Re: What's a good "proton reflector"?

In article <1133467214.065844.106200@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>, mechdan@xxxxxxxxx writes:
>mme...@xxxxxxxxxxxxxxxxxx wrote:
>>In article <1133445373.673837.215920@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>, mechdan@xxxxxxxxx writes:
>
>>>I'd previously done some analysis on performance based
>>>on .17c fast protons from a D-He3 reaction in the final
>>>stage of a multi-stage bomb, but I couldn't come to any
>>>final conclusions because I lacked any data on how much
>>>yield you could expect from D-He3 in a bomb.
>
>>Well, I'm not sure. Energetically, D-He3 is comparable to D-T,
>>fractionally higher, in fact. However, the cross section is
>>significantly lower, by nearly 2 orders of magnitude so I don't know
>>whether the confinement under explosion conditions is long enough to
>>achieve good burn. That'll take a fusion expert opinion (Bruce, you
>>there?)
>
>The two orders of magnitude lower cross section is what
>worried me the most. I should note that in my reading
>about fusion science, I don't actually "understand" the
>eaning of cross section units. I can understand
>differences in scale, but that only tells me a difference
>in relative terms. It's not like joules or newtons where
>I know how to do direct calculations.
>
The cross section is the effective target area that one of the
interacting particles presents to other, so the larger the cross
section, the higher the interaction rate. More specifically, given
two species of particles (Say, D and T) with present at densities
n1 and n2, moving with mean relative speed of <v> and with reaction
cross section sigma, the interaction rate (whether in terms of energy
released per unit volume per unit time, or interactions per unit
volume per unit time) is proportional to n1*n2*<v>*sigma. That's the
story in a nutshell. Note that:

1) At non-relativistic energies you've <v> proportional to sqrt(T)
when T is the temperature of the ensamble.

2) Even though it appears from the expression above that the
interaction rate is simply proportional to mean velocity, this is not
true since sigma itself is in general a function of velocity.

On the other hand it is true that at any given temperature the
reaction rate is proportional to the appropriate cross section.
Simply put, stuff with higher cross section "burns" faster. The term
"burns" applies since cross section is equally aplicable to chemical
reactions.

Now, same as with chemical processes, if you can have unlimited
confinement time, the reaction rate (as long as not excessively low)
is of secondary concern. The energy content is primary. Slower rate
just means that you need a bigger boiler. On the other hand, if the
confinement time is very short, you do care about reaction rate since
high energy content won't help you much if most of the fuel won't burn
and all (Thus, heavy oil is fine for home heating but gasoline is much
better for a car engine).

So here is the problem with He3 in a nuclear device. It reacts with
deuterium close to hundred times slower than tritium (at comparable
temperatures and densities) thus it may be (mind you, just may be, I'm
not privy to the calculations) that it is just too slow for the
confinement time available (which ain't much). If that's so, then a
nuclear device fueled with D-He3 will pretty much fizzle.

>And of course, .17c isn't fast enough for my tastes
>anyway. And of course, the protons go off in every
>direction uncontrollably, so only a small fraction of
>them will hit the starship...

And that's why you would like a charge that'll focus and, if possible,
accelerate them. Of course.
>
>For these and other reasons, I've been looking more at
>the fission side of things, based on the figure of about
>80% yield for tertiary U238 fission. That's a number I
>can wrap my head around, even if it means I'm almost
>ignoring the potentially more powerful fusion stage.

Well, this stage is powerful thoough the fussion fragments are much
slower than your protons. Still, they carry a hefty momentum and
significant charge.
>
>Based on your feedback in this thread, I've changed the
>focus of my attention to magnetic pinch. I'm reading
>up on theta-pinch plasma research, as it's closely
>related to what I'm thinking of. Current research in
>theta-pinch understandably revolves around electromagnets
>which do NOT self-destruct in a thermonuclear detonation
>upon use. Thus, the data is many orders of magnitude off
>of what I'm interested in (~35km/s vs ~100,000km/s).
>At least it's a place to start.

