Centre of Gravity: Real Effects on Stall Speed, Trim Drag, and Pitch Authority

A pilot and flight engineer in a cockpit at dusk, adjusting the trim wheel with a forward center of gravity, preparing for takeoff.

The load sheet arrives a few minutes before pushback: takeoff mass unchanged, but the last-minute cargo went into the aft hold and the centre of gravity moved several percent of mean aerodynamic chord aft. The trim setting changes, so do the speeds if the operator takes a centre-of-gravity credit, and so does cruise fuel burn. Centre of gravity is not a box to tick between two limits: it is the single loading parameter that simultaneously fixes the lift the wing must produce, the drag that must be overcome to stay in balance, and the pitch authority that will remain available. All three come out of the same moment equation, and knowing which of the three is limiting on a given flight is worth more than reciting the two limits.

One single moment equation, three distinct consequences

Moment equation In steady flight, two conditions hold the aircraft. The first is vertical: the sum of the lift from the wing and the tailplane balances the weight. The second is a moment condition: around the centre of gravity, everything sums to zero. The position of the aerodynamic centre of the wing-fuselage combination is fixed by geometry once and for all; that of the centre of gravity changes with every load.

Lever arm When the centre of gravity is ahead of the aerodynamic centre, the wing’s lift acts behind it and produces a nose-down moment. The tailplane must then provide a download — a load directed downward — to balance it nose-up. The vertical condition therefore requires the wing to produce more than the weight: it carries the weight, and additionally compensates for what the tailplane pulls down. Moving the centre of gravity aft shortens the lever arm, reduces this download, and eventually reverses it: the tailplane then lifts upward, and the wing produces less than the weight.

Balance terms Centre of gravity is obviously not the only term in the balance. The wing-fuselage’s own pitching moment, the effect of high-lift devices, the position of the thrust line relative to the centre of gravity, the drag of extended landing gear all contribute. But it is the only term that varies from flight to flight for the same aircraft in the same configuration, and the only one the operator decides.

Three consequences From this, three consequences are worth not confusing. The wing must produce more: it will reach its maximum lift at a higher speed. It produces more, and the tailplane produces something too: both pay in induced drag. Finally, the surface that supplies this balancing load does so with a deflection: that deflection is consumed and is no longer available for manoeuvring.

The centre of gravity affects how much lift the wing needs to produce. If it’s too far forward, the wing has to work harder, increasing stall speed and drag. If it’s too far back, the tailplane has less control authority, which can make the plane harder to handle.

On a Boeing 737, moving the centre of gravity from the forward limit to the aft limit can reduce stall speed by about 5 knots and reduce trim drag by 1-2% of total drag, saving hundreds of kilograms of fuel on a long flight.

Moment balance at forward centre of gravity and aft centre of gravityTwo side views of the same aircraft. At the top, the centre of gravity is ahead of the wing’s aerodynamic centre: the tailplane produces a download and the wing’s lift arrow exceeds that of the weight. At the bottom, the centre of gravity is behind the aerodynamic centre: the tailplane load reverses into a small upward lift and the wing’s lift becomes lower than the weight.forward CGliftweightCGaero centredownloadaft CGweightCGreduced lifttailplane lifting
Everything starts with the lever arm between the centre of gravity and the aerodynamic centre. At forward CG, the tailplane pulls down and the wing carries more than the weight: it is the same elongated arrow that explains both the higher stall speed and the trim drag. Moving the CG aft shortens the arm, then reverses the tailplane load.

Why a forward centre of gravity raises the stall speed

Stall speed The wing stalls at its angle of attack, not at a speed, and centre of gravity changes nothing about that. What it changes is the weight the wing can support when it reaches its maximum lift. Since at a forward CG it must produce the weight plus the tailplane download, the speed at which it reaches its limit rises. Put differently: the trimmed maximum lift coefficient of the whole aircraft is lower than that of the isolated wing, and the gap grows with the balancing load.

Flaps extended The effect is more pronounced in landing configuration than clean. Extending trailing-edge devices generally adds a nose-down moment that the tailplane must take up, on top of the one coming from the lever arm: the two contributions add together, so the sensitivity of stall speed to centre of gravity is greater with flaps extended. That is precisely the configuration where the margin is shortest.

