CZAR 52 Accident aerodynamics

[perrdan86]

I am doing a university assignment on a B-52 aircraft accident, the CZAR 52 Fairchild accident to be more precise. The assignment requires us to look at this accident from an aerodynamic perspective and im struggling tying all the elements together to prove what i think happened. I understand the aircraft used spoiler deflection to initiate a roll movement, however it kept rolling to the point where it overbanked, the wings stalled, lost lift and the aircraft crashed to the ground, as a very basic description.
However I'm struggling to support this, using the aircrafts aerodynamics, how it uses spoiler deflection to roll, its swept wings, adverse yaw, sideslip, overbanking, why bank angle continued to increase, roll rate.
any assistance would be greatly appreciated
Thanks for your time
Daniel

[john_tullamarine]

You've probably read the linked article (http://www.crm-devel.org/resources/paper/darkblue/darkblue.htm) by Tony Kern but, if not, start there to get some general background to the mishap.

The aerodynamics side of things is more generic and you will find a LOT of links on the net if your background is not, say, senior undergrad aero engineering.

[CabinMaster]

This sounds like you have no idea at all.

For starters: the B-52G & H don't even have ailerons, they use spoilers exclusively for Roll Control.

Getting an aerodynamic model with sufficient accuracy to predict transient dynamics might get difficult. The only models I came across were proprietary manufacturer's data, and then you have so many parameters that any student is surely lost.

[perrdan86]

cabin master, can you please tell me where i said the B52H uses ailerons for roll control? i mentioned twice in my initial post that they use spoilers...

Hi John! yes i have read many articles on the accident, the problem im having is that most discuss the human factors surrounding the pilot. I cant find any in depth aerodynamic analysis of the accident. Was just hoping there might be people on here with enough knowledge on aerodynamics that might be able to help me out.

anyone else got anything they can add?

[John Farley]

Have you considered what happens when a wing stalls?

[perrdan86]

hi John,
yes i have. My thoughts are that;
- he entered the initial turn, banking to about 45 degrees with an appropriate airspeed
- however, as can be seen in the still shots, as the aircraft rounds the control tower, both spoilers are deployed which is also the airbrake 2 position, therefore slowing the aircraft?
- i think then pilot attempts to move to a 60 degree AOB therefore flattening the right spoiler, leaving the left one deployed to further disrupt airflow over the left wing, incresing drag on this wing and reducing lift and increasing the AOB.
- however, i think, the aircraft was travelling below the stall speed for a 60 AOB, therefore causing the wing to stall,
- Th problem is compounded by the fact swept wings stall from the tip first. As the lowered wing reaches its critical angle first, it will stall before the top wing, it will produce a strong rolling moment due to the long moment arm from the outer sections of the wing to the centre of gravity.
- At this point not only is there a strong rolling moment produced by the stall on the lower wing, but the upper wing will have reached its critical angle also, stalling and losing any lift production it had left. The aircraft continues to bank to 90 degree, whilst doing so appears to be experience a yawing movement.
- due to all of this occurring at such a low altitude, and that a typical response to spoiler deflection is 3 seconds, any action taken by the pilot to correct once at around 60 degrees AOB it is to late.

does this seem like its on the right track? does sideslip play a roll in this aswell?
thanks for responding!!
much appreciated
daniel

[John Farley]

86

he entered the initial turn, banking to about 45 degrees with an appropriate airspeed

however, i think, the aircraft was travelling below the stall speed for a 60 AOB, therefore causing the wing to stall

Sorry my friend but there is no such thing as a simple stall speed/AOB relationship. For example you only stall at 1.414 times the wings level stall speed at 60 AOB IF you pull the stick back and try to hold height (ie actually pull the required g such that the vertical component of the lift vector equals the weight).

You can take an aeroplane to any bank angle you want but it NEED not stall. It will only stall if you reach the stalling AOA.

Good luck with what you are trying to do but my personal view is that the video by itself is not enough to establish everything that was going on aerodynamically.

If I was your tutor and you explained why you could not provide a definitive answer I would give you A+. He/she may sneakily be trying to make sure you understand that the much talked about 'stalling speed' is very often meaningless. Stalling is about AOA and little else. (I say little rather than nothing because under high IMN conditions the stalling angle of attack of a wing will reduce slightly)

JF

[bearfoil]

perrdan86

Leaving out the mindset of the flying pilot is inexcusable. This pilot's dream was to roll (360 degrees) the B-52. He had alienated most of his mates such that they refused to fly with him. He was a cowboy who should have been grounded years before this "airshow" practice circuit. His goal in this last flight was to tightly circle on base, and plant the a/c on the r/w. Once the roll left exceeded about 10 degrees, the a/c was doomed. With long and heavily loaded wings, any roll carries enough momentum to fatally affect handling. With everything BUT an agile response in the roll, this bomber takes to aerobatics like a bull to the color red. The commander was an accident waiting to happen.

rgds

[Jimmy Macintosh]

Just to throw a little something else in there:

At the low speed the aircraft is traveling, during the turn there is a delta lift between the wings causing a roll into the turn. Were the control surfaces sufficient to even over power the delta? (None of this involves a stalled wing either.)

