[Yum] Multiple Repository Hell

Troy Dawson dawson at fnal.gov
Wed Mar 12 18:57:14 UTC 2003

Robert G. Brown wrote:
> On Wed, 12 Mar 2003, Troy Dawson wrote:
>>Hmm ... sorta.
>>First we start with hydrogen, and strip the electrons off with a flash of 
>>electricity.  So we have a core with just a proton and an neutron.  And send 
>>that off the accelorator.  But then we have to get the neutron off because 
> Not exactly "hydrogen" (usually considered a single proton) but
> deuterium, yeah, exactly.  You NEED the proton, because one can't push a
> neutron with an EM field, because it is, well, neutral!  So you
> piggyback the neutron along with the proton it is nuclear-glued to.
> However, that particular nuclear bond is relatively weak.  You get
> energy OUT of fusion processes because the nuclear bond energy depth
> goes UP as you add more protons and neutrons (they are more tightly
> bound, which is negative energy, so give off more positive energy to get
> into the bound state), up until you make something like iron nuclei, at
> which point it goes the other way (so fission processes from heavier
> nucleii also release energy).  The neutron in deuterium is just barely
> bound.
> So you whonk it into something, and hope to transfer just enough
> transverse relative momentum to break the two apart, while still leaving
> the forward momentum of the separated neutrons dominant and mostly
> unaffected.  Just what and how you whonk it into likely depends on lots
> of things, e.g. the energy range of the neutron beam you are producing.
> Carbon films, metal deposit films, tubes of paraffin or polystyrene
> (neutron rich material)?  I don't keep up to date on the details,
> though.

Wow, you pretty much hit it.  I don't know the energy anymore (I did when I 
had to run it), but it's fairly high.
It's basically a wire mesh, just like a screen door's.  It's not even big, 
about 2 or 3 inches square (well, actually round).  The particles go through 
it, in our first ring, the neutron's go straight, the proton's keep bending.
I saw that thing for the first time, and it was like the first time you find 
out that superman is really just an actor.  It just burst my bubble.

>>But your guess is actually a sorta close.  And I think if it was a gum 
>>wrapper, I would have been just as disapointed.
> "Guess"?  I'm wounded.  Sure, I'm only a theorist, and nuclear physics
> isn't my bag, baby, but I have watched nuclear grad students here spend
> hours vapor depositing some particular metal on a piece of film,
> listened to them lament when a film blows and opens a beam pipe up to
> air (so it has to be pumped out again and can be messed up in other
> ways, which takes a long time and delays their disserations for ANOTHER
> six months as they lose their accelerator slot:-).  I've even spent a
> few nights of my life baby-sitting Duke's backyard accelerator -- just
> enough to know that it wasn't my baby, baby..:-)
> Heck, one of the two things I actually did for TUNL (our nuclear lab)
> back when I was deciding not to be a nuclear physicist was (using a
> computational tool called TRANSPORT -- mandatory caps for fortran IV
> code -- to do geometrical beam optics.  Transport was written by other
> humans, although I did find bugs and fix them and add features.  With it
> I computed the placement and settings of their quadrupole and sextapole
> magnetic lenses and their 90 degree bending magnets for their high
> resolution neutron experimental facility.  I apparently did a good
> enough job that when the leg was finally completed (twenty-plus years
> ago, mind you:-) they turned it on, twiddled the knobs a bit, and found
> their beam right on target with beam currents at or above their
> expectations and requirements.  They were even happy, I think.  I got
> petted on the head and given a bone.

Most people wouldn't know how impressive that is, but I do.  That's pretty 
good.  Every time we start up one, even if the engineers did a good job of 
designing, we usually get at least one manget that's powered backwards. :)

> So I still vaguely remember how accelerators work, although some of the
> details have doubtless gone south over twenty plus years...
> I actually think that the most impressive part of accelerators is the
> accelerators themselves, especially the sources (which are in a lot of
> cases FILLED with all sorts of ad hoc engineering and cleverness and
> even look like "real physics".  By comparison the experimental end is
> pretty boring.  A beam pipe leading right up to the target, a target (a
> plain old chunk of machined metal) suspended by something prosaic (like
> monofilament fishing line).  Various detectors for the crap that comes
> out.  Lots of shielding and indirect controls so that the crap doesn't
> affect your DNA -- much ;-) -- while you are operating all of same.
> Although you guys at fermilab probably have some pretty nifty detectors,
> which can also be amazing from the engineering perspective, especially
> the electronics and data collection stuff, especially compared to what
> was around when I was hanging out near nuclear target:-).  

Yes, we have our share of small little things like you are talking about.  But 
I've had to search and secure the big one's.  They are pretty impressive. 
D0's detector is like walking around the square borg ships .. but without all 
the special affects ... and borgs running around ... ok, well maybe it isn't.

> Of course, you probably have some pretty nifty accelerators, and
> sources, and bending magnets, and magnetic lensing, and control
> apparatus as well...:-)   Money buys cool toys.

Yep, one person's work is another person's toy.  It's funny but you can always 
see the people working here for the money versus the one's working here 
because to them this is the biggest playground they've ever been at.

> Nuclear physics is fun and interesting, no doubt.  Still, I like doing
> what I do now more, and tend to be first author in a string of two
> authors on all of my papers, instead of 83rd author in a string of 126
> authors, more typical of big-lab nuclear papers...:-)
>    rgb

Troy Dawson  dawson at fnal.gov  (630)840-6468
Fermilab  ComputingDivision/OSS  CSI Group

More information about the Yum mailing list