Net Second-Segment Gradient: Why the Limiting Mass Is Not a Coefficient

A twin-engine airliner climbs steeply after takeoff with one engine feathered, transitioning from the first to the second segment of its flight path under an overcast sky at dawn.

The net climb gradient exists nowhere in the air. No aeroplane has ever flown it, no instrument displays it, and the aircraft climbing away on one engine is doing something measurably better. It is a bookkeeping convention — a fixed increment subtracted from demonstrated performance so that obstacle clearance is computed against a path the fleet will beat almost all of the time. Understanding what that increment was bought to cover, and what it was never meant to cover, is the difference between reading a limiting mass and trusting one.

Two paths, one aeroplane

After an engine failure at or after V1, the take-off flight path is described in segments defined by configuration rather than by height. The first runs from the screen height at the end of the take-off distance until the landing gear is fully retracted. The second runs from that point — gear up, take-off flap, scheduled climb speed, operating engines at take-off thrust — to at least 400 ft above the take-off surface. Then come level acceleration and flap retraction, and finally the en-route configuration at maximum continuous thrust out to the end of the take-off path.

The second segment is the one that usually decides the mass, and for a simple reason: it combines take-off mass, take-off flap and one engine’s worth of missing thrust with the highest certified minimum gradient of any segment of the climb-out.

What is described above is the gross path — the demonstrated one, corrected from flight test to the day’s pressure altitude, temperature and mass. Alongside it, the code defines a second path built by subtracting a fixed gradient increment from the gross path at every point. That is the net path, and it is the only one allowed to be used for obstacle clearance.

Where the minimum gradients come from

Two families of numbers get confused constantly. The first is a set of design minima: with the critical engine inoperative, in each configuration, the aeroplane must be capable of at least a stated steady gradient. These are requirements on the aeroplane, written into CS-25 and FAR-25 in essentially identical terms, and they scale with the number of engines because losing one of four costs proportionally less thrust than losing one of two.

Certified minimum climb gradients (CS-25 and FAR-25), by number of engines
Segment and configuration Two engines Three engines Four engines
First segment — gear extended, take-off flap, take-off thrust Positive 0.3 % 0.5 %
Second segment — gear up, take-off flap, take-off thrust, at the scheduled climb speed 2.4 % 2.7 % 3.0 %
Final take-off segment — en-route configuration, maximum continuous thrust 1.2 % 1.5 % 1.7 %
Approach climb — approach configuration, gear up 2.1 % 2.4 % 2.7 %
Landing climb — landing configuration, gear down, all engines operating 3.2 % 3.2 % 3.2 %

The second family is the net increment. For the take-off flight path, the code requires the gross path to be diminished by a gradient of 0.8 % for a two-engine aeroplane, 0.9 % for three engines and 1.0 % for four. That reduced path is what must clear obstacles, and it must clear them by at least 35 ft vertically within the horizontal corridor defined by the operational rules.

Notice what is being compared. The design minimum is a floor on the gross gradient. The obstacle criterion is applied to the net. They are two independent tests, and each produces its own limiting mass.

The reduction is subtracted, not applied

This is where the word “margin” does the most damage. A margin suggests a proportion — a factor of safety, ten per cent here, a coefficient there. The net reduction is nothing of the kind. It is an absolute gradient, subtracted whole, and it therefore costs a completely different fraction of the performance depending on how much performance there is.

What 0.8 % costs a twin, expressed as a share of the gross gradient
Gross second-segment gradient Net gradient Share of the climb capability removed Typical situation
2.4 % 1.6 % One third At the certified floor — the climb-limited mass
3.0 % 2.2 % 27 % Hot and high, near the limit
4.0 % 3.2 % 20 % A normal day at moderate mass
6.0 % 5.2 % 13 % Light, cool, sea level
9.0 % 8.2 % 9 % Ferry or positioning mass

The net gradient is a fixed reduction from the gross gradient, not a percentage. On a hot day with a heavy plane, this reduction takes a bigger bite out of the climb performance than on a cool day with a light plane. It’s not a safety buffer you can “spend” in flight.

