system to interpret data "in such a way as to result in a nose-up pitch and loss of airspeed."

http://www.cnn.com/2009/US/02/19/buffalo.crash.ils/index.html

Quote
"It's not a navigation aid that would have applied to the approach."

I think what this means is the plane probably didn't even make it to the land formation (valley) where this might have been an issue.

This article explains it a bit further:

Southwest pilots union warns of Buffalo approach hazard
http://www.flightglobal.com/articles/2009/02/20/322868/southwest-pilots-union-warns-of-buffalo-approach-hazard.html
« Last Edit: February 20, 2009, 04:13:37 PM by kea001 »

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #28 on: February 21, 2009, 05:32:58 PM »
So they are thinking 3407 crashed outside the interfiered with are of the approach? Sound very similar to what was stated in the report.

"in such a way as to result in a nose-up pitch and loss of airspeed."

Colgan pitched to 31 deg nose up.


Offline iskyfly

  • Full Member
  • ***
  • Posts: 179
Re: Now that the BUF thread has been locked...
« Reply #29 on: February 22, 2009, 08:22:19 AM »
and if the autothrottle (if equipped)
It is not.

Here is a quote from a MD80 and 737 pilot;
http://www.airdisaster.info/forums/viewtopic.php?f=6&t=2054&start=150#p27072

Quote
Regarding the KBUF ILS 23 warning....this is not a new issue and has been known about. The problem comes when joining from the right (or west) side when the plane captures the G/S prior to the LOC. Colgan was joining from the left side and like many planes, the automation is such that the G/S cannot be captured prior to the LOC being captured. The 737, which SWA flies is one aircraft that I know can capture G/S prior to the LOC being captured.

While eye catching and eerily similar sounding, I think this is a red herring.

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #30 on: February 22, 2009, 09:41:57 AM »
I didn't think the -8 had autothrottle. Most turbo prop aircraft do not. So that makes sense to me then. The airplane was on autopilot, with "approach" selected. The glideslope deviated, the airplane tried to go to it. The pilots never added power as the airplane pitched up 31 degreees and lost airspeed, and stalled the wings. Before the pilots realized what was happening, it was too late.

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #31 on: February 22, 2009, 09:46:11 AM »
But it was the wrong side...

"The problem comes when joining from the right (or west) side when the plane captures the G/S prior to the LOC. Colgan was joining from the left side and like many planes,..."

This is bizzar. All the right symptoms, wrong side...very strange.

Offline iskyfly

  • Full Member
  • ***
  • Posts: 179
Re: Now that the BUF thread has been locked...
« Reply #32 on: February 22, 2009, 11:56:02 AM »
I didn't think the -8 had autothrottle. Most turbo prop aircraft do not. So that makes sense to me then. The airplane was on autopilot, with "approach" selected. The glideslope deviated, the airplane tried to go to it. The pilots never added power as the airplane pitched up 31 degreees and lost airspeed, and stalled the wings. Before the pilots realized what was happening, it was too late.
Well, you forgot this part;
Quote
Colgan was joining from the left side and like many planes, the automation is such that the G/S cannot be captured prior to the LOC being captured.


I don't think pursuing this "lead" will yield results.

It is time to wait for more facts.

Offline mhawke

  • Full Member
  • ***
  • Posts: 107
Re: Now that the BUF thread has been locked...
« Reply #33 on: February 22, 2009, 12:32:15 PM »
The glideslope deviated, the airplane tried to go to it. The pilots never added power as the airplane pitched up 31 degreees and lost airspeed, and stalled the wings. Before the pilots realized what was happening, it was too late.

Only problem with this theroy is that according to the NTSB, the autopilot was off and the 31 degree up was pilot commanded.  He overrode the stick pusher for some reason and pulled back when the stall warning came on.  That ws before the up angle.

Offline badger634

  • Newbie
  • *
  • Posts: 6
Re: Now that the BUF thread has been locked...
« Reply #34 on: February 22, 2009, 01:20:06 PM »
From what I understand, the captain had recently been flying Saab340's, which have the potential for tail stalls.  In a tail stall, you need to pull UP (among other things).

Dash 8's are not susceptible to a tail stall, but this pilot might have been relying on his Saab 340 days.  Instead, the stick shaker just indicated a run-of-the-mill wing stall, and his action of pulling up doomed the plane.

Does that seem reasonable?

Offline cessna157

  • Hero Member
  • *****
  • Posts: 708
    • facebook
Re: Now that the BUF thread has been locked...
« Reply #35 on: February 22, 2009, 01:44:38 PM »
You are correct in suspecting this is not the case.  When in approach mode, the autopilot follows the glideslope as a secondary case.  Pitch has greater control over the autopilot (it is more complicated than that, but let's just go with this as the baseline).  If the glideslope goes full scale deflection high, the autopilot will pitch the aircraft up to correct only to a certain point.  To use an extreme example:  If the glideslope remains at full up, the autopilot will not perform a loop trying to get to it.

