Flying on the 'step'
 

Reading 'Fate is the Hunter' - awesome book (where can I get a job flying stuff like that?)

In one chapter, Gann mentions climbing marginally above target alt (no such thing as a level bust then I guess), then descending to target in order to hit cruise speed at a more economical power setting. He calls it flying on the 'step' and claims it gives better performance than if you just level out at target and set cruise power.

I have previously heard of this technique and tried it in a DR300 - it does seem to work, giving us around 5kt more for a given cruise power setting.

Not sure how this works - can anyone shed any light?

Most transport type aircraft (as opposed to fighter or bomber type aircraft) have an Operating Data Manual with a specific target speed for a given FL and ISA deviation. You climb at the recommended power settings or IAS or whatever (as specified by the ODM) until your planned route FL. You are taught by your primary flying instructor as a basic flying technique to anticipate the level off, so these lessons continue in later life and you reduce the power and start the forward movement of the stick shortly before the levelling altitude (usually only about 50 - 100 feet before, depending on the climb rate of the aircraft). and so you arrive, with the altitude absolutely hacked, but still at the climb IAS, not the route IAS.

Once levelled, you now wait for the engines (at the cruise power setting) to do their job. Sooner or later, they do. Eventually, because you are level, not climbing, the speed increases and you arrive at the target speed.

But all of this takes time. Some people believe that if you climb to a couple of hundred feet above the cruise altitude, still at the climb IAS/power settings or whatever, and then convert the excess altitude into speed by diving, you will arrive at the target IAS a bit more quickly, which means that you can then settle down into the cruise a bit earlier.

For me, the jury is still out on this one. Looking at it theoretically, it seems to me that it is a transfer between kinetic energy (speed and engine power) and potential energy (altitude). As energy can be neither created nor destroyed, it shouldn't make any difference which way you do it. However, practically, whenever I've done it, it always does actually seem to me to appear to be working, and you do seem to arrive at the target cruise IAS a bit more quickly if you do climb above the target altitude and then descend to pick up the target speed. I'm sure it can't be true, but it just always seems that way when you're doing it.

Yes exactly! I too had been through the 'maths' of it and couldn't see how it works. However, like you, I found it does seem to somehow... hence the question!

The referenced technique was especially useful in large propeller driven aircraft...specifically DC-6/7, 1649 Constellation, and F.27...the types with which I am most familar.

Climbing several hundred feet above the target cruise altitude, with climb power still set, then beginning a slow descent, to the desired altitude, resulted in a more rapid acceleration to the desired cruise airspeed.
This was especially useful with the F.27/FH227B, wherein the climb power setting was also the cruise power used.

Having said all this, when it is all averaged out, the resultant cruise speed achieved using the above referenced technique, will not be much greater than otherwise, using the AOM method, over the long haul.

I think you've hit the nail on the head. When this came up elsewhere I did some very simple numerical integrations in the form of a spreadsheet...

Two aircraft with identical aerodynamics. Both aircraft climb at 55 m/s (just faster than min drag speed of 50 m/s) at an angle of about 2.5 degrees.

One aircraft gets to its chosen altitude and levels off. It accelerates to its cruise speed of 77 m/s. It reaches 65 m/s after about 30 s, 70 m/s after about a minute and is within 1 m/s of its cruise speed in a little over 2 mins.

The other continues the climb at 55 m/s for 20 s, so it gets about 50 m (150 ft) higher. It then pitches down and flies the reverse (2.5 deg downwards) for 20 s before levelling off. At that 40 s point, it's going about 1.8 m/s faster than the other aircraft. But it accelerates more slowly. After about 90 s, there's still a 0.6 m/s advantage in speed (about a knot and a half, probably the smallest noticeable increment that the step-believers would demonstrate).

And eventually, they reach the same speed of 77 m/s. The step is illusory, but the illusion persisted for a good while. The aircraft that performed the zoomy manoeuvre spent an extra 20 s in the climb at a lower speed. If you track the distances travelled, it never quite catches up with the one that simply levelled off.

I remember that "on the step" story. It seemed so plausible, pariticularly in a prop aircraft. And so many other little anecdotes from Fate is the Hunter. He was a great writer, simple and to the point, and I was enchanted.

