Index Home About Blog
Subject: Re: Ion Rockets [[was power]]
From: "Gordon D. Pusch" <pusch@mcs.anl.gov>
Date: May 31 1996
Newsgroups: sci.space.tech

The following message is a courtesy copy of an article
that has been posted as well.

In article <Ds88Bo.2tw%spenford@zoo.toronto.edu> Henry Spencer
<henry@zoo.toronto.edu> writes:

> In article <9604248329.AA832981562@fhu.disa.mil> "Terry Colvin"
> <colvint@fhu.disa.mil> writes:
> >
> > ...I suppose, however, that if it had turned out that nuclear bombs
> > could not be built, then maybe the nuclear arms race would have 
> > started off with the production of antimatter bombs.
>
> Very unlikely, actually.  Antimatter does not make good bombs.
> Even more ordinary nuclear bombs can "fizzle" unless carefully
> designed: the reaction gets going but too slowly, so the bomb
> blows itself apart before the reaction can proceed very far.

Ummm... I think I have to disagree with you on a number of points, Henry...

Fission bombs can ``fizzle'' because they rely on a =chain reaction=.
Hence, as you say, if the reaction gets going too slowly, one gets an
incomplete ``burn,'' since the bomb ``catastrophically disassembles
itself'' --- and in this ``disassembled'' state, the chain reaction
stops.

In the case of fusion bombs, the reaction is strongly temperature-
and density-dependent; unless the reactants stay hot enough and
dense enough for long enough (Lawson criterion!), the reaction
will not go to completion.


M/AM annihilation, by contrast, is =NOT= a chain reaction.
Furthermore, while the annihilation =rate= will depend on the
temperature of the reactants, the annihilation =efficiency= will not.
Hence, one is guaranteed that 100% of the antimatter *WILL* annihilate
with matter virtually 100% of the time, so long as the bomb and/or
detonation environment consists mostly of matter, and the matter and

antimatter are well mixed. If both the matter and antimatter are gases
or plasmas, it will =NOT= be hard to ensure good mixing --- *especially* 

since, barring someone discovering ``new physics'' that allows matter
to be ``flipped'' or ``rotated'' into antimatter, it is highly unlikely
that we will be able to manufacture and store any form of antimatter
other than antihydrogen ``ice'' in the foreseeable future. Hence, the
problem will instead be keeping it cold enough to =prevent= it from
evaporating and mixing !!!


> With antimatter this problem is far worse, because while fission
> and fusion occur throughout the reaction volume, the matter-
> antimatter reaction occurs only on a contact surface.  

This may =perhaps= be true of the infamous (and hypothetical)
``contraterrene cannonball,'' but it will certainly =NOT= be
true of antihydrogen gases or plasmas !!!  From the LEAR
experiments, we know that the annihilation lifetime of a slow
antiproton in condensed bulk matter is quite short; hence all
that will be required is a way of rapidly mixing the antimatter
with matter.  For example, one could implode a shell of normal
matter onto an antihydrogen ice nugget, and instead of trying to 
*suppress* turbulent mixing as one does in inertial confinement 
fusion, one would instead deliberately *induce* it; near-complete
mixing should be had after only a few Kelvin-Helmholtz instability
times.  Surround the whole thing with a normal-matter tamper to
thermalize some of the emitted radiation, and the mixture should
rapidly heat up to the point where thermal diffusion will
complete the mixing process.


> It's exceedingly difficult to get a major explosion with antimatter.
> (Tiny ones are not hard, since the square-cube law gives you more
> surface area per volume as the scale shrinks.)

The square/cube law will be irrelevant to sufficiently well-mixed
gases or plasmas of matter; as I've argued, it should be possible 
to achieve this on timescales comparable to a few Kelvin-Helmholtz
timescales, which can be made short compared to the implosion timescale.

Regarding production of an explosion:   contrary to common belief, 
the majority of the proton/antiproton annihilation energy is released,
not as gammas, but rather in the form of pions (as I believe you yourself 
have pointed out in other posts). One gets roughly equal numbers of
pi+, pi-, and pi0 particles, with energies in the ~400--800 MeV ballpark. 
The pi0's go to two gammas almost immediately; the gammas will have an
attenuation length of only a few tens of gm/cm^2 in matter.  The pi+'s
and pi-'s have a range of only a few tens of meters before they decay
to muons, even in vacuum; furthermore, in matter, both pions and muons
have a range of only a few tens to hundreds of grams/cm^2. Air has a
density of about a 1.25 kg/m^3; hence, most of the released energy
will be deposited within a few tens to hundreds of meters of the bomb;
I would expect this to generate a nice, hot fireball, just like a
fission or fusion device.