It is part of training that in an extreme ice encounter, the autopilot may hide what is truely going on with the aircraft.  That is true to a certain extent.  Along with a strong elevator force, there will also be a lot of trim indicated.  So that will be the first hint that something is awry. 

Also, another level of protection is a caution of such a condition.  I'm not sure exactly what the Q400 has for an avionics/autopilot suite, but the CRJ has a Collins package.  Something that we'd see on the older (let's use the phrase "well broken in" if you catch the hint) planes is, upon rotation at takeoff, the aircraft immediately wants to roll.  This is considered normal (1 wing is twisted, normal wear/tear) and can be easily fixed with aileron trim.  If this rolling tendency is not trimmed out (let's say the aircraft wants to roll right, so constant left input is required) and the autopilot is engaged, it will fly the aircraft just fine.  But if the autopilot tries to make a left turn, it feels that it is using a larger than normal force so it will generate a Master Caution, the pilots hear "Ding", and in this case we'd get a yellow caution message that says "AP trim is LWD".  This means the autopilot has sensed an out of trim position, and it is holding the left wing down.

The point I am making is in today's aircraft, there are protections to out of trim conditions by the autopilot.  This is not to say that they are absolutely perfect (the CRJ autopilot likes to hold up elevator for some reason at slower airspeeds instead of trimming it out, so a technique pilots use to disengage the autopilot is not to use the disengage button, but to use the up trim button).

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #36 on: February 22, 2009, 02:36:26 PM »
OK...I was not sure if the autopilot was engaged or not. I hadn't heard for sure...I know Colgans Company Policy is for the pilot to fly the airplane in icing conditions...that dosn't mean they always do. So it sounds more like pilot error then pretty much anything else.


Offline iskyfly

  • Full Member
  • ***
  • Posts: 179
Re: Now that the BUF thread has been locked...
« Reply #37 on: February 22, 2009, 03:30:38 PM »
OK...I was not sure if the autopilot was engaged or not. I hadn't heard for sure...I know Colgans Company Policy is for the pilot to fly the airplane in icing conditions...that dosn't mean they always do. So it sounds more like pilot error then pretty much anything else.


Dont confuse an autopilot in alt / hdg hold mode with an autopilot capturing and flying the G/S and loc. They very well could of had the autopilot on, holding altitude / heading.
 

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #38 on: February 23, 2009, 07:35:30 PM »



[/quote]
Dont confuse an autopilot in alt / hdg hold mode with an autopilot capturing and flying the G/S and loc. They very well could of had the autopilot on, holding altitude / heading.
 
[/quote]

It's all technically "autopilot" if you will. HDG and ALT mode would be engaged with APPR selected. Once the airplane begins to capture the LOC and GLD the HDG and ALT holds will automatically turn off, and join the airplane on the glideslope/localizer.

But Colgan should have been all hand flying per company policy. They were in known icing conditions.

Offline cessna157

  • Hero Member
  • *****
  • Posts: 708
    • facebook
Re: Now that the BUF thread has been locked...
« Reply #39 on: February 23, 2009, 08:07:10 PM »

But Colgan should have been all hand flying per company policy. They were in known icing conditions.

Yeah yeah, and when you're driving you should never exceed the speed limit and always keep your hands at 10 and 2.   :-D

Offline KSYR-pjr

  • Hero Member
  • *****
  • Posts: 1722
Re: Now that the BUF thread has been locked...
« Reply #40 on: February 23, 2009, 08:38:07 PM »
But Colgan should have been all hand flying per company policy. They were in known icing conditions.

Do you know their policy first-hand or are you just restating what the media is reporting?   Are you certain that this hand-flying policy covers all levels of icing, including light and moderate?

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #41 on: February 23, 2009, 10:42:26 PM »
KSYR-prj: To answer your question:
Most airlines do require pilots to hand fly in icing. No I dont know their policy first hand. But i do know when i was a dispatcher with Zantop, that WAS a requirement. Hand fly the airplane in icing conditions. PERIOD. And further more, an friend of mine who is a Captain with Northwest on the A330 agrees with me. Hand flying in icing is a requirement.

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #42 on: February 23, 2009, 10:55:50 PM »
Let me re-phrase that-It is required to hand fly the airplane on approach only in icing conditions. On climb-out, you can engage the autopilot when the aircraft is stable.   