Think to situations of differencial flap extension, this rapidly overpowers ailerons and rudder input.

Just a thought...

[PBL]

To reinforce the comments about indeterminism:-

there is nothing in the video as seen (that is, qualitatively) to show that the pilot didn't suffer from indecision or incapacity and let the nose fall through after going knife-edge.

To exclude that possibility, you would have to reconstruct the actual (quantitative) dynamics from measurements of the video, a skill which I don't recommend trying to pick up in the course of a student project assignment.......

What one might assume is that the pilot was highly skilled, well aware, and not able to roll level when the nose starts to fall through, because of some aerodynamic control impediment, for example spoilers not able anymore to generate differential lift on the wings to enable roll. But that is an assumption. You would have to validate it by analysing the true dynamics of the manoeuvre, which is not apparent without considerable reconstructive skill from the video. One might be able to tell from the video if wings-level roll inputs were attempted.

PBL

[Alber Ratman]

Simply the pilots put the airframe into a flight condition that was outside the parameters of the aircraft remaining in controlled flight. Oh, that sounds like the conclusion.

[Alber Ratman]

You will really have to find a copy of the accident report for all the details and for a verifable source of citation. However this will give you a clueon what you are looking for on the areodynamics of the accident.

1994 Fairchild Air Force Base B-52 crash - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/1994_Fairchild_Air_Force_Base_B-52_crash)

As said before, the captain put the aircraft into a stall by exceeding the limits that the aircraft could maintain that turn without stalling at that IAS. His high angle of bank and high AoA meant that the LH wing stalled first and the loss of lift would produce an continued roll to port and the nose to pitch down. Only increasing airspeed and having the height to carry out recovery action would have given him the chance to save the aircraft. He had neither the time to increase speed or height to carry out any recovery action.

Have fun with your research.:ok:

[Rigga]

Google is your friend (and easy enough to use too!)

"...The investigation found that as the B-52 entered its final turn sequence around the tower, its indicated
airspeed (IAS) was 182 knots (337 km/h). Although Holland applied additional engine power after
starting the turn, the late power application was not enough to maintain the aircraft's current airspeed
during the turn. Although the aircraft's airspeed indicator was available to all four aircrew members, the
aircraft's airspeed was allowed to continue to decrease. Eight seconds before impact, the aircraft's IAS
had deteriorated to 145 knots (269 km/h), the aircraft's bank increased past 60 degrees, and the aircraft
began to stall. Although Holland or McGeehan at this time applied full right spoiler, right rudder, and a
nose-up elevator, the aircraft continued to stall, exhibiting a behavior known as an "accelerated stall".
An accelerated stall occurs when the stall speed of an aircraft increases without an aircraft reducing its
airspeed because of environmental factors and/or the current attitude of the aircraft in relation to which
way it's moving.[6]
Due to the bank of 60 degrees or more, the stall speed for the aircraft at that moment was an IAS of 147
knots (272 km/h). Thus, flying at 145 knots (269 km/h) IAS the aircraft stalled without sufficient
altitude to recover before impacting the ground.[6]..."

"...An earlier incident occurred in 1991 when a B-52 piloted by Holland
performed a circle above a softball game in which Holland's daughter
was participating. Beginning at 2,500 feet (760 m) AGL, Holland's
aircraft executed the circle at 65 degrees of bank. Described by one witness as a "death spiral," the nose
of the aircraft continued to drop during the maneuver and the bank angle increased to 80 degrees. After
losing 1,000 feet (300 m) of altitude, Holland was able to regain control of the aircraft.[8]"

for the full report go to:
www.txwgcap.org/pdfs/B-52%20crash.pdf

[Alber Ratman]

Stall (flight) - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/Stall_(aerodynamic)#Accelerated_and_turning_flight_stall)

You really need to find a text book that states it..

The aircraft was entering a spin.. However it never got the chance to because the ground was there!

[Rigga]

...and another one!

