Harrowing screams form Learjet pilot

[3WE]

Also, why in the hell isn’t the automation giving better stall warnings?

After my 5,473’d Lear Jet landing, I might benefit from something telling me I’m getting a bit slow / high on AOA…

And to my knowledge most stall warning SYSTEMS have an AOA monitor.

I know when I did lower speed 172 stuff, there were little chirps reminding me it’s not the best time for pulling up blindly.

[Gabriel]

So fly 10 knots faster till you’re on final and be done with it!…

…And simulate some 270 degree circling approaches with winds and scattered scud.

Basically yes and yes, But why 10 knots? Why not 5? Or 25? Does 10 knots apply equally well for a Piper Cub and a Gulfstream? Or is the rule-of-thumb type-specific?

I don’t know, and I don’t care.

Run some math and give the bizjets dudes a figure. I expect it would take you 2 min to cipher something that would work for a 45-degree bank. (I could probably do it in 10 min).

I do know that lift and bank angles and speed closely follow some fairly basic mathomelogical relationships and lift ~ to the square of speed, so you get some extra bang for each knot.

It ain’t rocket surgery and it’s not_going to be something you can’t burn off when established on final for 98% of your operations.

Between Vref + 15% and Vref + 20%

[3WE]

Basically yes and yes, But why 10 knots? Why not 5? Or 25? Does 10 knots apply equally well for a Piper Cub and a Gulfstream? Or is the rule-of-thumb type-specific?

I don’t know, and I don’t care.

Run some math and give the bizjets dudes a figure. I expect it would take you 2 min to cipher it (It might take me 20 minutes).

Between Vref + 15% and Vref + 20%

So 20 knots…

Just talking out my ass- that seems a bit much with Vref already having its own safety buffer.

[Gabriel]

So 20 knots…

Just talking out my ass- that seems a bit much with Vref already having its own safety buffer.

Yeah, but the margin of Vref (1.23 Vs) is established over the stall, not over the stall warning. You don't want the stall warning to go off during the approach since that would (or should) trigger an approach-messing stall recovery in the best case, or a Colgan / Air France / Pinnacle reaction in the worse.

The number above (Vref+15/20%) has a 10% margin over the stall warning speed when you are in a 35-deg-bank coordinated and sustained turn and then pull (or are hit by a vertical gust that creates) an additional 0.5G.

[flyboy2548m]

Couple words about circling in general. In the US, it's a fairly rare maneuver. Perhaps because it's not required very often, most runways with any significant amount of traffic in this country have at least one straight-in approach of SOME kind. Hence, I would imagine most professional pilots, except those flying in a few select areas are not real proficient or at least not very current at circling approaches. I was actually surprised to hear that Simcom and FlightSafety teach it at all, never mind that they only teach "easy" circling. Yours truly, for example, is certified for circling only in VMC when exercising ATP and/or type rating privileges, but it's not something any of my airlines have ever formally taught or tested during checkrides. Granted, Gillespie field was supposedly VMC that night, but still...seems like a fairly iffy approach. This is an example of when regulations can lead you a bit astray. The approach plate says that circling is NA at night. Thus, if (for example) ceiling are 980' AGL, circling is no-go, but if they're 1,001' AGL, go ahead and circle all you want. Same with 3sm visibility. 3sm during the day is one thing, it's quite another at night, still another in mist.

Just some thoughts....

[Gabriel]

Couple words about circling in general. In the US, it's a fairly rare maneuver. Perhaps because it's not required very often, most runways with any significant amount of traffic in this country have at least one straight-in approach of SOME kind. Hence, I would imagine most professional pilots, except those flying in a few select areas are not real proficient or at least not very current at circling approaches. I was actually surprised to hear that Simcom and FlightSafety teach it at all, never mind that they only teach "easy" circling. Yours truly, for example, is certified for circling only in VMC when exercising ATP and/or type rating privileges, but it's not something any of my airlines have ever formally taught or tested during checkrides. Granted, Gillespie field was supposedly VMC that night, but still...seems like a fairly iffy approach. This is an example of when regulations can lead you a bit astray. The approach plate says that circling is NA at night. Thus, if (for example) ceiling are 980' AGL, circling is no-go, but if they're 1,001' AGL, go ahead and circle all you want. Same with 3sm visibility. 3sm during the day is one thing, it's quite another at night, still another in mist.

