Looks to me like Jones will eventually get a Nobel prize. The best
experiments seem to be getting Jones level fusion (down in a submarine etc).
Assuming that this is becoming more and more reproducible/understood
and that Pons never is able to demo anything, how long should Jones
have to wait? As I understand it in some cases 10 to 20 years can pass from
the time someone gets an amazing discovery to the time they get the prize.
What is a normal delay?
-- Vince
>
>Looks to me like Jones will eventually get a Nobel prize. The best
>experiments seem to be getting Jones level fusion (down in a submarine etc).
Whether jones gets a Nobel prize will depend on whether the process
occurs in abundance, naturally. If the mostly likely
scenario is correct---``fracto-fusion''---He will definitely _not_
get the Nobel prize. Fusing of D's that get accelerated across voltages
in microcracks is an interesting, but hardly Earth-shaking phenomena,
and not likely to occur much in nature.
You see, the essential importance of Jones research is in explaining
the unusual abundances of Tritium, etc, observed in the deep Earth
(and released volcanically). If he has found a process which does this,
that would be scientifically noteworthy. If he has just found an amusing
way to achieve a few fusions per second in a material, that would not.
In that case, the discovers of muonic fusion would deserve a Nobel
long before Jones, since their method is equally creative and much more
succesful.
If Jones' (unlikely) theory (that the fusion he observes is due
to the formation of a metallic D state) is correct, and if this
is achieved also within the Earth, then I would give him a moderate
chance for Nobel, a long time from now.
Remember that the founders of the ordinary routes to Fusion
energy have yet to receive their Nobels---anyone who wants a
Nobel _purely for fusion research_ must either
(a) take their place in line.
(b) make a major breakthrough in fusion energy production
(c) show their fusion effect explains some important natural phenomena
Jones only chance is (c) (given that the line in (a) is long and
that (b) is not forthcoming).
--
Barry Merriman
UCLA Dept. of Math
UCLA Inst. for Fusion and Plasma Research
ba...@math.ucla.edu (Internet)
Looks to me we will have to await *systematic* verification of those
results. For now, experiments done at the ultra-low background lab
beneath the Great Sasso in the Alps have produced negative results.
By the way, background also arises from construction materials (I
suppose that Jones did not use a nuclear submarine :-) )
Raul A. Baragiola \Internet: ra...@virginia.edu
Dept. Nuclear Engnr. and Engnr. Physics \Phone: (804)-982-2907
University of Virginia, Charlottesville, VA 22901 \ Fax: (804)-924-6270
Well, doesn't that overlook modern possibilities in material science
and, especially, forthcoming nanotechnology. If the microcracks are
sparse "in nature" -- that is, in contemporary palladium or titanium
bars, there would be an incentive to find ways to fabricate the stuff
to have more of those features -- say, in molar quantities. That
could increase the activity by factors like 10**15 or so, couldn't it?
Even if you obtained only one fusion at the cost of wrecking one
microcrack, you might exceed breakeven, if the fabrication process
isn't too expensive.
>Even if you obtained only one fusion at the cost of wrecking one
>microcrack, you might exceed breakeven, if the fabrication process
>isn't too expensive.
No, fracto-fusion could never reach breakeven, for fundamental reasons.
The rate at which deuterium ions lose energy in matter is simply too large,
by many orders of magnitude, compared to the fusion rate. Were this
not so, one could achieve breakeven simply using a particle accelerator
and a target.
Paul F. Dietz
di...@cs.rochester.edu
Possibly. So, before passing judgement, lets examine the ultimate
potential of the fracto-fusion approach.
In the fracto-fusion approach, electric fields
across micro-cracks accelerate D's (or D,T) into eachother at energies
high enough to creat fusion.
The essence is: chemical energy of the substrate is converted
into the kinetic energy of the D. So, the ultimate potential of the
method (ignoring details of crack phenomena) would be to use
all available chemical energy to accelerate D's.
The available chemical energy in a material is on the order
of 1eV/atom (if more energy than that were stored, the material
would ionize!). In order to fuse, colliding D's need on the
order of 1 MeV to overcome the coulomb barrier. Thus every 10^6
substrate atoms could enable one fusion. Since each fusion
releases around 10 MeV, we get Q = 10 (Q = energy out/energy in).
I doubt this best-case Q is high enough to be useful---for once you add in
inefficiencies in the acceleration process, conversion of heat
to useful energy, and the loss of substrate due to material damage
(all those 1 MeV D's zipping around inside it), you would be lucky
to achieve Q = 1.
And even if Q = 1 were achieved, since the input power is chemically
supplied, that means you have a nuclear power source
that is about as powerful as a usual chemical power source. This is
not at all like normal fusion and fission, which result in reactors
with much more power density than chemical power sources.
