Why is the t/o climb gradient requirement higher for quads than twins?

[NiclasB]

(Also known as: Yet another take-off climb gradient question)

Hi all,

AFAIK, the gross t/o segment 2 climb gradient requirements are
2.4% for a twin (1.6% net)
2.7% for a trijet (1.8% net)
3.0% for a quad (2.0% net)

Since these numbers are with one eng inop, I expected the quads to have a *lower* gradient requirement than twins, not the opposite. I figured that quads would have a larger "engine" error margin (3 op vs. 1 op) thus allowing for a smaller gradient error margin. :-O

Can anybody explain why quads have a higher certification climb gradient? Is there a flaw in my reasoning? Or is the reason that at the time of legislation, the twins simply could not produce 3.0% (without large TOW penalties)? Kind of like the 35/50 ft class A/B t/o screen height difference, which I understand was an adaption to contemporary (DC8?) performance available rather than "equal error margin" reasoning.

I would appreciate an answer very much. I've been banging my head :ugh: at this question for some time now. Have tried google, pprune, etc. to no avail.

Many thanks,

/Niclas, ATPL student, trying - perhaps too much - to understand the *reasoning* behind most of the numbers...

[Mad (Flt) Scientist]

The real rationale will be buried in the hostory of the FARs, BCARs, etc., and would be interesting to unearth.

However, here's a possible reason, purely speculative at this point.

The required gradient represents a "margin of safety" - the higher the gradient required, the more "safe" the aircraft.

Quads are more likely to suffer an engine failure on takeoff - roughly twice as likely as for a twin. So to counterbalance the increased risk of OEI operations, the requirement for a quad is higher.

In other words, quads are more likely to OEI, so have to do better when OEI, so as to offset the risk of other, additive, factors coming into play.

As I said, just speculation after the fact - may have nothing to do with how we got to this point at all ...

[NiclasB]

MFS, thanks for your suggestion. My thought *have* traveled that route, and I agree that a quad is more likely to have an (one) engine failure than a twin on pure statistics alone. However, since the gradients are calculated based on that we are already in OEI conditions, I figured that "safer" would mean "have a larger margin to disaster" and that 3/4 eng op would mean a larger margin than 1/2.

By the same statistical reasoning, you could argue that the probability of *another* engine failure of one of the remaining three quad engines is larger than failure of the remaining single twin engine, but I kind of thought that would be offset by the quad being more flyable on two remaining engines than the twin on none! :cool:

Still confused, not necessarily on a higher level... :confused:

[Rwy in Sight]

SLF here with practically no science background. Funny this question appear just today

May I approach the question "the other way around". The quad having lost an engine still has 75% of installed power available versus 50% on twin. Hence the numbers allow/impose the quad to have a better climb performance should a problem appear later.

Feel free to ignore my comment with or without explaining why.

Rwy in Sight

[NiclasB]

Rwy_in_sight: Well, your "we ask more of the quad because we can" would explain the gradient requirements. With my scientific background (math, computing science), I tend to like the "same risk for all planes" principle better.

Perhaps asking a legislation to follow simple, sound principles, is asking too much. Then again, who am I to ask...

At this point I have surrendered to accepting the fact that quads have higher requirements. Why? Because the legislation says so! Still would appreciate any more enligthened input.

[FlightPathOBN]

these are cert specs...

in some ways, you answered your own question, in regards to certification, why wouldnt an ac with 67% or 75% of power, NOT have a higher min CG than one at 50%?

[Mad (Flt) Scientist]

OK, a quick check of the historical FARs (http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgFAR.nsf/MainFrame?OpenFrameSet) shows that 25.121(b) has been similar since day 1 (in 1964):
Takeoff; landing gear retracted. In the takeoff configuration existing at the point of the flight path at which the landing gear is fully retracted, and in the configuration used in Sec. 25.111 but without ground effect, the steady gradient of climb may not be less than 2.4 percent for two-engine airplanes, 2.7 percent for three-engine airplanes, and 3.0 percent for four-engine airplanes, at V2 and with--

So we need to go back further ... CAR04b as of 1953, in the equivalent section 4b.120(b) required a rate of climb of not less than 0.035 Vs1 - so it used to be independent of number of engines, and greater than the current standard. At some time between 1953 and 1964 the rules changed - it'll be in the paper trail somewhere.

