Index Home About Blog
From: (Henry Spencer)
Subject: Re: EM Propulsion
Date: Fri, 20 Aug 1999 01:59:22 GMT

In article <01bee784$b2971e20$7c05bfce@default>,
M. James <> wrote:
>It would seem to be a good idea if the launch site were on the moon.  A
>linear induction motor would be easy to build.  (The LIM was sugested by
>Arthur C. Clarke in a story written about 1948, and was used right on Earth
>to get a spacecraft up to speed to start a ramjet from the Australian

Alas, it turns out that linear induction motors -- although quite simple
to build -- do not scale up well.  It's difficult to build a large one
that will produce significant accelerations.

Modern designs for such things use linear synchronous motors instead.
They have the disadvantage of needing coils on the moving assembly (so one
ends up with the O'Neill mass driver, with coil-equipped "buckets" which
are accelerated, release payloads, and then are decelerated for re-use so
you don't throw the coils away).  They have the advantage of being able
to operate at immensely higher accelerations.

(The early writers thought a few Gs might be feasible with the induction
motors -- probably not true.  O'Neill's first mass-driver prototype, built
out of the MIT electrical-engineering junk box, ran at 30G, and improved
later models demonstrated over 1000G.)
The good old days                   |  Henry Spencer
weren't.                            |      (aka

From: Doug Jones <>
Subject: Re: Mass driver simpler than ISRU?
Date: 04 Jul 1998

Cfrjlr wrote:

[in reference to high speed limitations of coilguns]

> So far nobody has provided any references to the "problems", I am
> beginning to suspect they do not exist.
> Charles F. Radley

This problem was identified by SSI many years ago, but never was solved.
While the energy storage per unit length of a coilgun is ideally
constant, the power level rises with the bucket velocity since the
energy must be delivered in a shorter period of time.

Higher speeds require faster switching in proportion to the bucket
velocity.  This is not a problem for the start transient (turning on the
current in a coil), but the coil current must be cut off as the bucket
passes through, else the bucket is not accelerated. Since V = L dI/dt,
the switches tend to fry when dI/dt is increased at higher bucket
velocities.  Even IGBTs (insulated gate bipolar transistors, the most
effective power-switching devices in existence) run out of current
cutoff capability at relatively low speeds and bucket masses.

Doug Jones
EE/Propulsion engineer

From: (Andrew Higgins)
Subject: Re: Mass drivers - will they work as advertised?
Date: Thu, 24 Sep 1998 01:19:00 +0200

In article <360935A8.203@nospam.comchangenospam2ti>,
mikecombs@nospam.comchangenospam2ti wrote:

> Mike Combs wrote:
> >
> > Thanks for posting, and for the references, Andrew.  I'll keep looking
> > into this.
> Well, I looked up the references Andrew kindly provided.  Although
> Cowan's coil gun is a somewhat similar technology, or at least has
> similar design goals as a mass-driver, I think we've been saying we know
> the apple's gone bad because the orange is rotten.
> I see significant technical differences between the Sandia Labs coil gun
> and a mass-driver.
> The coil gun is described as launching 1,050 kg projectiles.  I'm vague
> on the exact weight of a mass-driver bucket, but I'm pretty sure it's
> much, much lower.

But the coilguns *tested* at Sandia and elsewhere used projectiles that
were only 5 kg or so, and they *still* couldn't break 1 km/s.

> The coil gun had a solid, cylindrical barrel.  The mass-driver was an
> open framework.

I don't believe this difference is anything more than cosmetic.

> The mass-driver had super-conducting coils in the buckets.  As far as I
> can tell from the article, the coil gun armature is ohmic.  It is
> described as having an induced current.

This is a distinction worth pointing out, though it is not obvious that it
will help to overcome the limitations which have appeared to date.  Also,
using superconducting coils in the buckets has its own challenges, as I
understand it, such as getting power to the bucket.  I once heard Brian
O'Leary say (circa 1987) that this was one of the mass driver engineering
issues which SSI never really addressed.

> It is true that no coil gun or mass-driver has ever achieved greater
> than 1 km/sec velocity.  The only way to finally prove that a 160 meter
> long mass driver can achieve lunar escape velocity is to build the
> entire length in a metal jacket, pump out the air, and give it a whirl.

No.  This statement is tantamount to saying, "The only way to prove this
new launch vehicle design will make orbit is build the thing full-up and
fire it off."  But if the thrust of your new engine on the test stand does
not exceed the gross weight of the vehicle, you can *prove* it isn't going
anywhere without actually building it.

This is the situation mass drivers are in right now.  What is desperately
needed is more modeling and component development, *not* a test of the
full scale device.

Now, I did not say the SSI-style mass driver would never reach lunar
escape velocity.  For the record, I have always been supportive of SSI's
work and, ten years ago, became a Senior Associate of SSI.

If you recall, this thread was initiated a few months ago regarding the
comparative feasibility of Martian in-situ propellant production vs.
mass-driver launch of lunar LOX.  One of these concepts has been
demonstrated in the lab, the other has not.  My point in bringing up the
disastrous failure of coilguns was to show that some serious engineering
challenges remain in realizing a mass driver capable of lunar escape
velocity.  I feel we now have consensus on this issue.

> I wish somebody had handed SSI 1/10th of the money rail guns got.
> Throughout all of this, I'm trying to bear in mind that when a person
> becomes emotionally-committed to an idea, it often leads them to be
> resistant to suggestions that there are problems, and causes them to
> mentally-filter-out contrary data.

I openly admit I have a jaundiced eye when it comes to this issue, having
worked for six years on the prototype Ram Accelerator facility at the
University of Washington, Seattle.  If you've read the Sept. 1993 "Air &
Space" article, you now know just how thoroughly the EM launcher community
(railguns and coilguns) poisoned the funding well for development of
alternative hypervelocity launchers, such as the ram accelerator.  This is
a grudge I will likely take to my grave.
     Andrew J. Higgins            Department of Mechanical Eng.
     Shock Wave Physics Group     McGill University    Montreal, Quebec

Index Home About Blog