Pitch authority A second, often forgotten, mechanism overlaps this one. On some types, the demonstrable minimum speed is not set by the wing’s maximum lift but by pitch authority: at a forward CG, the deflection needed to hold the attitude reaches its stop, or the control force required exceeds what certification tolerates, before the wing stalls. The minimum speed is then a control limit disguised as a stall, and it degrades with centre of gravity far faster than a simple supported-load effect would suggest.

What certification gives you back, and what it does not

Reference speed The reference stall speed is determined at the most unfavourable centre of gravity of each configuration, and the scheduled speeds are derived from it by regulatory coefficients: on the order of 1.13 times this reference speed for takeoff safety speed, on the order of 1.23 times for approach speed in landing configuration — the exact values, their scope of application and the special cases being those of the certification text applicable to the type. If the flight manual publishes only one set of speeds, established at that unfavourable centre of gravity, any flight loaded further aft has a spare margin it does not claim. That is free conservatism, and it is silent.

Centre-of-gravity credit As soon as the operator uses centre-of-gravity-dependent data — takeoff mass increased at aft CG, recalculated reference speed — this margin is spent. The calculation becomes exact, hence fragile: it assumes the actual centre of gravity is indeed the one on the sheet. The centre-of-gravity credit is not a performance gain, it is a transfer of margin from certification conservatism to loading accuracy.

Speed monitoring The consequence for speed monitoring: a stall warning driven by angle of attack triggers at the same angle regardless of centre of gravity, but the speed at which that angle is reached shifts. A low-speed cue built on a fixed reference speed does not shift. The two indications therefore do not tell the same story at a forward centre of gravity, and it is the angle of attack that tells the truth.

Stall speed and trim drag as a function of centre-of-gravity positionTwo quantities in relative units plotted against centre of gravity, from the forward limit on the left to the aft limit on the right. The solid line, reference stall speed, decreases steadily as the centre of gravity moves aft. The dashed line, trim drag, decreases faster then passes through a minimum marked with a dot, located slightly aft of the centre of gravity that zeroes the tailplane load, before rising again up to the aft limit.relative valueforward limitaft limitstall speedzero tailplane loadtrim drag (dashed)drag optimumcentre of gravity moving aft
The two curves do not fall at the same rate, and one of them rises again. Stall speed follows the balancing load roughly proportionally; drag follows its square, hence a penalty that worsens near the forward limit. The drag minimum falls a little aft of the zero-tailplane-load point, and not at the aft limit, which is governed by stability.

Trim drag, the fuel-burn item that appears on no gauge

Three contributions Three contributions add up. The wing produces more than the weight: its induced drag, which varies as the square of lift, increases. The tailplane produces a load, downward or upward, it makes no difference: it pays its own induced drag, which varies as the square of that load. And the deflected surface, control surface or all-moving tailplane, modifies the pressure distribution and the flow at the junction with the fuselage: an increment of form drag adds to this.

Non-linear The dependence is not linear. The wing term grows roughly proportionally to the balancing load, the tailplane term as its square: the penalty worsens faster than the change in centre of gravity. The same shift of a few percent of MAC costs more near the forward limit than in the middle of the envelope. This is why chasing an aft centre of gravity mostly rewards operations that started from far forward.

Drag optimum The optimum, contrary to a common belief, does not sit exactly where the tailplane load is zero. For a two-lifting-surface system, total induced drag is minimal when lift is shared between them rather than concentrated on one alone, which implies a small positive tailplane lift, not a zero load. The drag optimum is therefore slightly aft of the zero-load point, and there is no reason for it to coincide with the aft limit of the envelope: that limit is set by stability, not by fuel burn.

Mach number Nor is there a single optimum. In transonic cruise, the aerodynamic centre moves aft as Mach number increases: for a given centre of gravity, the balancing load changes with Mach and altitude. The optimum is a surface, not a number, which is what makes the subject hard to reduce to a single loading rule.

Fuel burn Order of magnitude: on a transport aircraft, sweeping the usable envelope from one edge to the other shifts cruise fuel burn by roughly a few tenths of a percent, the value depending on the type, the tailplane geometry and the cruise regime. Negligible on a short sector where turnaround time governs the economics; significant on a long-haul flight, where the same percentage amounts to hundreds of kilograms per rotation, and considerably more when scaled to a fleet over a season. This difference in scale explains why some long-haul aircraft carry a trim tank in the tailplane and actively transfer fuel aft in cruise, while other operations simply avoid loading the forward end without reason.