[FlightPathOBN]

very good! :ok:

now...

sense the frustration when designing procedures for aircraft

when some of the FAR's were founded in 1964...

Reality, if one designed a approach/departure procedure ...based on the criteria..it would either never work, or operators would say, I am not using that...

[Rubber Dog]

From a performance theory book I have found the following sentence,

" To ensure the safety probability level of 10 to the power of -6 is preserved the gross gradients achieved by any class A aeroplane are diminished by
0.8% for a twin engined
0.9% for a three engined
1.0% for a four engined"

It could therefore be all about the probabilities. Is my reasoning the right way round? It's a bit early in the morning here ;)

[CliveL]

NiclasB
I tend to like the "same risk for all planes" principle better.
Perhaps asking a legislation to follow simple, sound principles, is asking too much.
At this point I have surrendered to accepting the fact that quads have higher requirements. Why? Because the legislation says so! Still would appreciate any more enligthened input.Rubber Dog
" To ensure the safety probability level of 10 to the power of -6 is preserved the gross gradients achieved by any class A aeroplane are diminished by ......
It could therefore be all about the probabilities. One of my treasured possessions is a very battered copy of a 1953 report from the ICAO Standing Committee on Performance which describes in great detail the methods used to set up recommended TO performance requirements.

The maths is fairly complicated, but in essence they calculated the standard deviation of flight path gradient coming from variations in a whole range of parameters from their 'nominal': Engine power (reciprocating engines), thrust (jet engines), drag, weight, atmospheric pressure, humidity, airspeed, probability of failure to feather etc. This was done for twins and four engined piston driven aircraft and four engined jets (there were no twin jets in 1953)

They then combined the resulting standard deviation with the probablity of powerplant failure to get graphs of the gradient 'margin' required as a function of 'incident probability'. This gave separate lines for twins and four engined aircraft, a difference which appears to be driven mostly by the additional failure probability of four engines. The ICAO team did not recommend any particular incident probability as that would have been outside their terms of reference, but their data allowed the Airworthiness Authorities to make an informed decision.

The NFP gradient was then derived by an assumption that at least 0.5% gradient would be required for all classes of aircraft plus another 0.1% as an allowance for an 8 deg banked turn plus whatever margins deemed necessary by FAA/ARB etc as determined above for their chosen incident probability.

So, as far as I can see, the "same risk for all airplanes; following simple sound principles" was followed, and it was "all about probabilities"

Hope this helps!

CliveL

Edit: looking back at earlier postings, this seems pretty close to what Mad (Flt) Scientist wrote.

Edit (2): On reflection that 0.5% basic gradient doesn't seem to fit with todays requirements, but you have to remember that in 1953 they were dealing principally with piston engined aircraft and that the airworthiness authorities could (and did) up the ante for jets. I don't think this changes the general principle though

[HazelNuts39]

Mad(Flt)Scientist and CliveL;
You may also be interested in the 'missing link' between the SCOP Final Report and FAR-25, which is Special Civil Air Regulation No. SR-422, SR-422A and SR-422B, see the FAA's Flight Test Guide for Certification of Transport Category Airplanes; Appendix 4: History of Jet Transport Performance Standards Advisory Circular No. 25-7B (http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgAdvisoryCircular.nsf/0/9f2b881724fefdf486257868003f38ee/$FILE/25-7B.pdf) (pdfpage 327)

Regards,
HN39

[CliveL]

Thanks for the link Hazelnuts - 1957 - that's just about where I came in !!