"Crash sequence

The investigation found that as the B-52 entered its final turn sequence around the tower, its indicated airspeed (IAS) was 182 knots (337 km/h). Although Holland applied additional engine power after starting the turn, his input came too late to maintain the aircraft's airspeed, because its turbine engines take up to 8 seconds to respond to throttle movements. Even though the airspeed indicator was available to all four aircrew members, the aircraft's airspeed was allowed to continue to decrease. Eight seconds before impact, the aircraft's IAS had deteriorated to 145 knots (269 km/h) and the aircraft's bank increased past 60°. At this time Holland or McGeehan applied full right spoiler, right rudder, and nose-up elevator, and the aircraft entered a turning flight stall (sometimes called accelerated stall). This phenomenon is a stall that occurs at a higher airspeed than the design stall speed – which always refers to straight and level flight - because of the fact that the aircraft is turning. Due to the bank of 60° or more, the stall speed for the aircraft was at that moment 147 knots (272 km/h). Thus, flying at 145 knots (269 km/h) IAS the aircraft stalled, without sufficient altitude to recover, before striking the ground."

[Alber Ratman]

You are writing a BSc level Uni assignment on areodynanics?

All things mentioned are backed up here.

Stall (flight) - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/Stall_(aerodynamic)#Accelerated_and_turning_flight_stall)

To all intent and purposes, the aircraft was entering a spin when it impacted the ground. All due to a pilot disregarding and exceeding the handling limits of his aircraft.

All his other antics were released on an flight safety video.

YouTube - Mishap of B-52 at Fairchild Air Force Base Washington (http://www.youtube.com/watch?v=YQa4PpIkOZU&feature=related)

[Alber Ratman]

A bit of padding out, a few equations etc and assignment done. How many words required?

[FluidFlow]

Hi Daniel, Perhaps you can compare some captured ‘still’ shots from the videos and try to get an idea of the wing flex (angle and shape of flex) and use this to justify your conclusions re the loading on the wings at various stages during the final moments.
Regards
Ian

[bearfoil]

I don't see anything in the video suggesting the a/c stalled. It was turning, and appeared to tighten the turn before impact, but the fact that it impacted the ground was due to standard loss of altitude in a turn of any kind.

1. He may have been attempting knife edge "flight". (around a point)

2. He may have been attempting a split ess.

3. He may have thought he was hook trapping onboard the ship

4. He may have gotten so far up his own tail everything tumbled.

I favor #4

Two things come to mind

This is a rather farfetched example of "aerodynamic" flight.

Once past the limits of control authority, all bets are off.

bear

[perrdan86]

hey guys!!
wow! thanks so much for your replies!! very much appreciated! Yes as you can see im no aerodynamics expert, far from it and this course is my first real introduction to this concept of flight. Here are the comments the lecturer made on last years group who did the assignment, ive been trying to use these when attempting to break down the sequence of events. For someone of my limited background in this area, it is quite hard as there is a lot of depth in regards to aerodynamics knowledge.

Here are the comments the lecturer made on last years group who did the assignment, ive been trying to use these when attempting to break down the sequence of events.

Fortunately, the majority of you seemed to understand most of the principles well enough. Areas where some of you still had difficulty were:

1. Making the distinction between angle of bank and load factor. Many of you insisted that “the stall speed increases with angle of bank” which is not always true (how does an aircraft do a slow roll?) and many of you used the formula “ sqrt(load factor)” to calculate a stall speed at a certain angle of bank ... which is NOT correct! It is possible to get to 90 degrees AOB without pulling ANY extra ‘g’ ... the formula is ONLY true in a STEADY, LEVEL TURN (where lift is being used to generate the CPF). If you look at the video again, you’ll see that not much of the manoeuvre fits that bill!

2. Many of you had a trouble distinguishing between not enough lift to maintain altitude, and stalling. They are completely different issues. Those of you who went for the simple “it stalled” solution really needed to state exactly ‘when’ (eg X degrees AOB at time Y), explain ‘why’, and then go on to explain why the bank angle continued to increase .. and the aircraft continued to ‘fly’ quite well, albeit in a descent. Those of you who said it didn’t stall needed to come up with a very convincing case indeed, and some of you did.

3. The stall characteristics caught a lot of you out ... many people just described ‘typical’ stalls from GA textbooks. Swept wing aircraft will likely stall wingtip first and pitch up, not ‘nose drop’ ... and recovery can usually be completed simply by ‘blasting out of it’ with lots of thrust ... well, that certainly worked in the 747 and 777 : )

4. A common misunderstanding was the spool up time of the engines. The eight seconds (which very few of you referenced ... where did it come from?) refers to the time from idle thrust to max thrust .... it is unlikely that either thrust setting was used. More likely, engine response was almost instantaneous from the thrust settings they were using. Similarly, just because the engines were at low thrust at impact does NOT prove that thrust was not added before then.