Then, many years later, I read a few of the books he wrote after retiring from flying, buoyed by what he'd earned writing Fate etc. They were about sailing and I was absolutely horrified at the risks he took. He was still very much to the point but what showed through - to my by then jaundiced and cautionary eye - an elemental carelesseness.

The chief pilot at the last company I worked for used to do this. I thought he was retarded. He used to also remain 200 ft high to remain out of the tops of cloud, instead of requesting a higher altitude. The funny part was, he thought he was being sneaky by resetting his altimeter so that is still read the correct altitude. Unfortunatly, his pea brain didn' t know that encoding altimeters send pressure altitude to the ATC radars, so it doesn't matter what setting you have in the altimeter.

Gann wouldn't have got caught out by an altimeter, he'dve dangled his HF antenna to feel the ground.

I'm not sure if the following applies to larger aircraft but I have observed the effects in light a/c and gliders:

Some things I have flown (high performance gliders, especially) seem to have a 'bistable' effect in terms of the wing performance.

The theory is that at some airspeeds, there exist two stable states for the airflow over the wing. You can be in either state, depending on whether you have speeded up or slowed down to get to it. I have heard aerodynamicists mutter about laminar attachment, separation bubbles and other techno-gabble but the effects are real (in some particular cases).

A sailplane I used to own climbed noticeably better if you approached the optimum thermalling speed from above, rather than below. It had boundary layer control devices over most of the span and I reckoned there was a small range of speeds where there was definitely some hysteresis in the performance polar.

It's similar to 'planing' seacraft, I think. You can motor along just below the 'critical' speed and be using a lot of power. You can then accelerate over the 'step' to get the boat in a lower drag configuration, then slowly reduce the speed to what it was previously, finding much less power is required to do the same job.

>You are taught by your primary flying instructor as a basic flying technique to anticipate the level off, so these lessons continue in later life and you reduce the power and start the forward movement of the stick shortly before the levelling altitude (usually only about 50 - 100 feet before, depending on the climb rate of the aircraft). <

Well if he taught you this, he taught you wrong, Oxford Blue!

Yes anticipate by, say, 50 - 100 feet (or perhaps 10% of the Rate of Climb) and then level off and accelerate to cruise speed WITH CLIMB POWER MAINTAINED until the target cruise speed is reached (or maybe one or two knots before because it takes a finite time to reduce to cruise power).

That said, the step technique does seem to have some credence.

A good point. However, the only polars I've seen with hysteresis were those depicted for low Reynolds number flow, aimed at model aircraft. And for those, the hysteresis was the wrong way round: approaching a particular lift coefficient from above resulted in lower drag than approaching it from below.

Interesting. I think it may have had something to do with 'turbulation' and this not being so effective when approaching from the low speed end. I remember talking to the designer about this particular glider when I went to pick up one of the very first ones from the factory. They had done extensive wind tunnel and free air testing and were sure that it would climb faster and glide better than anything else. In very smooth air this was true but 'microturbulence' was it's downfall. Unfortunately there's a lot of it about in you average thermal...

I don't think any manufacturers admit to the possibility of bistable polars, even though they talk about it privately. I think it happens over quite a narrow speed range and only in some implementations. Also Re. numbers are higher in this instance.

Just as it requires more energy to disturb an object from rest (to set it in motion) than to maintain an object in motion, it requires more energy to reach the cruise speed and altitude than to maintain it.

Nowadays this is achieved by accelerating to the a higher speed than the cruise speed and then reducing to crusie power.

If your max power is your cruise power then you can achieve your desired cruise speed either by waiting for the weight to decrease (thereby reducing the energy required) or by climbing above the selected altitude and converting the potential energy.

I think we are mixing several different scenarios.

1. Cruise by cruise power setting (piston and turboprops?)

2. Cruise by target cruise speed (jets?)

In the first case you are adding a constant ammount of ernergy and the speed achieved will depend on the overall energy equation. By climbing 200´ above your cruise speed and then descending and acelerating you are adding more energy than in a normal climb because you are using climb power for longer. I guess the same effect could be achieved if you maintain climb power for an equivalent amount of time while level at the cruise altitude.

In the second scenario you climb at climb power but you cruise at the target speed. This means that the energy added is the necesary to achive the selected speed. Climbing higher and descending to accelerate will only achieve a shorter time at max cruise power or a period of time at a reduced cruise power.