Clive

[NiclasB]

Rubber Dog: Your cited percentages (actually percentage units) correspond exactly to the difference between the gross and net climb gradients. I knew the net gradients were used for obstacle clearence. Now I know the rationale behind their calculation. Thanks. :)

CliveL: It sure helped! Thanks a lot! It then looks like the main factor was "more engines => higher likelyhood of an engine problem => higher requirements".

CliveL:
So, as far as I can see, the "same risk for all airplanes; following simple sound principles" was followed, and it was "all about probabilities"


Agreed. I am relieved. ;)

All: Aeronautical (legislation) archeology at its best! :D

[bubbers44]

We transferred our 4 engine Jetstar to Reno and wanted an engine out procedure so copied what the airlines used for their 2 and 3 engine aircraft. Because of our climb gradient of 3.0% we felt we were conservative because we may have been able to climb straight out rather than the turn around Rattlesnake peak that they used using DME to start the turn. Maybe the departure procedure is different depending on climb gradient?

[john_tullamarine]

copied what the airlines used for their 2 and 3 engine aircraft

Potentially very risky - the NFP is the sum of the lot, not just the gradient for whichever segment. No problem starting off with a procedure pinched from elsewhere but it needs to be checked rigorously against the AFM for the particular aircraft to identify any gotchas -

(a) all the runway numbers need to be looked at - including TOR, TOD

(b) V1/VR might have some significant variation - is clearway involved ?

(c) how do the first segment performance characteristics differ ?

(d) how do the third segment accel distances differ ?

Really it is a better approach to have the AFM on the table with the obstacle profile and figure out the clearances for your particular bird.

[oldchina]

"why wouldnt an ac with 67% or 75% of power, NOT have a higher min CG than one at 50%?"

Because a 4-eng plane has less total installed thrust than an equivalent twin.

Designers only give an aircraft the thrust it needs - to meet takeoff and other requirements - for example they didn't put four A330-300 engines on the A340-300.

[bubbers44]

JT, not as risky as climbing straight out with no performance data. We had no engine out performance procedure for Reno. Our speeds were compatible with the 727 so knowing we could outclimb that by .3% we were golden. We had a private jet so didn't have data for Reno. We used Western Airlines procedures. We had airport altitude and field length data but no procedure. I think we did the only thing we could for a safe operation.

[john_tullamarine]

not as risky as climbing straight out with no performance data

I'm not disputing that sort of consideration...

so knowing we could outclimb that by .3% we were golden

I'd still lose sleep over that approach - speed determines radius of turn and the first and third segments might well still present some problems.

If you had the procedure, presumably it gave some information regarding the critical obstacles - did you, at the very least, check your clearances from it/them ?

[FlightPathOBN]

I think we did the only thing we could for a safe operation.

Based on?

You didnt mention your aircraft, so it is difficult to quantify this...

[decurion]

Also interesting in this respect is the article by Joop Wagenmakers:

http://www.smartcockpit.com/data/pdfs/flightops/aerodynamics/Review_Of_Performance_Requirements.pdf

[aterpster]

bubbers44:

JT, not as risky as climbing straight out with no performance data. We had no engine out performance procedure for Reno. Our speeds were compatible with the 727 so knowing we could outclimb that by .3% we were golden. We had a private jet so didn't have data for Reno. We used Western Airlines procedures. We had airport altitude and field length data but no procedure. I think we did the only thing we could for a safe operation.

As I recall Western's OEI turned right from Rwy 16R over town, rather than the "death trap" (my term) left turn through Ratttlesnake gap.

With the performance of must of today's twins, straight-out for 16R OEI is the safest. That wouldn't work in many cases circa 1980.

I suspect J.T. would have heart failure looking at the Rattlesnake left turn.:)

[aterpster]

When I saw my company was going to use this for the 727-100 circa early 1980s I decided not to bid KRNO flights:

http://i201.photobucket.com/albums/aa214/aterpster/RattlesnakeOEI.jpg

[john_tullamarine]

That's why I prefer our splays - although I can't see the elevations from the graphic, I suspect that the scenic tour up the valley wouldn't accommodate the expanding splays. With a bit of wind analysis, the valley would just disappear totally ..