Offline KSYR-pjr

  • Hero Member
  • *****
  • Posts: 1722
Re: Now that the BUF thread has been locked...
« Reply #43 on: February 23, 2009, 11:45:27 PM »
KSYR-prj: To answer your question:
Most airlines do require pilots to hand fly in icing. No I dont know their policy first hand. But i do know when i was a dispatcher with Zantop, that WAS a requirement. Hand fly the airplane in icing conditions. PERIOD.

Even ILS CAT III C?  Hmmm.

edit:  I still question this "policy" with regards to the level of icing it references.  Cessna, did your company specify a level of icing where the AP must be disengaged on approach or was it any ice *period*?
« Last Edit: February 23, 2009, 11:57:59 PM by KSYR-pjr »

Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #44 on: February 24, 2009, 06:53:49 AM »
I will have to check on the CAT III requirement. Thats a good point.  :-)

Correct me if I'm wrong but I dont think KBUF has CAT III, does it? I think It's just simple ILS. 5/23 have ILS on each end, and 14/32 does not.
 


Offline iskyfly

  • Full Member
  • ***
  • Posts: 179
Re: Now that the BUF thread has been locked...
« Reply #45 on: February 24, 2009, 07:43:30 AM »
Regarding icing and autopilot;
http://www.airdisaster.info/forums/viewtopic.php?f=6&t=2054&st=0&sk=t&sd=a&start=50#p26750

Quote
Autopilot use in icing is not addressed for us in any of our manuals. I believe it was a recommendation that came from the Roselawn accident but was intended for smaller aicraft more susceptible to tail icing/stall. To be honest, tail stall is not even stressed on the big jets and I don't believe I've ever had any training on it. Boeing is apparently unconcerned enough about it that none of the Boeing products that you are familiar with (727/737/757/777) even have tail deicing. They are only equipped with wing anti-ice systems. My MD80 does have tail anti-ice as part of the airframe anti-ice system. I'm not sure if Airbus heats the tail or not. The only thing in our SOP that even touches on the possibility of tail icing/stall is a recommendation that when airframe anti-ice is required down to touchdown, you should select a tail cycle approximately 1 minute prior to selelction of land flaps (the anti ice cycle normally runs 15 mins on the wings followed by 2.5 mins on the tail. You have the ability to manually select a tail cycle and that is the recommendation prior to flap movement....this is only in our MD80 manuals as none of the other planes even have tail heat).

It sounds like Colgan's manuals 'recommended' hand flying in icing conditions. I can understand this as the autopilot will mask any strange control deflections or trim changes that you might otherwise feel when hand flying (the A/P will mask those changes and hold the desired attitude right up until it can't take it anymore and then will disconnect allowing those trim/control forces to do nasty things to the plane). However, it is still just a recommendation and the Captain could have felt that in IMC condtitions at night in adverse weather with a fairly new F/O (etc.) that it was better to keep the A/P on. No doubt that decision will be scrutinized. Their SOP requires handflying with severe icing and nobody is suggesting that severe icing conditions were present. The NTSB briefs keep mentioning 'significant icing', but inflight icing is rare enough that whenever you see a good amount cover the windshild bolts or the windscreen, pilots will always comment and say "wow, look at the ice building up'. That will go on the CVR, but taken out of context, it doesn't mean severe icing. You're not allowed to remain in severe icing....you are supposed to change altitudes/exit severe icing. You certainly wouldn't continue an approach if you felt you were in severe icing.

Offline cessna157

  • Hero Member
  • *****
  • Posts: 708
    • facebook
Re: Now that the BUF thread has been locked...
« Reply #46 on: February 24, 2009, 08:17:41 AM »
I still question this "policy" with regards to the level of icing it references.  Cessna, did your company specify a level of icing where the AP must be disengaged on approach or was it any ice *period*?

Well, after searching 1500+ pages of "ice" (it's surprising how many times offICEr and servICE show up!), I may have your answer. 

In the "Limitations" section of our flight manual, which describes all of the operating parameters (and unfortunately its the chapter that we have to have memorized for obvious reasons), I found absolutely nothing pertaining to the operation of the aircraft/autopilot in icing conditions.  Only saw the obvious stuff dealing with when the anti-icing systems must be on.  So, with that being said, it initially doesn't look like a hard ticketed item, to use the pun.

When looking through the Operating Policies, here is what I have found.  It is worded in such a way to provide the pilot with the choice (I learned in an Aviation Law class that many FARs are also worded this way, to provide loopholes, options, etc). 
For Cruise, it says "If there is a significant performance loss in icing conditions, CONSIDER disengaging the autopilot.  Leave icing conditions as soon as possible."  (emphasis added by me)
For Approach/Landing, autopilot disconnection is never mentioned, but it does say to consider adding up to 10 knots if visible ice is noticed.

