Warning: For Techies Only. Part 3 will return to a presentation suitable for the general reader. Diagram included, as supplied by the beautiful Russian spy “Natasha.”

Let’s reverse engineer the Avangard, using only the information (or disinformation) that is publicly available, which is:

• The length is stated to be 5.4 meters.
• It carries a warhead of at least 2 megatons.
• The body structure is made of carbon fiber plastic with ablative qualities similar to the U.S. “carbon-carbon.”
• It can make high-G maneuvers.
• It travels at suborbital velocity, below the Kármán line.
• It is powered by a solid-fuel scramjet.
• It blows up at the end.
• It was done on the cheap.

From this we can deduce:

• It is single-use only, so it’s tolerable if the airframe literally changes shape in the few minutes of flight.
• It is a “cold-body” vehicle. The carbon structural material ablates, shedding off the body, carrying heat away.  In comparison, the Boeing X-51 Waverider is a “hot-body” vehicle, where substantial heat is conducted to parts of the airframe, though critical parts are made of carbon-carbon.
• The weight of the warhead “physics package” is 1 metric ton or more.
• It has a dense concentration of mass, in the physics package.
• Since the final version travels at suborbital velocity, the requirement of airframe generated lift is minimal.
• Since the speed of the vehicle creates an ionization envelope, guidance must be entirely inertial. The thing is blind as a bat.

This is not a spy job. In what follows, the style should not imply that covert information was used. It’s a derivation, but it’s based on the facts that the Russians are not 10 feet tall, and they think like us, so the solution must be in plain view.

Although some loss of accuracy, compared to a ballistic vehicle, is allowed, it has to be controllable enough for the inertial guidance system to have something to work with. Hence these requirements:

• Control, without skinny control surfaces like fins, that would burn off or change shape in flight.
• The scramjet must have throttling capability. (Challenging, but there is an interesting solution.)

Since it was done on the cheap, yet achieving much higher speeds than the X-51, the designers exploited existing technologies and principles, of which there are a lot. While the U.S. effort attempts to create a new performance regime with hydrodynamic simulation on supercomputers, the Avangard takes another approach. Hypersonic “flight”, both mostly-aerodynamic and almost-ballistic, dates to the early 1960’s. Before computational hydrodynamics, there was intuition, which can be used here to reduce the size of the problem to apparently solvable proportions.

We know the Avangard designers used hydrodynamic simulation. This has come out with the jailing of some of the designers, for the sin of sharing their success with some of their former European collaborators. So how does their approach differ from the U.S., and possibly the Chinese? They chose an approach based on existing hypersonic vehicles, perturbing the designs to arrive at a new but nearby “regime” of flight.

“Perturb” means to start with something familiar, or comparatively simple, by small  steps making it different, until you arrive at the design you want. Here, the Russians had three starting points of great comfort:

• The elongated cone of a MIRV reentry vehicle, which has been studied ad nauseum.
• Nonlinear control theory based on the Lyapunov function, which has been applied by everybody and his brother to the above, to control the attitude of a MIRV with reaction jets.
• The prior use, way back, of moveable weights to control a blunt-cone reentry vehicle. The U.S. MARV design dates to at least 1963.

There were these new elements:

• Flat surface to generate compression lift.
• Scramjet inlet with some way to throttle.
• Control of attack angle by moving the heavy physics package within the vehicle body,  with further trim by reaction jets.
• High-g maneuver capability without loss of stability.
• A sophisticated control system that can work with control actuators that are highly interdependent in effect. The control actuators are simple, the control problem is complex.

A nonlinear system can be made conditionally stable. Some designs are more stable than others. Like almost all high performance aircraft, if a Boeing Waverider is pushed out of its stability envelope, it enters a regime from which the control system cannot recover. Yet a beach ball, or sphere of any kind, doesn’t have this problem. The more maneuverable the craft is, the more opportunities for loss of control.

This implies Avangard is compact, massive, and has only one aerodynamic control surface, which can be positioned only by changing the attitude of the entire structure.  It is immune to inertia coupling, and almost immune to mode couplings of any kind. This is possible because it is only marginally an aerodynamic vehicle.  Most of the “lift” results from the trajectory. Since “straight” for the Avangard is actually a curve with the radius of the earth, it is acted upon by the “fictitious” centrifugal force. Original tests at much lower speeds were done with deadweight much less than the 1000kg of an actual physics package.

The scramjet outlet, on the cold rear surface, was barely a problem, though inertia coupling had to be considered. This diagram was given to me by loving “Natasha” one fine spring day in Gorky Park. She asked me, “Would you like to feed the birds?”, and handed me a bag of bird seed containing the diagram. I’ll never forget her foot-long cigarette holder, or the clip-clop of her stiletto heels. Diagrams 1, 2, and 3 illustrate the perturbation:

1: The MIRV cone is cut by a plane, creating an initially small lifting surface. It is enlarged by iteration.

2.: The scramjet inlet duct is introduced the same way.

• Attack angle theta is introduced.
• The scramjet is throttled by changing the attack angle, which varies the area of the inlet duct exposed to the air stream. This is not independent control, but it’s enough.
• By shifting the position of the physics package “P”, L1 and L2 vary, changing the pitch (attack angle.)
• Reaction jets on the cold rear surface provide 3-degrees of freedom, rotating the vehicle body, and adding to pitch control.
• Location of “P” far forward, L1 >> L2, increases the force moment of the reaction jets.
• Concentration of mass at the forward end reduces the potential of inertial coupling.

3:  In further iterations, the  lifting surface is enlarged.  Shown also is increased theta for sharp maneuver and increased scramjet inlet duct area.

Avangard is not an airplane. It is not a general purpose solution, but there is elegance in simplicity. The purpose of deterrence is well served, at minimal cost. We should consider this without shame. The Russians realized that because strategic nuclear deterrent has no political purpose, the level of refinement implied by current U.S. hypersonic research is unnecessary.

You’re probably wondering what happened to Natasha. She exfiltrated Russia by a route I cannot disclose, and was resettled in the U.S. under a protection program. She made an indie film about her escape, which you can see here.