Just some thoughts....

Very interesting.

If you have to do a pattern (be it an IFR circling approach or a VFR circuit pattern) that, for any reason (visibility, terrain, traffic) would not allow for a long final, what speed and flap / gear config would you have at each stage? In particular, at what point would you have the plane fully configured and at Vref+ wind additive?

[Gabriel]

By the way... seeing that they overflew the field and did the circling at 300 / 400 ft AGL, and asked the lights to be set to a higher intensity when they were already maxed out, I don't think that they were legally VFR.

[3WE]

***In the US, it's a fairly rare maneuver.***

***This is an example of when regulations can lead you a bit astray.***

***Just some thoughts....***

Thank you for the thoughts, and forgiveness that we want to fix something every time a plane crashes. Not_Karl’s suggestions are probably the most effective, along with not_doing what the crashed pilots did.

[flyboy2548m]

If you have to do a pattern (be it an IFR circling approach or a VFR circuit pattern) that, for any reason (visibility, terrain, traffic) would not allow for a long final, what speed and flap / gear config would you have at each stage? In particular, at what point would you have the plane fully configured and at Vref+ wind additive?

No later than 1,000' AGL. There are some very limited exceptions, but for the most part the 1,000' AGL is a "hard" number.

[Gabriel]

If you have to do a pattern (be it an IFR circling approach or a VFR circuit pattern) that, for any reason (visibility, terrain, traffic) would not allow for a long final, what speed and flap / gear config would you have at each stage? In particular, at what point would you have the plane fully configured and at Vref+ wind additive?

No later than 1,000' AGL. There are some very limited exceptions, but for the most part the 1,000' AGL is a "hard" number.

And you'd be already established on final by then or you can be already at Vref and still have significant turns to make? (like base to final)

(Sorry, I intended to ask at what point in the circuit pattern rather than what altitude, I realize that I was not clear)

[3WE]

(Sorry, I intended to ask at what point in the circuit pattern rather than what altitude, I realize that I was not clear)

I'm not following what you are trying to ask here.

1) flyboy stated that he would only be doing a circling approach if the weather was firmly VMC.
2) While my trust in Dan-O has plummeted, I ass-ume, that maybe there is a DMMS that flyboy adheres to (or exceeds).
3) I do get the ideal that WRITTEN restrictions on this approach would keep flyboy from doing it, at all...the runway, and the final seem a bit short to be carrying too much extra speed...so, fully configured, turn final, chop the power as you line up/level off, and watch and see if you can slow up to Vref....(which you may be below 1000 feet going fast....) I'm guessing flyboy just ain't gonna do it.

[Gabriel]

I'm not following what you are trying to ask here.

1) flyboy stated that he would only be doing a circling approach if the weather was firmly VMC.
2) While my trust in Dan-O has plummeted, I ass-ume, that maybe there is a DMMS that flyboy adheres to (or exceeds).
3) I do get the ideal that WRITTEN restrictions on this approach would keep flyboy from doing it, at all...the runway, and the final seem a bit short to be carrying too much extra speed...so, fully configured, turn final, chop the power as you line up/level off, and watch and see if you can slow up to Vref....(which you may be below 1000 feet going fast....) I'm guessing flyboy just ain't gonna do it.

Ok so Dan says that
1. Vref is only good for when you are established in final approach when there is no more "maneuvering" to be done, but doesn't 'give enough protection while maneuvering around the pattern (or similar)
2. The airlines have a "minimum maneuvering speed" that give extra protection for maneuvering [1]
3. That GA lacks this "minimum maneuvering speed" so pilot either use Vref or improvise. [1]
4. He proposes that GA pilots calculate and use a "Defined Minimum Maneuvering Speed" (that he calls DMMS) and gives a formula for it. [1]

I tend to agree with him in that Vref doesn't give enough protection when maneuvering still needs to be done (not even in the Tomahawk) and that a minimum speed that has a margin of safety over the stall warning when banking, correction of the vertical path, atmospheric effects and less than perfect speed holding (or a combination of them) are taken into account.

