The centre-of-gravity envelope is drawn as a single closed shape, and that drawing quietly suggests that its two vertical edges are the same kind of thing — a permitted range with a wall at each end. They are not. The forward limit exists because the aeroplane must remain controllable in pitch; the aft limit exists because it must remain stable. Different physics, different failure modes, different warning signs. Treating them as one interval is how a legally loaded aircraft ends up unpleasant to fly, or worse, at exactly the point where the loadsheet said it would be fine.
One envelope, two unrelated problems
Longitudinal balance is a moment problem. The wing produces lift at the centre of pressure; the centre of gravity sits somewhere along the longitudinal axis; the horizontal tail produces whatever force is needed to close the moment equation. In a conventional configuration with the CG ahead of the wing’s centre of lift, the tail is pushing down. Move the CG forward and the required tail download grows. Move it aft and the download shrinks, may reverse, and the whole arrangement becomes progressively less willing to return to its trimmed attitude on its own.
That single sentence contains both limits, and it already shows they are not symmetrical. The forward limit is about how much force the tail can generate — a question of authority, and therefore of surface area, deflection, airspeed and thrust. The aft limit is about how strongly the aeroplane resists a disturbance — a question of stability, and therefore of the distance between the CG and the neutral point. One is a ceiling on a control force. The other is the disappearance of a self-correcting tendency. There is no reason for them to behave alike, and they do not.
The forward limit is about whether the tail can move the plane enough, while the aft limit is about whether the plane can stay stable on its own. They’re not the same, and treating them as such can lead to dangerous situations.
On a Boeing 737, a forward CG might require 15° of elevator deflection to rotate at Vr, while an aft CG might make the plane feel too light in pitch, leading to pilot-induced oscillations during turbulence.
The forward limit: can the tail still do its job?
At the forward edge of the envelope the aeroplane is at its most stable and at its least manoeuvrable. Every nose-up input requires more elevator, and there are three moments where the demand on the tail peaks.
- Rotation. At Vr the elevator has to break the nose free with the main gear still on the ground and the tail working in ground effect. A CG too far forward, or a stabiliser trim setting computed for the wrong CG, produces the classic heavy-nose rotation — increasing back pressure, a long ground roll past the expected point, and a rotation rate below what the take-off performance assumed.
- The landing flare. Low speed, full flap, high tail download requirement, and the need for a rapid pitch change close to the ground. This is generally the sizing case for the forward limit on transport aeroplanes, and it is why the certification demonstrations of controllability are flown at the forward limit in the landing configuration.
- The balked landing. A go-around adds a large thrust change to full flap and low speed, and the pitching moment that thrust produces depends on where the engines sit relative to the CG — nose-up under the wing, the other way when they are mounted high. Either way the trim is set for the approach, not for what the crew is about to ask of the aeroplane.
There is also a performance consequence that gets forgotten because it does not feel like a handling issue. A larger tail download must be carried by the wing in addition to the weight of the aircraft. The wing is therefore flying at a higher lift coefficient than the all-up weight alone would suggest, and the stall speed rises. Forward CG means a slightly higher stall speed, a slightly higher approach speed, more induced drag and marginally worse climb performance. It is a small effect on a large aeroplane and a very noticeable one on a light twin.
The aft limit: has the aeroplane stopped arguing back?
Static longitudinal stability is measured by the static margin — the distance between the CG and the neutral point, the point at which a change in angle of attack produces no net pitching moment. A large static margin means a strong restoring moment after a gust or an input, high stick force per g, and an aeroplane that returns to its trimmed speed on its own. As the CG moves aft, the static margin shrinks, and everything that depends on it shrinks with it.

What the crew actually notices, in order of appearance:
- Lighter and lighter pitch forces, then a tendency to over-control. A small input produces more attitude change than expected, the correction overshoots, and the aeroplane can be driven into a pilot-induced oscillation in pitch — most easily in turbulence, in the flare, or during manual handling after an autopilot disconnect.