There was significant work done many years ago (some may still be done
nowadays) on explosive compression of magnetic fields (basically
imploding coils with explosives). That's more along the lines of what
you're looking for. I have no references, unfortunately, but google
should help.

>
>>>There was also significant work done on high energy lasers
>>>within the Star Wars years, and getting a laser beam with
>>>useful divergence proved beyond the reasonable. Technology
>>>improved in the '90s, making high energy lasers reasonable
>>>thanks to "new things under the sun", which people hadn't
>>>thought of before.
>
>>No, you're wrong on this. There was no problem with
>>laser divergence there. Laser divergence was, from
>>the beginning, pretty much diffraction limited.
>
>Not within the atmosphere, which is where most laser
>weaponry proposals live. Even Space Based Laser
>hoped to kill missiles within the atmosphere where
>aerodynamic forces would turn even minor damage into
>catastrophic failure.

That's true, but secondary, separate from the divergence of the laser
per se. The primary killer was still the fact the the lasers didn't
have enough oomph to do what was required, even out of the atmosphere.
>
>>>>Not to mention that the interaction with the
>>>>neutralizer itself introduces beam blooming well in excess of the
>>>>picoradian you need. And there is no further cooling after the
>>>>neutralizer. There is no magic in this business.
>
>>>You can cool a beam after the neutralizer;
>
>>Nope.
>
>I got into a long argument with Luke Campbell on the
>question of laser cooling a particle beam. I was
>arguing the other side, but really couldn't come
>up with any reason why it couldn't be done in principle.

Well, for this you've to see how laser cooling works. I suggest you
google for this a bit the, if this is not satisfactory, I'll be glad
to oblige further. But the key point is that lasers cool
one-dimensionallyly, along the direction of the beam. So for full
cooling you need three pairs of lasers, one along each dimension (each
pair consists of one laser tuned just slightly below an excitation
frequency of the atom/molecule (corrected for the doppler shift) and
one just slightly above. So, yes, you could use a pair of lasers
colinear with your particle beam to cool the longitudinal direction,
but that's the direction you don't care about. You care about the
transverse one (that's the one that gives you divergence) but for this
you would need lasers flying parallel to the beam, firing at 90
degrees to it.

You could reduce the number to less than six, down to four in the
appropriate geometry, but that still gives you 3 of axis ones.
>
>Of course, just because something can be done in principle
>doesn't mean that can be done easily or even at all,
>once the practical difficulties are taken into account.
>
>>>even I'm aware of at one possible mechanism--laser cooling.
>
>>Sorry, no. Laser cooling is fine for *stationary* trap, not for a
>>beam. Check the geometry required.
>
>What's the problem with the geometry? You need lasers
>aimed at the beam from at least 4 different directions
>(in a tetrahedral formation, for example). One of
>those can be along the axis of the beam, and can be
>a narrow ray. The others need to be optically "spread"
>into an arc rather than a single ray, but that's no biggie.

They cannot be spread into an arc, the angle the laser beam makes with
the particle beam needs to be very precisely defined since this angle
determines the doppler correction, thus the tune. If you spread into
an arc, only a very small portion of the laser beam will have the
right tune. And, since you cannot spread the laser beam, each parcel
of the particle beam will be crossing the side beams in some
nanoseconds, not enough time for any decent cooling.

>>You do not cool an atom assembly by a single laser. It takes
>>criss-crossing beams to do it, In a stationary trap (and I count
>>slow, thermal, atomic beam as stationary, that's easy to do. With the
>>beams we're talking about, you'll need laser flying alongside the
>>beam, on all sides, in formation, to get what you need. Not very
>>practical.
>
>There's no need for the lasers to be flying in any
>direction. You just need to adjust the frequency of
>the lasers according to the necessary doppler shift.

See above.

>Of course, with a wide enough "arc" angle, the necessary
>doppler shift will vary with aimpoint.

