I don't know much at all about ion drives. But in (e.g.) an x-ray tube, electrons are accelerated to high speeds by arranging for them to begin in a location with non-zero potential (a hot negative filament) so that they will gain speed as they go away. So at a very vague guess, I'd expect that ion drives arrange to ionize the atoms in a region at a high potential, after which the positive ions quickly accelerate toward some negatively charged but porous "target" and then (mostly) pass through and continue at constant speed.
My broad take on the distinction (assuming I have the right general idea about ion drives!) would be that if you're the one providing the material to be ionized, you get a lot of control over its initial conditions. But if you're planning to use matter that's externally available, your control is a lot more limited. In the schematic that I've seen for this dipole drive, the ions begin far away (so, at zero potential) and end far away (again at V=0), so I don't see a way to avoid that.
> In the schematic that I've seen for this dipole drive, the ions begin far away (so, at zero potential) and end far away (again at V=0), so I don't see a way to avoid that.
The closest I can see to the dipole drive is a gridded ion truster:
It, like the ion drive, consists of two grids with an electric potential between them. Ions are emitted into the first grid, get accelerated and propelled out the back.
Linear accelerators uses the same mechanism to accelerate particles though using alternating current to accelerate the particles through multiple stages instead of just one.
Of course in both of these cases they contain particle emitters "built in".
My physics background is severely lacking, but my understanding of your objection to the dipole drive is that it'd effectively neutralize any trust gained. And that would be the case for ion thrusters if you just do as described above as well, as the exhaust would be charged and attracted to the vehicle , neutralizing the thrust. To gain net thrust for the vehicle you therefore additionally need to make the ion bean electrically neutral by using a cathode to emit electrons into the beam.
So in a gridded ion thruster you have a reservoir of propellant that is then ionized before being passed through the first grid, and a cathode to emit electrons into the stream after the second grid.
My lack of physics background means it is not clear to me what happens if you replace the propellant reservoir with a scoop, but it sounds like what Zubrin argues is that the reflection of the electrons exerts so little force on the scoop that it can be largely ignored relative to the thrust provided by the protons that gets accelerated through the grids, and that the reflected electrons will be sufficient to neutralize the overall charge created.
It's not clear to me why this could not be possible given that we know that ion thrusters work. It seems to me that the only difference is that you effectively have an outside medium (the plasma) you transfer energy to. Conceptually not much different from how a turbofan works, where you're also relying on the amount of additional energy you transfer to the medium to overcome the drag of the portion of the medium that hits your vehicle. I realise of course that the types of forces involved are entirely different, so it might be an idiotic comparison, but the point being that scoop type propulsion systems are well known, and while we know that the medium will eventually lose energy some way or the other and return to the same potential, we can still achieve propulsion by transferring energy to it locally. We know we can do that by ionizing a propellant carried with the vehicle.
I don't grasp what is conceptually different with transferring energy to increase the potential of particles that happen to enter the engine/screens.
It would seem to me that the two big questions is whether he is right that the reflected electrons would be enough to neutralize the overall charge sufficiently, but neutralizing the charge of the exhaust is a known problem from ion thrusters with a known potential solution (add a cathode to emit electrons into the exhaust; doing so would seem to negate some of the benefits of the design, though), and whether or not there'd be enough force exerted on the scoop to counter the thrust.
I'm totally accepting that I might be missing something totally obvious to someone with an actual physics background here, but I don't understand what based on your previous comments. I'm sure you're busy but I'd love it if you have time to expand on what you believe the problem is in a suitably dumbed down way...
While I'm aware of the dangers of appealing to authority, Zubrin is not some random nobody proposing something with no history of work in this field - he's worked on analyzing scoop based and ion based thruster designs for a very long time, so I'm inclined to give him some benefit of doubt [this idea should at least be "very simple" to test given the predictions in the paper - he's suggesting a small but measurable amount of thrust should be possible to generate with screens with a diameter down into a few meters radius if he's right].
E.g. he published work together with Dana Andrews analyzing the effects on a Buzzard ramjet design as far back as 1985 that was already looking at what it would take to allow accelerating into a plasma stream (specifically they found that their hypothetical Buzzard ramjet design would be unable to accelerate into the solar wind), and subsequently co-invented the mag sail propulsion idea with Dana Andrews in 1988 as an attempt to fix the flaws they believe the Buzzard ramjet idea has, and he's worked on these types of concepts ever since.
My broad take on the distinction (assuming I have the right general idea about ion drives!) would be that if you're the one providing the material to be ionized, you get a lot of control over its initial conditions. But if you're planning to use matter that's externally available, your control is a lot more limited. In the schematic that I've seen for this dipole drive, the ions begin far away (so, at zero potential) and end far away (again at V=0), so I don't see a way to avoid that.