Space Forum / Space Flight / November 2003

Existing plasma thrusters use mechanical gimbals, because they get the
job done in a very well understood manner.  Magnetic thrust vectoring
is not out of the question, and is appealing on several grounds - not
the least of which is that anything that gets rid of unnecessary moving
parts is a Good Thing.

However, plasma dynamics is a Wierd Science, and it's hard to predict
exactly how a plasma exhaust will respond to some arbitrary, asymmetric
magnetic field.  Also hard to test on the ground, because the walls of
your test chamber interfere with both the plasma exhaust and the applied
magnetic field.

So this is a long-term thing, for people with a lot of experience operating
plasma thrusters in space and trying to optimize Nth-generation designs.
The Russians have been flying plasma thrusters for thirty years, and are
still using mechanical gimbals.

Many of them do.  Other methods of thrust vectoring are sometimes used,
e.g. fluid injection into the side of the nozzle.

As John has already noted, this is feasible in principle but hasn't yet
been used in practice.  Similarly, you can vector the thrust of an ion
engine by shifting the grids in relation to each other -- this was
demonstrated nearly forty years ago -- but operational ion-thruster
systems still do vectoring with mechanical gimbals.

Non-mechanical thrust vectoring is the sort of sophisticated refinement
that would start showing up if these systems were being aggressively
developed and improved.  By and large, they aren't.

I didn't see any mention of this on the VASIMR site. It's a neat idea,
but I'm sure a mechanical method would win because:
1. It would be cheaper, given the low thrusts involved
2. There would be no cosine losses

Though thinking about it, you'd probably just turn the whole ship.
That would work at the low forces and accelerations typical of
electric propulsion.