Aha. Now, check what tuning precision is required and estimate the
width of the arc allowable. It is very little.

> Luke Campbell
>noted that you'd need to conceptually split up the beam
>path into multiple cooling zone segments, each with
>different doppler shift settings.
>
Now we're getting into an "escape to complexity", again. Lets throw
more lasers on it. Not enough? Lets add some nanorobots, perhaps
some VN machines and whatever else we can think of. With enough
buzzwords it is bound to work, right?:-)

Any scheme which depends on a large number of separate components, all
of which are required to work just right to very high precision, at
distances which do not allow for any real time feedback, is a no go.
That's a basic rule in robust design, and when you're taking a shot
that is so expeensive that in case of failure you may have to wait a
very long time before you can repeat it, the design will better be
very robust.

>In principle, appropriate doppler shift adjustments
>alone are good enough to accelerate a beam pulse
>in the first place--no ionization necessary, and the
>neutral beam pulse is cool from start to finish.
>This wouldn't be an efficient way of doing things,
>of course.
>
Run the numbers and you'll find how much inefficient.


>>>I know. I just wish someone else was interested enough in
>>>this sort of propulsion to publish such a "what if" article.
>
>>Perhaps somebody will be. The relatively easy part is to calculate
>>what sort of impulse can be extracted from a given proton puff by a
>>given magsail (but I trust this has already been more or less done).
>
>Yes, I found plenty of literature on that part in the
>context of particle beam propulsion.
>
>>The not so easy part is estimating what sort of puff is achievable.
>>Alamos or Livermore could probably give a decent answer but I doubt
>>they will:-)
>
>The experts who have first hand knowledge of the data
>can't talk about it. :(

I know:(
>
>>>I appreciate your thoughts and advice. Thank you.
>
>>You're welcome. Sorry if I've been less then curteous
>>but, being an experimentalist I'm trained to seek holes
>>in schemes.
>
>Hey, I understand. I know I'm what you call "an enthusiast".
>Maybe even a borderline net-kook, at times. Throughout
>the years, I've had a lot of far-out ideas.
>
Nah, you don't qualify as net-kook. There is nothing rare about
far-out ideas. The characteristics of a net-kook are:

1) He just knows, with absolute certainty, that he's right.
2) He's forever locked in "send" mode, never switching to "receive"

You fail the test on both counts, sorry:-)

Mati Meron | "When you argue with a fool,
meron@xxxxxxxxxxxxxxxxx | chances are he is doing just the same"
.

Relevant Pages

  • Thrity Years From Now
    ... The 'plug in the wall' could be any number of energy creating tech- ... ped with laser energy outputs of 10**4 joules, ... provide initial damping for the lense, mirror housings, pulse ... The reason for beam shaping is to compensate for spatial gain ...
    (sci.space.policy)
  • Re: Whats a good "proton reflector"?
    ... >>>imploding coils with explosives). ... CONCEPTUALLY splitting up the beam into multiple cooling ... Optical prisms and/or scanning the laser back ... >>bomb units are exactly along the starship path centerline rather ...
    (sci.physics)
  • Re: Whats a good "proton reflector"?
    ... >>>up with any reason why it couldn't be done in principle. ... for this you've to see how laser cooling works. ... along the direction of the beam. ... >randomizing nature of gas particle collisions. ...
    (sci.physics)
  • Re: Whats a good "proton reflector"?
    ... but I expect the energy content of chemical explosives ... > CONCEPTUALLY splitting up the beam into multiple cooling ... Optical prisms and/or scanning the laser back ... >>>bomb units are exactly along the starship path centerline rather ...
    (sci.physics)
  • Re: A modification of Fizeaus toothed-wheel experiment
    ... To get a short laser pulse, ... > picosecond laser and a timing control/recorder device. ... > At the other end is a particle beam moving at a significant ... > the beam is shooting into the track). ...
    (sci.physics.relativity)
Loading