Now: I don't know if the point 2 is true. I agree with the concept, but do airline pilots really have a minimum maneuvering speed that they need to keep before being established in final? So that is my question for Flyboy. I think there was something in the FAR(s) but I was looking and couldn't find anything neither in part 25 or in 121. (I couldn't find anything about minimum selectable speed either, which I know the airline pilots use).

So my question to Flyboy is if he would make a significant turn (say base to final) at Vref. And if not at what speed (and how he knows that speed).

[1] Dan and Flight Chop made a video (https://www.youtube.com/watch?v=m_tKShlf_gU) about DMMS. They asked several airline pilots and GA pilots... "Is maneuvering speed a minimum or a maximum?" All airline pilots answered "a minimum" (so point 2 has to be true), all GA pilots answered "a maximum". Obviously, there are 2 different "maneuvering speed".

One is Va, the "design maneuver speed" which is basically the speed below which if you pull up hard you will stall before breaking the wings (you will also not break the plane by a single sustained full deflection control input, although combination of control inputs like full elevator and aileron or alternative applications like rudder reversals can break the plane even below Va). That speed is written in every GA airplane POH and placarded in every GA airplane cockpit. That is the speed that the GA pilots had in mind when they answered that "maneuvering speed" is a maximum.

The other one is the airlines' maneuvering speed which supposedly gives a certain margin of safety over stall (warning?) while maneuvering. Perhaps it is the same than the minimum selectable speed. In either case, I didn't find it in the FARs and I don't know how it is established and used and what margin it gives. To add to the confusion, Vref has to be greater than the minimum selectable speed or airplanes could not fly the final approach at Vref, not on autothrottle at least.

The rest of the video goes into the importance of this concept and how GA pilots can calculate their own DMMS.

So here I am . Sorry, no answers. Just questions.

[3WE]

A lot of words to say that Vref is maybe a little too slow for turning (plus all the other nuances that might work towards a stall…)

In the light plane world we have a lot of slop 1) because our runways are long and 2. Our “fast
speeds are actually still slow speeds. Also, we probably need more gust protection (on a percentage basis)…Still, a little inattention, a little cramped, a little low, a little slow and another thing or two and 150s bite like Lears…

I am impressed with the supposed 1 knot accuracy that flyboy targets…

Theoretically, shouldn’t flyboy have a “good safe speed” for every flap setting…with 1 knot accuracy?

As Bobby would say, we are outsiders and don’t know.

One of my newer phases: “Apologies, but I am curious” if there’s anything formal or informal done with speed in an airliner maneuvering for a “tight” landing vs truly lined up on final?

PS: I recall a few ‘maneuvering’ landings during airplane rides:

-A seeming 2-3 mile final at MCI from a ~90 degree base leg.

-STL had used to have an LDA approach with a somewhat late side step (did a little banking in a 727, with corrections going on through initial flare.)

-I can recall some last minute maneuvers at Honolulu as a Hawaii airlines lined up with those diagonal runways.

I am curious if there were extra knots involved, or if Vref contains enough speed buffer??

And for the 35th time, we’re the pilots due 5 more knots of stall warning?

One final comment, we may need a thought recorder, but I’m wondering if the dude didn’t see something weird (hills+clouds) and THINK he needed severe evasive action: “Oh
Shit, Oh Shit” = I’m going to hit the hill… the scream when the plane is spinning???????)

[Gabriel]

I am curious if there were extra knots involved, or if Vref contains enough speed buffer??

To be clear... Vref may be enough speed buffer say 99.99% of the times. But the 0.01% is still too much (it would mean "not enough safety margin once every 10,000 flights, which is still dozens of times per year)

One final comment, we may need a thought recorder, but I’m wondering if the dude didn’t see something weird (hills+clouds) and THINK he needed severe evasive action: “Oh Shit, Oh Shit” = I’m going to hit the hill… the scream when the plane is spinning???????)

I wondered the same thing.

[3WE]

*** Vref may be enough speed buffer say 99.99% of the times. But the 0.01% is still too much (it would mean "not enough safety margin once every 10,000 flights, which is still dozens of times per year)***

Repeat comments:

So tack on 10 knots or Vdmmswtf…until established on final. DO IT AND BE DONE WITH IT!