You are quite correct about constant cruise power settings in old 4 engine piston transports, calypso.
These were always cruised at a constant BHP/BMEP, and as the weight reduced, the TAS increased slightly.

Jets of course turned this scenario upside down...some old piston drivers simply could not adapt...or found themselves in a jet upset situation when they climbed too high (for their weight) and slowed down (sometimes way down) for turbulence.
Bad news if tried.

411A,

That's almost but not entirely accurate. I flew one of those old 4 engine piston transports (Wright R3350 turbo compounds), albeit a military one for which range was more important than schedule.

In that case, the optimum approach was to reduce RPM (with max BMEP) to cruise at the best range cruise speed for the weight. This meant that the last portion of the flight would be flown at some 15% slower than the initial cruise. (Made for a long night: I once flew a 20 hr flight, returning to Nova Scotia while retaining Jacksonville FLA as my alternate. Very glad wx above minima on arrival.)

For fast cruise, which commercial airliners used as you said,(unless on the hairy edge of range), indeed the TAS would increase by as much as 7-9% at max cruise power.

Was much more comfortable when I could get back to jets which have essentially no range penalty to fly at high cruise speed when at high flight levels.

Your coffin corner situation appears to have been found by a Pinnacle crew.

EGBK, you say you tried this in a DR300, with the Robin cranked wing. If the DR300 is at S&L flight, the outer cranked part of the wing produces no lift, and hence reduced drag, the lift is therefore from the main part of the wing. But to get to the 120 kts required you climb above the required height and the dive, when you reach 120kts you find you can maintain that speed in level flight.

However if you pull back on the stick just slightly the speed reduces, and you come off the step. Lowering the nose again to S&L you wil find you are at 110kts S&L, unable to get back to 120kts.

How the step applies to aircraft without this cranked wing I dont know.

Yakker - having a brain-dead moment here. How does a cranked wing produce no lift s&l? Surely it will develop a lift vector perpendicular to the surface?

I also dispute that by raising the nose and lowering it again you won't get back to 120. If you raise the nose, you'll slow down and probably climb. If you lower it, you'll speed up and probably descend. Once back in equilibrium (which may take some considerable time - see other posts), power+attitude = performance: the same as before, since nothing in the force equation has changed.

This step thing is fascinating isn't it!?

Does lift not vary with angle of attack. At a certain angle of attack the wing produces no lift. The cranked outer part of the wing has a different AoA to the main part of the wing. Hence at S&L the main wing produces lift, but the outer part does not.

You will not get back to 120 flying S&L as the drag is a problem. By diving you reach the Step, and can maintain S&L at 120.

Personally, I'm sceptical of the entire concept of "the step" in cruising. Whatever...

It should, of course, be pointed out that such a technique is a total no-no in RVSM.

Doh! Cheers Yakker - worked it out meself in the meantime. Must be these long shifts...

I personally think this whole business of flying on the step is a load of c**p.

What decides a final cruise TAS is an equilibrium between thrust and drag, remember, excess thrust can either make an aircraft accelerate or climb, once the total thrust is equaled by the total drag you will have level, unaccelerated flight.

I can appreciate that climbing above and then diving down will seem to yeild a higher speed or the same speed in less time than a normal level off and acceleration, but unless the two methods can be made under the exact same thrust / weight / ambient conditions any conclusion that is derived is purely subjective.

If anyone can post scientific data derived from proper flight testing that shows existence of the mythical "Step" I would be very intersted.

What about the Concorde on final approach, using ~60% power (not completely sure about this though) to fly at 15% of cruising speed?

Dufo,

All aircraft will require more power to fly at slow airpeeds such as on approach vs higher airspeeds such as LRC or max endurance, this is due to the effects of the two primary types of drag an aircraft has to overcome; Induced drag, which is greatest at higher AOA's and low airpeeds and gets less the faster you go, and Parasite drag which is proportional to airspeed and gets greater the faster you go. The aircraft has to overcome the combined effect of these and hence if you were to draw a graph of TAS vs total drag you would get a U shaped curve. From this it is possible to see that an aircraft can actually be flown at two different airspeeds with the same power setting, albeit the lower of the two speeds is substantially less than the minimum speeds typically used on approach. This is one of the reasons that primary students are taught slow flight.