> Also, with production technology we can reasonably foresee,
> antimatter is impossibly expensive for weapons applications.
> Even the US military has finite budgets.  The cost of burning
> a city down with conventional weapons is large but not infinite.

Now, on =this= point I have to agree with you, Henry... :-) 
Even =Bob Forward= doesn't think we'll get the price down below 
US$ 60.e6/mg using foreseeable Earth-based technologies --- and,
at 43 kT/gm of antimatter, we're talking roughly =US$ 1.4e9= per
kiloton !!!!!!!!!

As you say, even the Pentagon's budget isn't =THAT= large... ;-)


Gordon D. Pusch                     |  Internet: <pusch@mcs.anl.gov>
Math and C.S. Div., Bldg.203/C254   |  FAX:      (708) 252-5986
Argonne National Laboratory         |  Phone:    (708) 252-3843
9700 South Cass Ave.                |  
Argonne, IL  USA  60439-4844        |  http://www.mcs.anl.gov/people/pusch/

But I don't speak for ANL or the DOE, and they *sure* don't speak for =ME=...

From: "Gordon D. Pusch" <pusch@mcs.anl.gov>
Newsgroups: sci.space.tech
Subject: Re: Antimatter ``bombs'' [was Ion Rockets [was power]]
Date: Sat, 15 Jun 1996 02:00:01 -0500

The following message is a courtesy copy of an article
that has been posted as well.

In article <DsvCy6.3z6%spenford@zoo.toronto.edu> Henry Spencer
<henry@zoo.toronto.edu> writes:

> In article <phbuj5fesu.fsf@destiny.mcs.anl.gov> "Gordon D. Pusch"
>   <pusch@mcs.anl.gov> writes:
> >Hence, one is guaranteed that 100% of the antimatter *WILL* annihilate
> >with matter virtually 100% of the time, so long as the bomb and/or
> >detonation environment consists mostly of matter, and the matter and
> >antimatter are well mixed. If both the matter and antimatter are gases
> >or plasmas, it will =NOT= be hard to ensure good mixing...
>
> Ah, but on what time scale?  That's the heart of my objection.  
> To get an explosion -- in other words, a real live bomb -- you
> need not just good mixing, but rapid mixing.  And in this case,
> the reactants will be working pretty hard to keep each other at
> arm's length.

I've already argued that by exploiting the Rayleigh-Taylor instability
of an imploding shell, it should be possible to get thorough mixing on
a timescale comparable to the implosion timescale, which should be on
the order of microseconds. I'd have to look up the annihilation-rate 
of antiprotons in bulk matter as determined by the LEAR experiments,
but if memory serves me correctly, in liquid hydrogen and helium the
annihilation lifetime was also on the order of microseconds, and should
=decrease= with increasing temperature (the annihilation reaction has
an essentially ``geometric'' cross-section at low energies, so the
annihilation rate should be nearly proportional to the thermal
collision frequency).


RE: the reactants ``working to keep each other at arm's length'' ---
I've always found the ``Leidenfrost layer argument'' to be implausible
for mixtures with scales smaller than the mean energy-deposition length
of the annihilation pions and gammas. 

In the case of an interface between astrophysical-scale domains of
bulk matter and antimatter, I accept Alfven's argument that the
``pressure'' exerted by the annihilation products at the interface
will tend to drive the two domains apart.

However, as I observed in my earlier post, in the case of a reaction
with a ``point-like'' geometry, the released energy will be deposited
over a spherical region tens to hundreds of meters in radius. In such
a geometry, I suspect it is more likely that thermalized annihilation
radiation will act to =contain= rather than disrupt the reactants,
analogous to radiation-induced compression in a thermonuclear device
(particularly if the ``bomb'' includes a tamper). Since only a miniscule
fraction of the annihilation pions and gammas will be reabsorbed by the
reactants themselves, I expect the effective ``pressure'' they exert
to disrupt the mixing matter and antimatter will be very much smaller
than that of the thermalized radiation, because the thermalized radiation 
will have a much shorter mean scattering length and hence will transfer 
momentum much more strongly to the reactants.

At least, that =my= gut feel on this problem. To get a definitive answer
would probably require certain hardware and substantial modifications to
certain software that the U.S. Gov't would rather I not have access to... ;-/


Gordon D. Pusch                     |  Internet: <pusch@mcs.anl.gov>
Math and C.S. Div., Bldg.203/C254   |  FAX:      (708) 252-5986
Argonne National Laboratory         |  Phone:    (708) 252-3843
9700 South Cass Ave.                |  
Argonne, IL  USA  60439-4844        |  http://www.mcs.anl.gov/people/pusch/

But I don't speak for ANL or the DOE, and they *sure* don't speak for =ME=...

Index Home About Blog