AND ADJUST THE STALL WARNINGS because it’s not too hard to recover from a stall WARNING, but the list of crashes where you can’t recover from a stall continues to lengthen…including Hui Thieu Lo who wasn’t turning.

[flyboy2548m]

And you'd be already established on final by then or you can be already at Vref and still have significant turns to make? (like base to final)

(Sorry, I intended to ask at what point in the circuit pattern rather than what altitude, I realize that I was not clear)

I can carry as much as Vref+15 if I need it past 1,000' if the specifics of the approach require it. I should still have no problem being at Vref by 500". An example would be at LGA for the Expressway Visual which has since been replaced by RNAV Visual. I can carry extra speed until I finish the turn at the second Home Depot and still be good by 500'.

[Gabriel]

And you'd be already established on final by then or you can be already at Vref and still have significant turns to make? (like base to final)

(Sorry, I intended to ask at what point in the circuit pattern rather than what altitude, I realize that I was not clear)

I can carry as much as Vref+15 if I need it past 1,000' if the specifics of the approach require it. I should still have no problem being at Vref by 500". An example would be at LGA for the Expressway Visual which has since been replaced by RNAV Visual. I can carry extra speed until I finish the turn at the second Home Depot and still be good by 500'.

Thank you.

When you do that, are you following any specific procedure or similar that "mandates" that or are you using your judgement and airmanship?

Again, asking because in that video every airline pilot said that the maneuvering speed in a minimum, and I could not find any reference to any maneuvering speed other than Va (the structural maneuvering speed, which is a maximum).

[3WE]

….Again asking…I could not find any reference to any maneuvering speed…

I think you might be asking if DR is full of caca.

DMNS may be on the list of all the things we aren’t going to change in Aviation, along with TOPMS, stall warning adjustments, and Evan’s extensive screening and training and cultural programs.

[3WE]

I hit www.internet.com

Google will tell you that DMMS is 1.404 times the clean stall speed. (I’m sure that.004 is important )

That being said, DMMS hits are somewhat lacking…many go to ummmm…obscure discussion forums other than here.

I get the impression that this is a “neat idea” that does not_ really exist in the flyboyuniverse.

So, we are left wondering if Vref is not quite enough for everything plus a “steep” (30.404 degree) turn and a little pull up.

Ass hat comment: 40.4% slop over the clean speed seems a bit excessive.

Mild flame of flyboy: he hints that he might carry a few extra knots for those low, tight turns, but he isn’t throwing us a bone as to whether there’s some mathematical guidelines, or if it’s just based on his super genius?

[elaw]

1.404 sounds hella close to 1.414 which is the square root of 2, and isn't that the load factor in Gs when making a 45-degree banked turn?

Or in other words it sounds like that number was chosen to keep the plane from stalling when maneuvering, the assumption (or instruction) being that during said maneuvering, one should not exceed 45 degrees of bank.

[3WE]

2^0.5

Likely…

That being said, the significant digit thing thing is still a bad joke…1.4 might be close enough…especially if you are going off the clean stall speed…

ALSO you don’t hit 2 G until a 60 degree bank, so 1.4141414 covers you to 60 (but no slop).

[Gabriel]

1.404 sounds hella close to 1.414 which is the square root of 2, and isn't that the load factor in Gs when making a 45-degree banked turn?

Or in other words it sounds like that number was chosen to keep the plane from stalling when maneuvering, the assumption (or instruction) being that during said maneuvering, one should not exceed 45 degrees of bank.

Yes. No.

Yes, sqrt(2) is the load factor in a sustained, coordinated, constant vertical speed, 45-degree-bank turn.

No, sqrt(2) is not the correction factor of the 1G stall speed when in a sqrt(2) load factor. With sqrt(2) * 1G stall speed you can pull quite more than sqrt(2) Gs. Actually, you can pull 2Gs, which is the load factor in a sustained, coordinated, constant vertical speed, 60-degree-bank turn. That is because lift goes with speed squared, so if you keep a fixed AoA and increase the speed by a factor of sqrt(2), the load factor will increase by a factor of [sqrt(2)]^2, which is of curse 2.