The other reason an aircraft will use more power on approach is simply to overcome the drag of the landing gear and lift enhacing devices such as slats and flaps, and lets not forget in the case of the concorde that drooped nose has to add quite a bit of a drag.

And before anyone jumps on board and says "See, you just said it, you can go faster with less power!" just remember, the minimum power required occurs at L/D max for any aircraft which is typically around the airspeeds required for max rate of climb (since max ROC occurs at a speed where there is the greatest excess of thrust) as you go any faster than this speed (typical cruise speeds are substantially faster) you will require more power, period.

Diving down to an altitude may get you a higher TAS initially but sooner or later the laws of physics will prevail and you will get the same airspeed. In fact, if you think about it, the diving down method will be less efficient since you will have climbed higher than necessary and wasted fuel in the process.

Calypso,
If you look at your power tables for a single engine piston you will be able to find values for TAS and power required to achieve this. These are the required power settings to maintain the speeds not the required power (energy) to reach them because to accelerate a body requires more energy.

If you use the technique stated by firefly bob you will also reduce the shock cooling effect on the engine.

Strange no-one has considered the total drag curve in this thread.

Recall that it starts high at low speed and then reduces to min drag speed from which point up is the only way it goes. I know there may be a little excursion around transonic.

So for a fairly wide range of power settings we have two speeds having equal total drag.

A delta wing has a broad range of speed around min drag where there is little variation so if we use that as an example the effects are magnified.

Assume a flat bottom of the total drag curve over about 30 Kts as with the Vulcan. Now you have a 30 Kt range of speeds at which you can stabilise for a particular power setting.

If you want to cruise within that 30 Kt band of speed then you would be foolish not to accelerate beyond the speed band, set the power and let the speed stabilise back to the high speed end. Otherwise you could be stuck all day at the low end.

Most non deltas will also have a small band of speeds at the bottom of the total drag curve.

So - best method is to hold climb power after levelling, overshoot the expected cruise speed, reduce to cruising power and you should stabilise at the higher value.

Sometimes turbulence or other disturbances will cause you to slip back to the lower speed. More power will be needed for a while to regain the faster speed.

All this is irrelevent if your speeds are further up the total drag curve.

Milt:

If you had checked back a couple of posts you will see that I had mentioned the total drag curve. The problem that exists with flying at the bottom of the curve as you point out, is that these speeds are far slower than what we usually cruise at, as I mentioned before, the TAS that equates to minimum drag (L/D max) is in the same vicinity as the speed for max ROC

Cheers! :ok:

This is what I love about flying - always something to learn and there are still some things that aren't fully understood!

Thanks for all your input folks - really interesting and educational for this newbie...

Yakker and EGBKFLYER,
I think the logic is a little incorrect regarding the cranked wings: there will simply be a differing degree of lift depending on the AoA, not lift and no lift. I don't think aeroplane designers are in the business of adding bits that 'just travel along for the ride' while adding drag!

212man

It was my understanding the Jodel/Robin wing was efficient due to the tips being at a lower angle of incidence than the rest of the wing to reduce drag at cruise attitudes. The tips only providing significant lift at take-off and landing. The tips are also angled upwards to provide dihedral for stability and to reduce wingtip vortex drag.

Why would you bolt on a section of wing that rides along at a less efficient AoA unless there was an overriding necessesity? I suggest the washout is for stall development control ie to induce a root stall first.

More or less what I thought!

This link may explain more about the original post topic:

http://web.usna.navy.mil/~dfr/flying/step_wide.pdf

In the P2V7 (aka SP2H or 4eng Neptune) you could engage the autopilot in baro hold and then quickly gain up to 12kts indicated (190 to 202 rings a bell) by trimming a load onto the elevators via the autopilot's electric trim-wheel (with auto-trim declutched).

My theory on that was that it was loading the control surface and streamlining the trim-tabs - thus reducing trim drag. It's also possible that the resulting airplane attitude was more streamlined (and thus benign) in terms of the Thrust-Weight/Lift-Drag couples (and their equilibrium position).

It wasn't an induced static error either, because the NAVS would always confirm the extra TAS. I mulled over this for ages way back then and tried to factor in what else might have been happening with the varicam (variable camber tailplane horizontal stabilizer). Never did find a bottom line there.

Trim tab position does make a difference. In an MB326H you could hold the nose off down to ultra low speeds (for aerody braking) if you ran the elevator trim full nose-down after touchdown. It was then acting as an elevator augment.....and boosted elevator effectiveness tremendously.