What a shift of the centre of gravity changes, at constant mass and configuration
Quantity Forward CG Aft CG
Tailplane balancing load Large download Small download, or even positive lift
Lift the wing must produce Above weight Close to weight, or below
Reference stall speed Higher Lower
Scheduled speeds derived from it Higher, distances increased Lower: this is the centre-of-gravity credit
Trim drag Maximum across the envelope Close to the minimum, without exactly reaching it
Static margin, stick force per g Comfortable Reduced, down to the regulatory minimum
Remaining nose-up authority Consumed by trimming Largely available
Natural nose-down moment at the stall Clear-cut Attenuated, recovery more demanding
Load on the nose landing gear High, often sizing Low, degraded ground steering
Case it governs in certification Pitch authority, minimum speeds Stability, stall, ground behaviour

Pitch authority: what an aft centre of gravity takes away without saying so

Static margin Static margin is the distance between the centre of gravity and the neutral point, the centre of gravity at which the pitching moment ceases to depend on lift. The smaller it is, the smaller the restoring moment produced by a change in angle of attack: the aircraft holds its speed less firmly, deviations settle in rather than being resorbed, and manually flying an approach demands more attention well before any limit is reached.

Manoeuvre margin Manoeuvre margin is a different quantity, larger than static margin because it includes the damping contributed by pitch rate. It is what governs stick force per unit of load factor. This sensation — an aircraft that picks up load factor without one really having pressed — signals a very aft centre of gravity before any display does. Certification imposes a minimum on this quantity: the aft limit of the envelope is often its direct expression.

Short-period mode The short-period mode also shifts: its frequency drops as the centre of gravity moves aft. The aircraft responds more briskly and settles down less well; in turbulent air, the crew compensates without necessarily identifying the cause.

Stall behaviour At the stall, the small margin is paid for twice. The natural nose-down moment that helps reduce angle of attack is attenuated, and the control surface must do a larger share of the work — at the very moment the tailplane is operating in degraded airflow. On architectures where the tailplane sits in the extension of the wing’s wake at high angle of attack, this is the territory of deep stall.

Takeoff rotation The forward limit consumes authority the other way, and two cases fix it: rotation at takeoff, where a sufficient nose-up moment must be produced at a still-low speed, and the flare with reduced thrust and flaps extended, where the same requirement arises with decreasing dynamic pressure. Add the go-around: with engines mounted under the wing, hence below the centre of gravity, applying thrust produces a nose-up moment. That moment does not depend on centre of gravity; what does depend on it is how vigorously the aircraft responds — at an aft centre of gravity, less opposed by stability, it pitches up more sharply.

Fly-by-wire There remains the case of fly-by-wire airframes. Some deliberately exploit a centre of gravity that natural stability alone would not cover, the control law restoring apparent stability; the trim-drag gain is then real and permanent. The operational consequence lies elsewhere: the centre of gravity acceptable in normal law is not necessarily acceptable in a degraded law. Hence the centre-of-gravity restrictions attached to certain flight control failures, and the forward fuel transfers that accompany them. Centre of gravity then stops being a problem solved on the ground.

What actually sets the two edges of the envelope

Envelope edges A mass-and-balance envelope is not a rectangle, and its edges do not all belong to the same discipline. Each portion of the boundary is the trace of a different criterion: knowing which one you are approaching tells you what will degrade first.

Criteria that bound the centre-of-gravity envelope, and the associated warning sign
Criterion Limit concerned Mechanism What is observed first
Nose-up authority at rotation Forward Insufficient available deflection at low dynamic pressure Heavy rotation, extended takeoff run
Authority at the flare, reduced thrust Forward Same requirement at decreasing dynamic pressure Late flare, hard touchdown
Minimum speeds and scheduled speeds Forward Tailplane download compensated by the wing Increased takeoff and landing distances
Nose gear structural load Forward Increased load transfer, worsened under braking Nothing in flight: a calculated limitation
Minimum static stability and stick force per g Aft Insufficient manoeuvre margin Sensitive aircraft, involuntary load factor
Stall behaviour and recovery Aft Attenuated natural nose-down moment Deeper stall, slower recovery
Minimum load on the nose wheel Aft Insufficient grip and ground steering Degraded directional control, especially on contaminated runways
Tail tip-over on the ground Aft Resultant behind the main gear Event during loading or refuelling
Controllability in degraded control law Aft Artificial stability lost Published restriction, fuel transfer

Two remarks Two remarks are called for. The limits depend on mass, because the criteria themselves depend on it: pitch authority is assessed at a given mass and dynamic pressure, nose gear load grows with weight. And the envelope used in operation is generally narrower than the certified envelope: it absorbs in-flight shift, the dispersion of standard masses and last-minute changes. Confusing the two leads to believing available a margin that is already allocated.