An Airbus A320 at its climb-limited mass (MTOW on a hot day at 5,000 ft) might have a gross second-segment gradient of 2.4%. After subtracting the 0.8% net reduction, the net gradient is 1.6%. That means the plane must clear obstacles using this lower gradient, not the higher one it could actually achieve.

Read the first row again. An aeroplane at exactly its climb-limited mass has one third of its remaining climb capability removed by the convention before a single obstacle is considered. That is not a rounding allowance. It is the dominant term in the obstacle calculation on precisely the days when obstacles matter — hot, high, heavy, and flat out.

Intuition transfers badly between days. On a cool morning at low mass the net path is a slightly flattened version of the gross one; on a limiting day it is a different departure altogether, and the obstacle cleared comfortably last winter is not cleared by the same aeroplane at the same mass in August.

The trap. Treating the difference between gross and net as spare performance available in flight. It is not spare, and it is not available: the increment is the statistical allowance that makes the published obstacle clearance meaningful across a fleet and a pilot population. An aeroplane that flies the gross gradient on the day has not gained 0.8 % of headroom — it has performed exactly as the calculation assumed it would.

What the increment was bought to cover

The certification reduction was introduced to absorb dispersion, not to create comfort. Certification performance comes from a small number of flights, flown by test pilots in an instrumented aeroplane and reduced to a model. Three sources of scatter sit between that model and a line flight.

The first is technique. The scheduled climb speed is a target, not an achieved value: rotations vary in rate and in the speed captured, recognition of the failure varies, pitch control after it varies. Each knot and each degree costs gradient, and the errors are not symmetric in their consequences.

The second is aeroplane-to-aeroplane variation. Two airframes of the same type off the same line do not have the same drag. Rigging, seal condition, surface finish, repairs, antenna fits and the accumulated history of a fifteen-year-old fuselage all move the polar. Engines of the same rating do not all deliver identically either.

The third is the reduction of the flight test itself — the corrections from test-day conditions to arbitrary pressure altitude, temperature and mass, each carrying its own uncertainty.

Put together, the intent behind the reduction is that the net path should be one the aeroplane will achieve or beat with high confidence, so that the obstacle clearance computed against it is not merely an average expectation but a covered one. That is a statistical statement about a fleet, and it is the reason the increment is fixed rather than proportional: dispersion in achieved gradient does not shrink just because the aeroplane happens to be light.

Climb-limited and obstacle-limited masses are different questions

A take-off mass is not computed. It is bounded, several times over, and the answer is the lowest bound. Confusing which bound is active is a routine source of wrong conclusions — most often the belief that a longer runway would help when the binding limit has nothing to do with runway length.

The bounds on take-off mass, and which path each one uses
Bound The question it answers Gross or net What relieves it
Field length Can the aeroplane stop, or go, within the declared distances? Neither — distances, not gradients Runway length, headwind, more flap
Climb limit, often called the WAT limit Does the aeroplane still meet the certified minimum gradient in the limiting segment? Gross Colder air, lower field elevation, less flap, an improved-climb speed schedule, bleeds off, less mass
Obstacle limit Does the net path clear every accountable obstacle by 35 ft? Net A different departure track, a turn procedure, less mass, sometimes more flap
Tyre speed, brake energy Are the wheels and brakes inside their limits? Neither Less mass, less tailwind, cooler brakes, sometimes more flap
Structural and landing masses Is the aeroplane legal on arrival as well as departure? Neither Fuel, payload, alternate choice

The climb limit is the one that surprises people, because it is entirely indifferent to the surroundings. An aerodrome with a five-kilometre runway on a featureless plain still has a climb-limited mass, and on a hot day at altitude it may be the binding one. It exists because the certification code says a twin must produce 2.4 % gross in the second segment with an engine out — nothing about the terrain enters that requirement at all.

The obstacle limit is the opposite: it says nothing about certified capability and everything about what is in front of the aeroplane. It is computed on the net path, over a corridor whose width and treatment of turns are set by the operational rules rather than the certification code — rules worth reading rather than remembering, since they differ in detail between states and regulatory systems.