OTOH, overtalk, one could always fire up the two jets for that added few knots....while reducing the range ah...just a bit.:E

As far as I can see, the only situation where an on the step level off would provide benefit is where there's a drag rise & then drop with incr. speed eg transonic to supersonic.

If the a/c has insufficient thrust to overcome the maximum drag, but sufficient to once the drag has reduced a bit then trading altitude for speed + max thrust could allow a final speed to be greater than without using a step. Effectively the additional height was used to store & then rapidly release a bit more energy to overcome the drag 'bump'.

Otherwise I can't see it. If you have thrust to climb then you have thrust to accelerate and eventually the steady state speed will occur when T=D.

OVERTALK, how did the navs confirm the increase in TAS? If it was by applying density correction to the CAS, if the CAS was wrong, their calculation of TAS would be wrong.

Unless you had Doppler and they noticed the groundspeed increasing on the Doppler, their calculations were no better than you reading the raw IAS.

tinstaafl,

Absolutely right! with the possible exception of the transonic range, the existance of the step is a myth.

We had Doppler and VOR/DME (and a very accurate radar - the APS20E).

It was a genuine TAS increase alright. Some Sqn pilots claimed up to 15kts extra IAS but I only ever saw a pretty standard 12kts.

We had the manual spark that meant that you had to go to climb power every hour (for a minute) and check the plugs out on the ignition analyzer. No matter how fatigued, you only ever got to get that wrong once.

Reclutch auto-trim
Mixtures auto-rich
Spark - from ADV to Retard
RPM 2600
50" dry Torque (cowl flaps cracked) & oil-coolers auto

Balfanging the donks and waking up the whole retinue of back-enders wasn't worth the grief (nor the half-eaten apples that eventually got to the sharp end from the galley together with the tangy coffee). Of course once they had us started into wearing chute harnesses you could almost guarantee that you'd be creating an emergency as you squeezed down back past the Tactical station and your harness tripped multiple breakers. So one always had to rely upon good relationships with your crew for sustenance.

Eventually I got so paranoid about the supercharger upshift I resolved never to use that again. Shifting both R3350-32WA's into high gear at once was the only option, you couldn't do it one at a time. Shearing a shaft was too easy.

After a few Neppies were lost due to high speed exhaust PRT breakups I was a tad paranoid about that accessory too. CSD disconnect was also a grim business if it didn't work. But by far the best trick I ever recall was the guy who lit off both AVGAS fuelled jets one dark night - without dumping them first (from standby) and forgot to turn boost pumps off and low. It's twice as impressive as the F-111 dump and burn when you see the flames from those Westinghouse J34-WE-36 dump-valves extending twice the length of the airplane.

And 411A is right. If you were prepared to go on oxygen and cruise at 30,000ft you could true out at over 330kts on all 4 - but the anmpg was down a bit.

What I miss most is arming the spoilers and making all the rear-enders sick as dogs in a MAD-trapping pattern. You could generate roll-rates as good as any fighter of the day. But once you shut down and left the airplane, you had to sprint for your life.

I was going to say the same as Tinstaafl.

The transonic regime is analogous to the planing hull scenario.

From a mechanics of flight point of view, I can imagine some interesting things happen if you are climbing at a speed which corresponds to the neutrally speed stable point at the bottom of the drag polar i.e., V_md. In this case, how you reduce the power as you approach the target altitude is critical as you may or may not wind up on either side of V_md whereupon speed stability is qualitatively different. It tends not to be a problem in commercial jets because we don't climb at speeds approaching V_md. Au contraire, I suppose in gliding, its all about V_md!

From an aerodynamic point of view it doesn't get any more difficult than unsteady (transonic) aerodynamics and without doing the testing, I wouldn't like to comment about the potential hysteretic effects FullWings mentions but essentially unless you've somehow sustained a qualitatively different flowfield, in the steady state case, there should be a unique value of velocity for the given thrust. My summation would be that in the world of in-flight Reynolds numbers, turbulence will destroy any laminar type effects that you'd prefer to persist.

OVERTALK

The fact you only saw a miserable 12kts as opposed to 15kts, is surely a reflection on your good airmanship and your view on where its sensible to keep the CG.

;)

My $0.02.

:ok: