Lorrey Drive

Hara Ra (harara@shamanics.com)
Fri, 27 Dec 1996 22:12:56 -0800


Ah, memory.... I once invented something very similar. Imagine a
cardioid with its cusp on the Y axis, much like the usual drawing of a
heart. At the tip of the cusp place an axle with a hole in it through
which a shaft is placed. As the axle rotates, the shaft slides through
the axle guided by the cardioid. (The width of the cardioid is constant
regardless of the shaft angle.) The center of mass of the shaft
describes a circle, and as it rotates at a constant speed (as driven by
the shaft), the net result appears to be a centrifugal force pointing
downward.

I actually tried to build a physical model, but my limited ability with
metalwork (a teenager with hand tools) kept me from making one which
would actually turn. Years later I realised that all that would really
happen is that the device would jiggle a bit around the center of the
circle described by the shaft's center of mass, with no net momentum
transfer outside of the system....

O---------------------------------O
| Hara Ra <harara@shamanics.com> |
| Box 8334 Santa Cruz, CA 95061 |
O---------------------------------O

Eliezer Yudkowsky wrote:
>
> Okay, I've taken a look at the Lorrey drive, and I've found what appears
> to be a gaping mechanical flaw.
>
> > Imagine a centrifugal drum, like that seen edge on in the animation
> > below. If you had several masses in the drum, rotating around the
> > geometric center, they would all produce equal centrifugal force at
> > every point in their orbits, so you would essentially have the
> > workings of a gyroscope. However, if you were able to speed each
> > mass up at one point in the orbit, and slow it down again on the
> > opposite side of the centrifuge, you would have a point in the
> > centrifuge where all masses were going slowly on that side, and
> > fast on the opposite side. This would produce excess centrifugal
> > force in a direction straight out from the point of highest velocity.
> > you would also impose a large amount of torque on the drum and
> > chassis of the device.
>
> Positions | Velocities | Accelerations
> _______ | _______ | _______
> / A \ | / A-> \ | / >A \
> / B \ | / ^ B \ | / <<<<B \
> \ D / | \ D V / | \ D>> /
> \ C / | \ <-C / | \ >C /
> ^^^^^^^ | ^^^^^^^ | ^^^^^^^
>
> Take a mass M, weighing one gram. M goes around in the centrifuge
> shown. At D, M travels at 5 meters/second, at B, 10 meters/second. At
> A, M is pushed forward by an extra 5 m/s and at C, M is slowed by 5 m/s.
>
> The total system, including the centrifuge, works like this. Start at A
> before M is accelerated, assuming the whole device to have zero
> momentum. At A, M is pushed forward by 5 m/s. This transfers
> -5gm/s of momentum to the centrifuge. M goes around the centrifuge
> through B, coming out at C going -10 m/s instead of 10, transferring 20
> gm/s of momentum to the centrifuge. M is then slowed down at C,
> transferring another -5gm/s to the centrifuge. M then goes from C to A
> through D, going from -5m/s to 5m/s, transferring -10gm/s of momentum to
> the centrifuge.
>
> Net change in momentum: -5 + 20 + -5 + -10 = *0*.
>
> It does not work. It will not go. Sorry, Mike.
>
> Trust me, I know how you feel - I, too, had a cherished Great Idea, but
> luckily found the flaw *before* it was published.