Such a device might be useful as a neutron source, but it would
have little to recommend it as a power source (you get the
rad-waste of nuclear energy coupled with the low power density
of chemical energy---the worst of both worlds).
But this energy gain is valid only for those D that fuse. The inmense
majority will slow down in energy due to excitation and ionization of
the atoms in the solids. This is because atomic cross sections are more
than a million times larger than fusion cross sections. This was pointed
out by a netter a few days ago, who noticed that otherwise we could get
breakeven with a particle accelerator.
-Kevin Flanagan
Student R.I.T.
------------------------------------------------------------------------------
I love antiques; This terminal is 11 | My other computer is a HP 32S
years old and preforms better than | Calculator
most emulations I see out there |
_____________________________________________________________________________
B.T.W. Please pardon all spelling errors, I can't opperate the spell
checker on this system...ULTRIX illiterasy...I could give you the
address of my 12th grade english teacher if you want a scape goat.
Along with the tritium from volcanic emissions it also explains the extra
heat coming out of Jupiter and the helium in the diamonds. These sort of fit
with Jones' old paper (~1986 ?) about rates of fusion under extreme pressure.
His general theory and experiments explain several curious puzzles (not just
one). I can't see how else one could explain the heat of Jupiter or the
tritium from volcanos - seems Jones has to be right. Are there any other
working theories for these?
-- Vince
PS I still say Jones is headed for a Nobel prize.
This is *not* a topic on which I claim any special expertise (in spite
of the fact that it is, technically, astronomy), but before all this started
up I was informed by Roman Smoluchowski that Jupiter's heat is produced
by the gradual settling of the core, in particular by the buildup of
Helium in the core (and its depletion in the outer layers). This explanation
is widely regarded as plausible, if not completely compelling.
--
I'm not afraid of dying Ethan Vishniac, Dept of Astronomy, Univ. of Texas
I just don't want to be {charm,ut-sally,emx,noao}!utastro!ethan
there when it happens. (arpanet) et...@astro.AS.UTEXAS.EDU
- Woody Allen (bitnet) ethan%astro.as....@CUNYVM.CUNY.EDU
In two words, alpha decay. The earth's crust is surprisingly rich in thorium
and uranium ("surprisingly" in the sense that they're relatively abundant for
their high atomic numbers), and their decay chains proceed via several steps
involving emission of alpha particles, which after all are just helium-4
nuclei. Cosmic-ray-induced spallation reactions also produce some helium, but
alpha decay of actinides is more important.
The question of the conditions under which helium collects underground is a
complicated one from a materials-science point of view, but there is no need
to invoke fusion, or any other unexpected reaction, to explain the presence of
the helium to begin with.
--
Bill Johnson | "A man should never be ashamed to own he
Los Alamos National Laboratory | has been in the wrong, which is but saying,
Los Alamos, New Mexico USA | in other words, that he is wiser to-day
!cmcl2!lanl!mwj (m...@lanl.gov) | than he was yesterday." (A. Pope)
>Looks to me like Jones will eventually get a Nobel prize.
Well, I hope so. Of course since I am a BYU alumnus I couldn't be
at all biased. :-)
Realisticly I'm afraid there is a too much unknown at this time to
have any confidence in Vincent's prediction. This could turn out in
any of at least the following ways:
1. Jones et al may be proven wrong. The neutrons they have
detected may be only statistical fluctuations or erroneous
measurements.
2. They may have something on the order of muon induced fusion.
Interesting but not significant.
3. They may explain some things like volcano emitted tritium,
making the work significant but probably not of Nobel Prize
significance.
4. This work may eventually (over many years) lead to a controlled
fusion energy source. This would be of Nobel significance but it
does not appear likely at present and even if it occurs is a long
way off.
>And even if Q = 1 were achieved, since the input power is chemically
>supplied, that means you have a nuclear power source
>that is about as powerful as a usual chemical power source. This is
>not at all like normal fusion and fission, which result in reactors
>with much more power density than chemical power sources.
I don't think power density is the main issue here (although high
power density would be an advantage in many applications). The
issue is power available from a *large* fuel source. Most of the
current sources of fuel are likely to become quite scarce in a few
decades. However if we can find a way to burn deuterium we should
have centuries worth of fuel available.
>Colliding D do not to have energies of the order of 1 MeV to overcome
>the Coulomb barrier and fuse, since they can tunnel through the barrier
>at much, much lower energies. Otherwise, the sun would not shine.
This is certainly true, but as you point out:
>But this energy gain is valid only for those D that fuse. The inmense
>majority will slow down in energy due to excitation and ionization of
>the atoms in the solids.
And because of this, I think my 1 MeV fusion threshold is more reasonable
for the above calculation.