What the number assumes about the way it is flown

Every gradient in the take-off data is conditional on a specific way of flying. The conditions are rarely stated in the flight deck, and each one is a way for the real gradient to fall short of the computed one.

  • Speed. The gradient is scheduled at the climb speed the data assumes. Above and below it, the aeroplane climbs less well, and the loss below the target is steeper than the loss above. An engine failure flown ten knots slow is not a conservative choice.
  • Configuration. Take-off flap, gear up, and the bleed configuration the calculation assumed. Turning packs on, or selecting engine anti-ice that the computation did not include, moves the answer in the wrong direction — which is why the anti-ice state is an input to the take-off calculation and not a matter of comfort.
  • Thrust and time. Take-off thrust is time limited — typically five minutes, extended for many types when an engine has failed, with the exact figure in the flight manual. A second segment prolonged past that limit is outside the data.
  • Bank angle. A turn costs gradient. The operational rules limit bank in the early climb-out and require the flight path to account for the loss in any turn; the flight manual carries the correction. A turn flown at more bank than the procedure assumed does not merely tighten the ground track, it lowers the path.
  • Wind. The operating rules credit no more than half of any reported headwind and charge no less than one and a half times any reported tailwind when the net path is built. That asymmetry is deliberate, and it means a headwind never helps as much as it appears to.

The right habit. When the numbers are tight, ask which bound is active before asking what to change. Flap that relieves a field-length limit often worsens a climb limit, and the reverse; adding thrust helps a climb limit and does nothing for brake energy. The question “what is limiting” has a single answer on any given day, and every useful action follows from it.

Buying gradient, and what it costs

Where the climb limit binds, the standard remedy is to fly faster. The improved-climb or increased-V2 technique accelerates to a higher speed on the runway, spending field length to place the second segment nearer the speed for best gradient in that configuration. On a long runway with a low climb limit it can add real payload. On a short runway it is unavailable, because the length it spends was already the binding constraint.

Flap setting is the mirror image. More flap shortens the run and relieves a field-length limit, and costs climb gradient through drag; less flap does the opposite. That is why the flap schedule at a given aerodrome is a performance decision rather than a preference.

Reduced thrust deserves a specific note because two different mechanisms are routinely called by one name. Under the assumed-temperature method the calculation is performed as though the day were hotter, so every limit — including the climb limit — is evaluated at that assumed temperature, and the actual thrust delivered is never less than the computation assumed. A fixed derate is different in kind: it establishes a lower rated thrust, with its own minimum control speeds, and it is not a reduction from the full rating. Mixing the two up produces control-speed errors that no gradient calculation will reveal. The distinction is worth working through methodically rather than by rule of thumb; the material on the FL330 training platform takes the same route from certification requirement to line application.

What is not inside the increment

The last point is the one most worth carrying away. The 0.8 % covers dispersion around a demonstrated capability. It does not cover departures from the assumptions themselves, and a short list of those makes the boundary concrete.

A pilot in a cockpit watches an airplane taking off at dusk with landing gear down and engines emitting smoke.
Takeoff dynamics depend on more than just mass coefficients.

It does not cover contamination: airframe icing is not in the take-off data at all, and an aeroplane with ice on the wing has a different polar rather than a slightly degraded one. It does not cover engine deterioration beyond whatever the manufacturer states its scheduled data is based on — a manufacturer and continuing-airworthiness question rather than a certification one. It does not cover an obstacle absent from the database, which is not cleared by 35 ft but simply missing from a calculation that then means nothing. And it does not cover a second failure, a configuration error, or a departure flown on a track the data did not assume.

The limiting mass is therefore not a coefficient applied to a capability. It is the lowest solution of several independent inequalities, one of which carries a fixed gradient subtracted for statistical reasons that have nothing to do with prudence in flight. Read as a coefficient, it invites the idea that some of it can be spent. Read as what it is — the price of turning a handful of test flights into a claim about a fleet — it leaves one conclusion: the aeroplane at its limiting mass has exactly the performance the calculation says, and not one gradient point more.

Commentaires

Leave a Reply

Découvrez l’écosystème FL330

FL330, c’est toute l’aviation professionnelle — explorez nos autres espaces.