- Weak or absent speed stability. The aeroplane no longer tries to return to trimmed speed, so speed excursions have to be flown out by hand rather than being damped by the airframe.
- Degraded stall behaviour. Recovery from a stall depends on reducing the angle of attack, which the nose dropping helps; an aft CG works against both. In an aeroplane with a T-tail or with a strong nose-up moment at high angle of attack, an aft CG is a contributor to a deep stall condition from which the elevator may have insufficient authority to recover.
Note the asymmetry in the warning signs. Forward CG announces itself early and honestly: everything feels heavy, and the crew adapts. Aft CG feels pleasant right up to the point where it does not. Light, responsive, low trim drag — the handling qualities of a well-balanced aeroplane, until a disturbance arrives and nothing restores the attitude.
| Forward limit | Aft limit | |
|---|---|---|
| What sets it | Elevator and stabiliser authority — the tail’s ability to produce the required download | Static margin — the distance from CG to neutral point |
| Sizing case | Landing flare at low speed, full flap; rotation at Vr; balked landing | Manoeuvring stability, stick force per g, stall recovery |
| What the crew feels | Heavy pitch forces, slow rotation, a flare that needs more than expected | Light, twitchy pitch; a tendency to over-control; no return to trimmed speed |
| Performance side effect | Higher stall and approach speeds, more trim drag, slightly reduced climb | Lower trim drag and reduced fuel burn — the reason operators want to be there |
| Failure mode if exceeded | Control authority runs out — the aeroplane will not rotate or will not flare | Stability runs out — the aeroplane will not recover on its own |
| Warning before the failure | Progressive and obvious | Deceptive; the handling improves before it degrades |
Why the envelope is a polygon and not a rectangle
The two limits do not move together with weight, and that is why the certified envelope is a shape rather than a pair of parallel lines. The forward limit is often weight-dependent, because at high weight the tail download required for a given CG position grows and the authority available at low speed becomes the binding constraint. The aft limit may also move with weight, and on some types it is additionally constrained by structural considerations at the extremes rather than by aerodynamics alone.
Two consequences follow. A CG expressed only in percentage of mean aerodynamic chord is meaningless without the weight that goes with it — the same %MAC can be inside the envelope at one weight and outside it at another. And the envelope has to hold for the whole flight, not just when the loadsheet is signed: fuel burn moves the CG in a direction set by the tank sequence, so a departure heading toward an edge is a departure that will leave it.
Why operators fly close to the aft limit anyway
Trim drag is real money. The tail download has to be carried by extra wing lift, and that costs induced drag on every hour of every sector. Moving the CG aft reduces the download and the drag with it. The gain per flight is small in percentage terms and enormous across a fleet and a year, which is why some long-range types carry a trim tank in the horizontal stabiliser and transfer fuel aft in cruise to hold an optimum CG, then forward again before descent.
That is why the aft region of the envelope is routinely occupied rather than reserved for unusual loadings. An aircraft loaded near the aft limit is inside its certification and behaving exactly as designed, and it will still be less forgiving in manual handling than the same aircraft loaded mid-envelope. That is not a defect and not a reason to refuse the loadsheet. It is a reason to brief it.
The trap. A loading error that keeps the total mass correct is invisible to every check that looks at mass. Swap the contents of a forward hold and an aft hold, seat a group in the rear of a cabin the loadsheet assumed was spread, or load a container in the wrong position: the take-off weight is right, the performance calculation is right, and only the moment is wrong. Mass and balance are two independent quantities, and the second one has far fewer natural cross-checks than the first.
Where the numbers come from, and where they drift
The CG on the loadsheet is a computed value, not a measured one. It is built from the dry operating mass and index of that specific airframe, plus the moment contributions of fuel, passengers, baggage and cargo at their assumed positions. Each of those inputs is an assumption with a tolerance attached, and the tolerances are not the same size.
- Dry operating mass and index come from the aircraft’s weighing record and the operator’s configuration data. They drift with modifications, cabin reconfigurations and accumulated items, which is why periodic re-weighing exists.
