Brian Gaff wrote:
That is what I was thinking, hence my thought on the deceleration motors,
but this is weight taken away from payload, I suppose.

Whichever way you look at it, you cannot just assume that the two bodies
coasting for a short time will give the desired separation. If you coast
too long, in order to protect your first stage from plume damage, you run
the risk of sloshing in the upper stage suddenly released from thrust with
no continuation within the time you have before it goes float about in the
tank.
Brian


There are two ways {at least) around that problem.
1.) Use ullage rocket engines on the upper stage to seat the propellants
before engine ignition.
2.) Use some sort of a membrane inside the tank that separates the
propellant from the tank proper and which allows gas to fill the area
between the tank wall and the membrane - squeezing the propellant toward
the feed pipe rather than letting it float around in the tank.
SpaceX's report still doesn't explain the roll oscillations on the video
during ascent.

Pat