How does the A350 variable camber system decrease drag?

Question

Just curious how the A350 reduces drag/fuel burn by extending the flaps ever so slightly during cruise. Everything I've read about extending flaps says that extending them pushes the centre of pressure rearwards. This causes a pitching down moment, which means you need to increase the AoA, but this affects drag and fuel burn negatively (a pitch up command increases the down force from the tailplane which has the result of increasing the effective weight of the plane, which requires an even higher AoA to compensate, which also increases the induced drag).

Here's how it works as per the (A350 Flight Deck and Systems Briefing for Pilots)

Differential Flap Setting and Variable Camber

The Differential Flap Setting and Variable Camber enable to optimize the loads and drag on the wings.
Small flaps deflections (4° maximum) either symmetrically or asymmetrically, enable to automatically:

• Optimize the wing camber to reduce wing loads and drag
• Perform an optimized Lateral Trim function.

Answer 1

Henning Strüber, one of the Airbus engineers behind this system, has written a paper on it.

In cruise:

This can be applied in early cruise phases to shift the center of lift more inboard and by that reducing the wing root bending moment, which can be transferred into a structural weight saving.

A plane that can be built lighter will have lower drag [for the same payload].

For a heavy and/or a hot and high takeoff:

In high-lift configuration a more outboard loaded lift distribution can be achieved to reduce induced drag during take-off.

For how that works, see here. The answer there by @PeterKämpf confirms that outboard loading requires a heavy wing relative to the whole mass (scaling laws work for an insect, but not for an albatross or an airplane). So those two regimes may appear to be contradictory: if the cruise system allows a lighter wing, then how can that lighter wing achieve more outboard lift for a heavy takeoff.

Accounting for the gust loads in cruise is the key here, which are smaller for takeoffs. (Thanks to @PeterKämpf for this insight; see comment posted below.)

Answer 2

Just complementing the answer from @ares which is quite good and refers to the main effect. I would like to refer to another "secondary" effect that implies also drag reduction.

When designing an airplane the structure is designed taking into account several factors, one of them, is the maximum load that the airplane can be exposed to.

Airbus has designed a system that during flight optimises the loads over the wing. Let's say, for example, that, without the variable cambering system we have a determined maximum load (let's say A). Using the variable cambering system the airplane is capable of reducing the load A (maybe increasing drag) to B (being B < A).

So, when designing the structure, assuming that the variable cambering system will be used, the airplane will use B as design point and not A. As B < A the size and weight of the structure with the variable cambering system will be lighter.

A lighter structure will imply less lift force needed and so less drag produced to achieve such lift. So, from a design optimization point of view, the variable cambering system is, essentially, providing new design variables which will allow a better opitmization of the airplane, reducing the drag.

Essentially, @ares has properly described the "active" mechanism, and I have described a "passive" one.

Answer 3

Briefly speaking, there at two types of drag: profile drag and induced drag. Profile drag can be said to be composed of viscous drag and shock-wave drag (these two often interact). Induced drag is a result of the local effective velocities on the wing due to the -vortex- circulation. I don't know what is your background and thus I won't go more in depth. I will just mention that, if you want to simplify things, profile drag can be seen as the drag that a 2D airfoil would have, while induced drag is a measure of the efficiency of a wing geometry. In reality, these two types of drag may strongly interact when the aspect ratio of wings is small.

For example, it is well known that elliptical spanload distributions are optimal (when induced drag is the objective). This 'elliptical spanload' can be reproduced if you design a wing of i) constant chord with elliptical distribution of twist, ii) elliptical distribution of chord with constant twist, iii) a combination of twist and chord that gives an elliptical distribution.

What AIRBUS says, is that we can vary the chord in-flight to optimize spanload and root-bending moment. What they actually do, in simple terms, is to optimize for induced drag and possibly control, i.e. stability and flight controls by changing the chord distribution. Now, if you change camber, you will also optimize for profile drag (because camber is a property of the 2D airfoil) and twist (in the sense of aerodynamic twist). For commercial jets, flying in transonic conditions, this means that the camber is modified to reduce shock-wave drag loss.

Clarification: Aerodynamic twist refers to changes in camber along the span. It is often more efficient to change the airfoil shape along the span instead of using the same airfoil at a different angle.

If you want more details, please let me know.