From the load sheet to the trim setting

Centre-of-gravity calculation The centre-of-gravity calculation is a sum of moments reduced to an index system: each station receives an influence coefficient, and the addition remains checkable by hand. This elegance has a cost: the index hides the physics. The same index change does not correspond to the same number of kilograms depending on the station: a few hundred kilograms in the aft hold move the centre of gravity far more than a tonne near the reference point. The loading errors that matter are not the heaviest ones.

A military aircrew member reviews a load sheet while another loads cargo.
Load distribution affects aircraft performance and handling.

Standard masses Standard masses are a statistical tool: they correctly describe a full cabin, less well a lightly loaded flight, where individual dispersion no longer averages out. Add cabin distribution: free seat allocation on a light flight can produce a distribution the calculation did not assume, if the operator imposes no zoning.

The trap. A centre-of-gravity error never shows up as a centre-of-gravity error: it shows up as a trim setting that does not match the actual loading. Centre of gravity actually further aft than calculated, hence trim too nose-up: rotation runs away on its own, brisk, with a risk of tail strike. Centre of gravity actually further forward, insufficient trim: rotation is heavy, long, and the runway keeps rolling by. In both cases, the first clue arrives at rotation, when there is no longer an option. The useful check is done on the ground, on the consistency between the picture of the loading and the printed figure.

The right reflex. Read the centre-of-gravity position within the envelope before reading the trim figure, not the other way round. Locate the point — rather forward, rather aft, how far from the edge — then check that the announced setting goes in that direction and that its order of magnitude is what is expected for this kind of loading. A figure that is correct within a wrongly pictured envelope is still wrong; a figure consistent with a mental picture of the loading is already half-verified.

Last-minute changes Last-minute changes are bounded by two tolerances, and the usual oversight concerns the second. The mass tolerance is remembered without effort; the one on centre of gravity, expressed as index or as percent of chord, much less so. Yet a modest load placed far from the reference point can respect the first and breach the second. The test is not how many kilograms, but how many kilograms and at which station.

The trap. Taking the aft centre-of-gravity credit without checking that the assumed distribution is the one that will actually be achieved. Takeoff mass increase is calculated for an announced centre of gravity; actual cabin fill, the dispersion of standard masses and a hold swapped at the last minute all work in the direction of a deviation. And the credit has already spent the conservatism that precisely absorbed that deviation: the calculation is no longer protected by certification margin, it is protected by the quality of the loading.

In-flight movement Centre of gravity moves in flight. Fuel consumption empties the tanks in a set order, an order chosen as much for wing bending as for centre of gravity; on airframes fitted with a trim tank, it also moves deliberately. The practical consequence is that the landing centre of gravity is not the takeoff one, and it is the landing centre of gravity that governs authority at the flare. A flight that takes off near the aft limit and arrives near the forward limit has met both constraints within the same flight without ever exceeding either.

Shifting cargo Then there is shifting cargo. It is the only mechanism that takes the centre of gravity outside the envelope in flight without anything signalling it: no alarm, no message, only a trim demand that has no reason to exist. A trim setting that drifts nose-up during climb, at known mass and configuration, is loading information before it is flight control information. Restraint is the only lock, and it is checked on the ground.

Movement of the centre of gravity during the flight, between the two limitsPosition of the centre of gravity through the phases, from loading to landing, with aft toward the top. The light band marks the usable envelope between the forward and aft limits. Both traces start from the same loading and move aft to the same position at takeoff. The solid line, with fuel transfer aft in cruise, stays close to the aft limit then returns forward during descent. The dashed line, without transfer, drifts steadily forward as fuel is consumed. Both arrive at landing near the forward limit.CGaftaft limitforward limittakeoff loaded aftsolid line: fuel transfer aft in cruisedashed line: consumption only, CG returns forwardloadingtakeoffcruiselanding
The takeoff centre of gravity is not the flare centre of gravity. Fuel burn brings the point back forward, and it is the landing centre of gravity that governs authority at the moment it is needed. A flight can thus approach both constraints successively without ever exceeding either.