True, you get plenty of fusion from a 10 keV plasma---but, any given
D undergoes _many_ collisions before actually fusing, since the probabilty
of fusion is low for any given collision.
In my simplest picture of fracto-fusion, I imagine the D accelerated across
a crack, colliding with another D, and if they don't fuse, they
head off into the material and slow down rapidly by losing energy
to atomic collisions. So, I figured that unless the D energy were
high enough to make ``one shot'' fusion likely, it would
not fuse. For a single collision to have a good chance of fusing,
the D had better have around 1 MeV.
To do it ``right'', one simply needs to work out a Lawson criteria
for fracto fusion---but this will involve details of the cracks
(say volume fraction), since the fusion can only occur in or on the boundary
of the cracks---inside the lattice, the D would cool down quite quickly.
Well, the helium point is a very good point. I was just recently at a
talk on granites and slate belts and one of the main "problems" brought
up was how sediment "mud" get as hot as it gets. There is some
anamolous heat source which causes a much higher geothermal gradient
(35C/km as opposed to a more normal 15C/km) in slate belts (which
were previously basins much like the ones oil wells are found in)
that occur BEFORE the intrusion of the granites and are most likely
responsible for the granite melting in the first place.
One place where this high gradient is observed directly is in some
Venesualian oil fields and current thought is that it is due to
oxidation of organic matter. Of course where does the oxygen come from?
No, I don't know if there is any He in the Venesualian oil fields but
if I get the time I'll look it up.
Jim Bowers
[1--3 omitted]
>4. This work may eventually (over many years) lead to a controlled
>fusion energy source. This would be of Nobel significance but it
>does not appear likely at present and even if it occurs is a long
>way off.
If this were to occur, I think Jones would get the Nobel prize
for ``largest scaling up in the history of science''. Right
now they claim on the order of 1 neutron/sec, and a 1 GW reactor
would imply about 10^21 neutrons/sec---scaling up by 21 orders of
magnitude!
>In the fracto-fusion approach, electric fields
>across micro-cracks accelerate D's (or D,T) into each other at energies
>high enough to create fusion.
>
> . . . . In order to fuse, colliding D's need on the
>order of 1 MeV to overcome the coulomb barrier. Thus every 10^6
>substrate atoms could enable one fusion. Since each fusion
>releases around 10 MeV, we get Q = 10 (Q = energy out/energy in).
>
> . ...
>And even if Q = 1 were achieved, since the input power is chemically
>supplied, that means you have a nuclear power source
>that is about as powerful as a usual chemical power source. This is
>not at all like normal fusion and fission, which result in reactors
>with much more power density than chemical power sources.
Fission devices CAN develop very high power densities as evidenced
by their "bomb" applications. In MOST OTHER APPLICATIONS they do NOT,
for the engineering consideration, that the temperature can't exceed
the melting (or pressure) limitation of the confining structure.
Magnetic fusion with solid vacuum chamber walls has the same problem
only worse. Their problem is that the walls must remain an extremely
high grade vacuum barrier in spite of the "wished for" megawatts/meter
squared of soft X-rays and copious neutron and other particle insults.
A high pressure coal burner works better simply because of the high
density of cooling tubes that can matrix the reacting (burner) region.
These potential and real applications of nuclear technology to
energy (power) production are rather pathetic, since while the energy
density ratio of the fuels favor nuclear by the millions, the power
density realized is subnormal to advanced chemical technology devices.
We are interested in correcting that problem, but it's going to be
a difficult task to get our solution developed, since it competes
with a very troubled (Inter)National Fusion program.
>Such a device might be useful as a neutron source, but it would
>have little to recommend it as a power source (you get the
>rad-waste of nuclear energy coupled with the low power density
>of chemical energy---the worst of both worlds).
IF.. you need a few neutrons, a pocket sized source is better
and much, much easier to do environmental containment on.
+---------------------------------------------------------+**********+
| +Commercial*
| Paul M. Koloc, President (301) 445-1075 ***FUSION***
| Prometheus II, Ltd.; College Park, MD 20740-0222 ***in the***
| mimsy!prometheus!pmk; p...@prometheus.UUCP **Nineties**
+---------------------------------------------------------************
Well, that's an easy one. The answer is quantum black holes! Some
Russion chappie has recently proposed this as an explanation for several
geological phenomena, including the earth's magnetic field,volcanic hot
spots and gravitational anomalies. He has even proposed a way to test the
hypothesis, by sensitive gravimetric measurements around mountains.
I don't understand how a qbh can be locked in position though (the theory
requires at least a hundred of the little devils dotted around the
earth - with a bigger one at the center).
Maybe qbh's are responsible for cold fusion! :-)
Dick Jackson
A more interesting thing to know is the He3/He4 ratio. The presence of
He can be acounted for much better by present therory than can the
variations in this ratio.