- Standard masses for passengers, hand baggage and hold baggage are values fixed by the authority, based on surveys, with distinct figures by category and by season in some frameworks. They are statistically valid for a full aeroplane and progressively less valid as the number of occupants falls — which is exactly why small-aeroplane operations often require actual masses instead. The applicable figures are set by the regulator and the operator’s approved scheme; they are periodically re-surveyed and they are not a number to quote from memory.
- Seating distribution is an assumption about where those masses sit. A cabin loaded by zone with a free-seating policy, a light load allowed to spread, or a group moved late in boarding can invalidate it without anyone noticing.
- Cargo position is the input with the largest moment arm and the least redundancy in checking. It is also the one handled furthest from the flight deck.
| Input | Typical failure | Effect on CG | Realistic defence |
|---|---|---|---|
| Dry operating mass and index | Stale after a cabin or equipment change | Small but systematic, on every flight of that airframe | Weighing programme and configuration control |
| Passenger masses | Standard masses applied to a small or atypical load | Larger scatter as occupancy falls | Actual masses where the approved scheme requires them |
| Seating distribution | Late group move, free seating on a light load | Can be substantial on a short fuselage | Zone control and a headcount by zone, not just a total |
| Hold loading | Containers or bulk in the wrong compartment | Large — long moment arms, both directions | Load instruction report signed against what is physically there |
| Fuel | Tank sequence assumption differs from the actual burn | Progressive movement during the flight | In-flight CG monitoring where the type provides it |
The last-minute change is where the two limits meet
A last-minute change is permitted within limits set by the operator’s approved procedure, and it is bounded precisely because it bypasses the recomputation. A handful of passengers added or removed changes the mass by an amount that is easy to reason about; it changes the moment by an amount that depends entirely on where they were sitting or where the bag went. That is the asymmetry the procedure exists to contain.
The same logic applies to the downstream use of the CG figure. On most modern transport types the CG feeds the stabiliser trim setting for take-off and contributes to the speed calculation. An erroneous CG therefore does not stay confined to the balance sheet — it propagates into a trim setting the crew will select and into speeds they will fly, and the resulting rotation behaviour is the first place the error becomes visible, at the least convenient moment. The cross-check that catches this is not arithmetic; it is a sanity check of the trim setting against what the crew expects for that loading, followed by a willingness to stop the take-off if the aeroplane does not behave the way the paperwork promised.
The right reflex. Read the loadsheet as two separate answers, not one. First: is the mass right and is it within structural and performance limits? Second, and independently: is this CG where I expect it to be for this loading, in which part of the envelope, and which way will it move as fuel burns? A crew that can say out loud “we are aft of mid-envelope today, expect light pitch forces in the flare” has already done the useful part of the briefing.
On the ground, the aft limit has a second life
Balance is not a flight-only concern. During loading and unloading, a CG behind the main gear will sit the aircraft on its tail, and the sequence in which holds are filled matters as much as the final state. It is a real ground-handling hazard on freighters and on any aircraft loaded out of its normal sequence. Tail stands, loading sequence rules and nose-gear ballast exist for this, and none of them appear on the loadsheet the crew reads.
An aft CG also unloads the nose gear, reducing steering authority on a contaminated surface and changing the weight transfer under braking — one more reason the aft limit is not the mirror image of the forward one.
What to take to the flight deck
The envelope is a compromise between two independent guarantees: in front, that the tail can still do what the crew asks of it at the worst combination of speed and configuration; behind, that the aeroplane will still return to where it was pointed when something disturbs it. Cross either edge and a different guarantee lapses.
In daily practice that is modest and specific: know which half of the envelope you are in before you rotate, know which way the CG will travel during the sector, treat a mass-neutral loading error as fully capable of hurting you as a mass error, and expect a legally aft-loaded aircraft to feel lighter in pitch than the one you flew yesterday. Type-specific behaviour near the edges is worth more than generic theory, and it gets argued out on forums.fl330.com.

Leave a Reply
You must be logged in to post a comment.