Also in the recommendations it says to consider leaving the flaps up as long as possible to avoid additional aircraft icing, and if flaps are deployed in severe icing conditions for an extended time, airframe buffeting may start and is considered normal.  If retracting the flaps reduces the buffet, a landing may be made at the discretion of the pilot, applying speed, performance, and runway penalties.


When it says to "consider" disconnecting the autopilot, that goes along with all of the icing training I have received throughout my career, not just with the CRJ.  So, there are in fact no real limitations on how much ice we can fly through or when to use the autopilot.  It merely gives the pilots the choice of what they want to do.  That being said, the CRJ has excellent ice shedding abilities, especially the -700 and -900.  There have been a few accidents with the 50 seat series in icing due to the nature of its critical hardwing, but these have only occurred during takeoff at time of rotation, nothing airborne.
« Last Edit: February 24, 2009, 08:21:01 AM by cessna157 »

Offline KSYR-pjr

  • Hero Member
  • *****
  • Posts: 1722
Re: Now that the BUF thread has been locked...
« Reply #47 on: February 24, 2009, 09:56:49 AM »
Cessna, thank you for taking the time to look into the operations policy under which you flew (hint:  search for " ice " should limit the results to just the relevant sentences.  :) ).  Iskyfly - thank you also for reposting that quote.

The reason I am questioning this black-or-white finality of the icing policy is exactly what the bolded point in iskyfly's repost stated.   The risks of light-to-moderate icing to an aircraft certified for flight into known ice are indeed much lower than that of other hazards related to night, IMC flight.   Therefore I do not believe there was an operational prohibition of AP use in known light to moderate icing.  

Keep in mind that all other aircraft that night, including a similar model some 23 minutes behind the accident aircraft, reported only to a level of moderate icing on approach into Buffalo.  More importantly, the NTSB has not officially determined the level of icing that night, despite the publicly released comments from the accident flight deck indicating "significant icing."  As was stated elsewhere, caution should be exercised in taking this comment out of context since "significant icing" is not an official, defined level of ice.  For example, the pilots could have been commenting on the ice build-up on the windshield wiper, a part nonessential to the aerodynamics of that aircraft.

Placing the FAA-worded but controversial definition of known ice aside for a moment, consider that just about every cloudy day in the Northeast US from October to March (and there are a lot - too many, actually, and we are all sun-deprived) has a light-to-moderate icing potential.  Had there been such a prohibition of AP use in light-to-moderate ice, this would suggest that every air transport flight over the Northeast skies would be hand-flying approaches, most notably at night, the majority of the time.  Are we really to believe that every air transport aircraft is hand-flying approaches the majority of the time during the six months that make up icing season in the Northeast US?   Absolutely not.


Offline joeyb747

  • Hero Member
  • *****
  • Posts: 1745
  • Nothing Like A 747!
Re: Now that the BUF thread has been locked...
« Reply #48 on: February 24, 2009, 06:24:27 PM »
Ok-I have some new information. After consulting a couple of different people, I have this information for you all.

First of all, I was misinformed, and shared incorrect info, and for that i apologize.

The correct info is this: There is no such "requirement" on not using autopilot on approach in icing conditions on any type of aircraft.

However, there are "recommendations". All are type-specific, and vary by model. These "recommendations" are aimed at turboprop aircraft that have deicing boots. Aircraft that use heated leading edge do not have any such "recommendations".

Mesaba has a recommendation on its Saab 340 fleet that the pilot hand-fly the airplane on approach in moderate and higher icing conditions. This is per a NWA Captain. I am not familiar with Mesabas' policies/recommendations or whatever you would like to call them.

American Eagle has a similar recommendation on its ATR fleet, its Jetstream 31 fleet, and its Saab 340 fleet.

The -8, or Q400, has leading edge boots. I'm guessing the recommendation follows those same lines.

I've been told these recommendations came about after the American Eagle ATR crash in Roselawn, IN.
 
But of course, these are just recommendations. It's pilots discretion when to use autopilot or not.

I hope this will help clear up some things. And again, I apologize for the wrong information.     :-D

Offline napper505

  • Newbie
  • *
  • Posts: 9
  • 75 th anniversary
Re: Now that the BUF thread has been locked...
« Reply #49 on: February 25, 2009, 09:43:30 PM »
Hey

Was listening to Kbuf tonight

(mesaba?)  was shown as DhD 400 Colgan was redirected to London, contact Toronto Center etc..   between 9:00 pm and 9:30 pm local time

Current weather mixed snow and now mixed freezing rain 30 miles West of YYZ.

Quote from pilot "where do you want us to go?"   :?

There is a front moving thru .  Now listening to both YYZ and kbuf whenever I can


Napper505



« Last Edit: February 25, 2009, 09:55:04 PM by napper505 »