5. Many of you quoted the lift formula, then asserted that the ONLY way to increase lift in a turn is by increasing AOA. Well, what about increasing speed!? Also many stated that “at 90 degrees AOB the aircraft is not producing ANY lift.” Well, there is a LOT of lift coming from the wings ... but it just isn’t opposing weight, and there is also quite a lot of lift coming from fuselage, fin, thrust etc.

6. Many people did not understand how spoilers work. They do not work like elevators or rudders. They do NOT produce an angle of bank. They generate a roll RATE ... when the required AOB is reached, spoilers are retracted (and possibly deflected slightly on the other wing) to STOP the rate of roll, then retracted.

7. There were also a lot of people who stated that there is NO lift being produced after the wing stalls. Have another look at the CL v AOA curves ... there is a loss of lift, but the wing is still producing a LOT of lift ...

8. A small number seemed to want to relate the increase in induced drag to a corresponding increase in thrust ... without realising that parasite drag probably made up a larger proportion of total drag at the time.

9. Another small number did not understand the effect of flaps (no, they do NOT increase the stall speed, nor do they “reduce the available range of AOA” either)

10. Those of you who explored the spiral instability idea had their work cut out. Quite a few simply mentioned that swept wing and high wing are stable, and that anhedral was the opposite, then somehow ‘concluded’ that the overall effect was “unstable” ... that’s a bit of a leap of faith! There was much more to that argument, and a lot of referencing work would have to be done to support it.

11. The understanding of the effect of head/tailwind was poor. Many of you would have scored more marks by forgetting it completely.

Thanks again all for your input, again i cant emphasize how much it is appreciated!
Daniel

[perrdan86]

Hi John,
with regards to what you said about speed and stall, i was wondering, from what ive read at for example 60 degree AOB the effective weight of the aircraft is doubled and the stall speed is increased by 1.41. For eg if an aircraft that has straight and level stall speed of 50 KIAS, during a 60 degree banked turn, due to increase weight/load factor the aircraft now has a stall speed of 70 KIAS.
Is this the correct principle? my way of thinking was that the aircraft at a bank of 45 degrees was travelling at a speed above its "stall speed, however due to slowing down because of the spoiler being in the speed brake 2 position, the aircrafts speed was decreasing and was therefore below its minimum speed for a banked turn at 60 degree.
Or would it have to do with the lower wing reaching its critical angle therefore stalling first as it has a higher AOA than upper wing. I would have thought there was a relationship between speed and stall as in the forumula for lift contains dynamic pressure, therefore if this is not sufficient, not enough lift is produced??

sorry if that is a silly question, im just trying to understand what you are saying and what my txt book seems to say.
Daniel

[bearfoil]

perrdan86

What your instructor also failed to take into account was that the a/c had executed a similar manouver up wind, bleeding off any energy that could have extricated him from the final hole. Full flaps (gear down?), and a second energy consuming exercise in high bank angle turns, was his downfall. That and his love affair with looking like the hot pilot he was not. If you notice his record prior, he satisfied himself with a single risky manouver at each meeting. Here at Fairchild, he introduced the "Second Phase", upping the risk. "Breaking new ground" as it were. Following a steep turn with a second (or 'continuation' of the initial one), He was behind in everything, but mostly in energy.

Bonehead.

[Intruder]

You'd best do a search on the flight and find the public versions of the mishap report. It will give you the most accurate description of the event that is available. Those mishap reports are usually quite detailed.

[PBL]

Daniel,

it was helpful to read the commentary your instructor gave on last year's group, because it indicates to those of us who have university teaching experience the level of student understanding to which the course, and the task, are pitched.

Your comment
I would have thought there was a relationship between speed and stall
touches on a topic which many people (including apparently some professionals) find hard to grasp.

An aircraft can stall at any speed.

The "stall speed" that is defined for many purposes is the speed at which, in level 1g flight, the wing will lose lift because the airflow separates. It is sometimes designated "V_s1g" to emphasise the "1g" condition. Stalling at higher speeds will involve a higher load factor (that is, greater force in the direction of the airplane "z" axis); stalling at lower speeds a lower load factor.

People also say that a wing stalls at a fixed angle of attack. Well, that is also not quite true as stated. The angle of attack at which a wing stalls is dependent on the Mach number at which the aircraft is flying. For example, the stall AoA of many commercial jets in the Mach 0.8-0.9 range is very roughly speaking 7°-10°, whereas at much lower speeds (say, M < 0.2) it is very roughly speaking twice as great.