Heart failure would only be a consideration were I to be on the aeroplane ...

Surely that could only be countenanced as a day VMC departure option ?

[aterpster]

J.T.:

This was in the pre-splay days. 300 feet each side of centerline. It was the book OEI track, which made it weather independent.

The procedure, as I recall, was a flaps 5 takeoff, which made it easy to obtain the required KIAS, something on the order of 170 or 180. You began your 15 degree banked turn on an ILS DME fix then held the bank until you achieved the heading that would hopefully take you out of Dodge.

I believe the 727-100 with an engine failure just above V1 would cross the physical end of the runway a fair amount higher than 35 feet. But, to retain that margin the airplane was weight limited on a warm day.

So far as I know, it was never used.

[PhilWebb2]

I believe you have made a mistake in your chart listing the climb gradients.

2 eng gross gradient is 3.2 and net is 2.4.
4 eng gross gradient is 4.0 and net is 3.0

You have taken the net and subtracted the .8/1.0 instead of subtracting it from the gross

[bubbers44]

The rattlesnake peak turn was fun but noticed on departure if you lost the DME where you start the turn if you lost an engine in the sim you were on your own. We had an approach in there one morning and only the NDB worked. We followed United and they didn't clear the runway so went around asking do we follow the NDB missed approach procedure which was a left turn back to the north. They said yes so we started our turn and the controller said hold your heading, which was right at a mountain and when I asked what heading they disregarded and did the published missed to avoid terrain. Sometimes controllers forget about mountains and just think about midairs in a panick situation. I loved flying there but you had to pay attention.

[aterpster]

bubbers44:

JT, not as risky as climbing straight out with no performance data. We had no engine out performance procedure for Reno. Our speeds were compatible with the 727 so knowing we could outclimb that by .3% we were golden. We had a private jet so didn't have data for Reno. We used Western Airlines procedures. We had airport altitude and field length data but no procedure. I think we did the only thing we could for a safe operation.

So, taking off to the south if you had an engine failure just above V1 you would do a right 180 (more or less) over town? I recall that is what WAL did many years ago.

As to the 727, the option for the biggest engines would seem to make a difference.

War story: When TWA first planned to operate into Reno in the early 1980s they planned to use only 727-100s and only fly to KLAS. They told our ALPA safety committee with OEI that would permit a straight out climb to a fix in the Carson Valley where a climb in hold would occur. I got the FAA form for the fix and advised them they would enter the hold 1,800 feet below MRA for the fix. Lots of red faces on the company side of that table.

Of course, they also asserted all TWA airplanes could climb straight-out everywhere with OEI. Our ALPA safety committee convinced a senior VP to get us the altitude each mile for a 727-231 at max structural takeoff on KLAS Runway 25 at the critical temperature for that weight. It took 31 miles to get to 1,500 agl. The airplane would have hit the first ridge of mountains west of town. That changed TWA's OEI procedures, big time.

I found this incredible from an airline that sold performance data to other airlines, etc.

Since then, I generally do not trust performance engineering, except JT of course.
:)

I think a couple of these for-hire performance engineering departments often "cook the books" so to speak.

P.S. I think more likely than not, with an engine failure just above V1 in IMC or at night, more likely in those days there would have been CFIT into Rattlesnake Peak regardless of airline.

[Intruder]

AFAIK, the gross t/o segment 2 climb gradient requirements are
2.4% for a twin (1.6% net)
2.7% for a trijet (1.8% net)
3.0% for a quad (2.0% net)

Since these numbers are with one eng inop, I expected the quads to have a *lower* gradient requirement than twins, not the opposite. I figured that quads would have a larger "engine" error margin (3 op vs. 1 op) thus allowing for a smaller gradient error margin
A "quad" has 75% power (1/3 of it on the side with the inop engine) available with 1 engine inop, a tri has 66%, and a twin has 50%. It is VERY logical that the 1-engine-inop numbers would reflect that fact.