The NASA News item NASA’S First Asteroid Deflection Mission Enters Next Design Phase describes a test the kinetic impactor technique that could someday be used to adjust the orbit of an asteroid in order to avoid a collision with Earth in the future by making a small change at a time much earlier than the predicted event.
The article says:
The target for DART is an asteroid that will have a distant approach to Earth in October 2022, and then again in 2024. The asteroid is called Didymos -- Greek for “twin” -- because it’s an asteroid binary system that consists of two bodies: Didymos A, about one-half mile (780 meters) in size, and a smaller asteroid orbiting it called Didymos B, about 530 feet (160 meters) in size. DART would impact only the smaller of the two bodies, Didymos B.
The Didymos system has been closely studied since 2003. The primary body is a rocky S-type object, with composition similar to that of many asteroids. The composition of its small companion, Didymos B, is unknown, but the size is typical of asteroids that could potentially create regional effects should they impact Earth.
[...]“A binary asteroid is the perfect natural laboratory for this test,” said Tom Statler, program scientist for DART at NASA Headquarters. “The fact that Didymos B is in orbit around Didymos A makes it easier to see the results of the impact, and ensures that the experiment doesn’t change the orbit of the pair around the sun.”
Can someone explain the physics of this explanation to me more completely? Is it a small change in the orbital period of the pair that will be detected? Wouldn't this be predictable from just conservation of linear and angular momentum and Energy? Radar would only work during close Earth approach, so can the rotation be phased by optical means? What if the impact changes the rotation rate of the primary about its own axis and brings it out of tidal lock. Wouldn't this distort the brightness curve unpredictably?
Or is is the details of how the impact process itself proceeds, how efficiently momentum is transferred. In this case wouldn't the unknown composition of the B component make this more difficult to interpret? Will the spacecraft release an "observer" spacecraft to record the collision from a safe distance, say a few hours behind the impactor?
I don't understand what the observables are here, and what can be learned from them.
above: Delay-doppler data from Arecebo in 2003, found here. The small line above the primary is caused by the relative orbital motion of the secondary.