Also, the point of stall is not necessarily a well-defined point, especially at high speed. It is more like a range of increasingly unpleasant phenomena, and a point is chosen by engineers or test pilots at which the phenomena are "sufficiently unpleasant", and there is some judgement involved in that. At low speeds, at a given load, as with many physical relationships we see in engineering and science (when things go up and then down again), there is an angle of attack at which the lift generated by the wing is maximised. Many people call this the "point of stall", but it is important to understand that there is often plenty of lift generated at higher AoA than this, just not as much as at the maximum!

Some airplanes I have flown, such as the Cessna C-152, have quite a well-defined stall "break", that is, at Vs1g going slower the aerodynamic characteristics change quite abruptly, quite noticeably to the pilot and any passengers (the low-speed lift versus AoA curve "drops off" very sharply after max lift). Then there are other aircraft, such as the Morane-Saulnier Rallye MS880B, which I have just flown and which I anticipate flying quite a lot in the future, in which you notice absolutely nothing abrupt; the airplane just starts to descend (the lift versus AoA curve reduces benignly after max lift).

So when you think about "the wing stalled" as a possible phenomenon to consider in why the B-52 crashed, you would have to know whether loss of lift with the B-52 is a gradual phenomenon, as it is with the MS880B, or a sudden phenomenon, as it is with the C-152. And here's betting you don't know!

Moving on to general considerations. Some contributors here have indicated that they think it is completely clear from the video what is going on aerodynamically. I emphasise again what I said earlier: I don't think it is obvious. Consider: if it were obvious, you would only be getting one uniform opinion from contributors here, whereas you are getting many! Also, if it were obvious then your instructor would likely not be giving you, in advance, his comments on last years answers (for example, a set of comments on answers to the question "What number is 1+1?" is going to be helpful in so far as it consists of the one word "2"!).

Bearfoil touched on a topic which others have not yet mentioned. An airplane without kinetic energy along its x-axis is like anything else: it falls. And if you start out with low energy, and then you generate lift by loading the wing or any other part of the body, then energy is going to dissipate rapidly; it is going to be passed from the airplane into turbulent air movement. And if you run out of energy, aerodynamics is moot, because the "dynamics" part ain't there, which then means, at low level, that the "aero" won't be for long either.

I think this is a good exercise! I hope you are learning a lot!

PBL

[QJB]

Stall Stick Position :):):)

You have to pull to get a stall, generally speaking at a constant speed the lift that you can create with the wing increases linearly with angle of attack which you increase by pulling back on the stick. It reaches a maximum at the stall where there is a marked increase in airflow separation over the wing. Roll alone cannot cause a stall. Otherwise how would aerobatic aircraft do 360 degree rolls with out falling out of the sky.

[john_tullamarine]

We've combined the OP's three or so threads on this subject (very naughty to have multiple threads going simultaneously).

Hopefully the edits have not made the result difficult to follow ?

[PBL]

Also a bit naughty of contributors to quote the answer extensively from investigative documents when the OP had identified it as a study project. Explanations, clarifications and hints from third parties are, for most teachers, welcome, and part of the learning process. Answers are not!

PBL

[John Farley]

you only stall at 1.414 times the wings level stall speed at 60 AOB IF you pull the stick back and try to hold height (ie actually pull the required g such that the vertical component of the lift vector equals the weight).

I should not have put 'try to' in this. What I said only applies if you DO hold height.

Sorry if you don't understand this.

Only if you DO hold height with 60 deg bank will you increase the lift vector to a length of double the weight (needed to keep the VERTICAL component of the lift vector equal to the weight) and so put up your stall speed to 1.414 times the wing level case.

Rolling to a high banck angle without pulling back and increasing the AOA will NOT affect the stall margin. What will happen to your flight path is another matter though.

[sycamore]

Perrdan86, if you have the time or inclination ,why not find a `flying school`,and go and speak to the CFI,or even a simulator,not a microsoft one, to do a practical sortie, when you have looked at all the relevant information,and done some analysis.( I`m assuming you have`nt done any practical flying ).It doesn`t have to be in an exotic jet,as the same accident occurs to people who don`t pay attention to what the aircraft is `telling you`,either thru` `fixation outside`,trying to line-up by overbanking,pulling too much in a descending spiral instead of `rolling-out`,then pulling,or any combination.
We normally say ` stall ,flick ,spin,crash, die !`; it happens in Piper Cubs,to B-52s.......

edit later; having watched the `u-tube` videos,at that height there was no going back ,after about 60 deg. of bank,and there was little or no power applied,going by the smoke trail,compared to the turn after initial t/off. There was full opposite spoiler applied, but the nose had gone down,and it appears little `g` applied(wing-flex),and it appears full flap,with lots of drag...