So no, it doesn't make sense that the 1.404 factor for the DMMS over the 1G stall is related to the sqrt(2) load factor in a 45-degree bank. Actually, Dan explains it in the video (exact timestamp here: : it is a 30% speed margin over the stall speed in a 30-degree bank (sqrt(1 / cos 30), so the final DMMS factor is 1.3 * sqrt(1 / cos 30) which, do you guess how much is that? 1.404? Nope. It is 1.397. But Dan says that the increase in stall speed in a 30-deg bank turn (which in fact is 1.0746...) is "approximately 8%" , and 1.3*1.08=1.404. And I agree with 3WE's sentiment. 30% speed margin sounds like a lot, but it is the margin over what?

https://youtu.be/m_tKShlf_gU?t=680 (goes to the timestamp in the video where Dan explains the calculation)

Many years ago I came up with a more or less similar approach, except that I didn't want to put margins over the actual stall but over the stall warning, and tried to add more real-world conditions than a sustained, coordinated, constant vertical speed, 30-degree-bank turn. So my reasoning was, you are trying to keep a 30-degree bank but you are not perfect so make it 35 deg (speed factor 1.1), and you are typically not just keeping it sustained, coordinated, and at constant vertical speed so let's consider an additional 0.2G pull to correct the vertical flight path if you are low (or similar load factor increase due to turbulence) (speed factor 1.1) and I want that to have a 10% speed margin (another 1.1) over the stall WARNING which is required to have at least a 5% margin (1.05) over the 1G stall speed. So we get 1.1*1.1*1.1*1.05 = 1.398 which, by chance, is almost identical to 1.397 (the simpler 30% margin over the ideal 30-deg-bank stall speed). But (in my opinion) my rationale has a better justification. It is not an exaggerated 30% speed margin but just a 10% speed margin, not over stall but over stall warning, and not in an ideal and unrealistic 30-deg bank turn, but a slightly imperfectly held 30-deg bank turn and a slight additional load factor for vertical correction or turbulence. That's where MY "40% over stall speed" comes from.

[Gabriel]

Ass hat comment: 40.4% slop over the clean speed seems a bit excessive.

I agree. It seems excessive. But it is not.

And it is not over clean stall unless you are clean, it is 40% over the stall in the configuration you have. I know, in the part of the video that I linked he says clean stall, but it was a mistake, look at this earlier part:

https://youtu.be/m_tKShlf_gU?t=586

"The clean stall is 68 knots, dirty is 64. So I did the math, 68 knots times 1.404 comes up to 89.9 knots, so I'm going to call it 90 knots"

Except that it does not come up to that.
68 * 1.404 = 95.5, not 89.9
Now 64 (the dirty stall) * 1.404 = 89.9 indeed.

[3WE]

***1.1*1.1*1.1*1.05***

The other thing you have to factor in is sloppy altitude management in a turn…

The plane’s nose will want to dip, you may lose some altitude…so then you pull up a little more.

I did it A LOT.

And, I’m pretty sure it’s happened in airliners I’ve ridden.

First a turn, then 10 seconds later a few more fractional Gs.

Maybe it was arriving at assigned altitudes versus a little nose drop, but, who knows.

Same for our poor pilot here, he had a few fluctuations.

EDIT: Ok, I see you DID factor that in.

[Gabriel]

The other thing you have to factor in is sloppy altitude management in a turn…
The plane’s nose will want to dip, you may lose some altitude…so then you pull up a little more.
I did it A LOT.

Me too... intentionally.... sometimes... It is not the bank what increases your stall speed but the pull-back. So lettings the nose go down and gain speed before pulling is a good way of not_stalling.

EDIT: Ok, I see you DID factor that in.

Yes, I factored in a 1.2G pull. If that seems like not too much, boosting the 1.4 factor up to 1.45 (for context, 2 to 3 additional knots in the Tomahawk, 5 or 6 additional knots in the Learjet) buys you a 50% more pull-up power, from 1.2 to 1.3 (yes, that's 50% more because the "no pull-up" baseline is 1G, so you are increasing the pull-up part of the load factor from .0.2 to 0.3)