What to open in your own documentation

Five points Five points are worth examining away from the urgency of the moment. At what centre of gravity are your type’s published speeds established, and does your performance tool take a centre-of-gravity credit or not: the answer changes the nature of the margin you have. Is the envelope you use the certified one or a reduced envelope, and exactly what does that reduction absorb. The last-minute-change tolerance expressed in centre of gravity, not just in mass. What the centre-of-gravity restrictions attached to degraded control laws say, and what action they demand in flight. Finally, which indication triggers a doubt about the loading in your operation, and who decides when the doubt is resolved on paper but not on the aircraft.

Grey area A grey area remains, a structural one. Centre of gravity is calculated, never measured. No transport aircraft weighs its centre of gravity in flight: the only available witness is the trim required to hold level flight at known speed and configuration, and this indicator has no published tolerance. The devices that display an in-flight centre of gravity derive it from the same aerodynamic models and the same load sheet; they do not constitute an independent measurement. Reconciling a suspect trim setting with a calculation that comes out right therefore remains an exercise in judgement, at the interface between crew, loading and technical control — and these practices are exchanged more than they are published, on the forum as much as in the crew room.

Type rating For anyone preparing a type rating, one question is enough to separate the candidates: why does a forward centre of gravity increase stall speed when the wing always stalls at the same angle of attack? The answer through the moment equation — the wing must produce the weight plus the tailplane download — is worth more than reciting two limits, because it gives in the same move the origin of trim drag and that of the loss of authority. The rest is maintained through periodic training, on the simulator and on real loading cases.

Applicable text The applicable text, the type’s certification specifications, the flight manual and the operator’s approved documentation prevail over this article, which sets out a mechanism and does not stand in for a loading procedure.

Questions fréquentes

Why does a forward centre of gravity increase stall speed if the wing stalls at a constant angle of attack?

Because it is not the stall angle of attack that changes, but the weight the wing can support at that angle. At a forward centre of gravity, the tailplane produces a download to balance the wing’s nose-down moment, so the wing must produce the weight increased by that download. It reaches its maximum lift at a higher speed: the stall speed rises while the stall angle is unchanged. The effect is more pronounced with flaps extended, because the nose-down moment from the trailing-edge devices adds to that of the lever arm.

Does an aft centre of gravity really save fuel?

Yes, through a reduction in trim drag: less tailplane download means less lift the wing has to produce, hence less induced drag on both surfaces, and a smaller trim deflection. The order of magnitude remains modest: typically a few tenths of a percent of cruise fuel burn between the two edges of the usable envelope, depending on the type and the flight regime. It is not noticeable on a short sector, but it amounts to hundreds of kilograms per rotation on a long-haul flight, and considerably more scaled to a fleet, which is what justifies tailplane trim tanks and active cruise fuel transfers.

Is the drag optimum located at the aft limit of the envelope?

No, for two distinct reasons. First, the total induced drag of a two-lifting-surface system is minimal when the lift is shared between them, which implies a small positive tailplane lift rather than a zero load. Second, the aft limit is not set by fuel burn but by stability, stick force per g, stall behaviour and ground handling: nothing guarantees it falls near the aerodynamic optimum. The optimum also depends on Mach number and altitude, since the aerodynamic centre moves aft transonically.

Should centre-of-gravity-dependent performance data be used?

They are legitimate once approved, but they change the nature of the margin. Speeds published at a single centre of gravity are established at the most unfavourable one: any flight loaded further aft benefits from free conservatism. Taking the centre-of-gravity credit means spending that conservatism, and the calculation then assumes the actual centre of gravity is indeed the one on the sheet. The decision is therefore as much a question of the reliability of the loading process as of performance.

Is the landing centre of gravity the same as the takeoff one?

Rarely. Fuel consumption empties the tanks in a set order, chosen as much for wing bending as for centre of gravity, and some airframes additionally deliberately move fuel aft in cruise. A flight can therefore take off near the aft limit and arrive near the forward limit without ever leaving the envelope. It is the landing centre of gravity that governs the authority available at the flare, with reduced thrust and flaps extended, and it deserves to be checked in its own right.

How can a load-sheet error be detected in flight?

The only witness genuinely available is the trim required to hold level flight at known speed, mass and configuration: a trim setting that departs from the expected value, or that drifts nose-up during climb, is loading information before it is flight-control information. No tolerance is published for this indicator, which leaves room for judgement rather than a criterion. An unrestrained load that shifts is exactly the case where no alarm triggers, which is why so much weight is placed on checking ground restraint.

Commentaires

Leave a Reply

Découvrez l’écosystème FL330

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