[ RadSafe ] Monte Carlo Modeling

[JPreisig at aol.com]

Dear Radsafe,

     This is from:    jpreisig at aol.com    .


     Greetings Radsafers:

           Hope all is well where you live and work.  A recent posting 
discussed
     using a computational code for solving a health physics problem, 
without
     resorting to Monte Carlo modeling.  Hey, to solve Health Physics 
problems
     try to get an analytical solution with pad and paper.  If that fails, 
try getting
     a solution using a computational code.  If this doesnt work, then 
resort
     to Monte Carlo modeling, if you can.  Not all workplaces allow Health 
Physicists
     the time and/or facilities to do Monte Carlo modeling.

           These days, one can get significant computing memory <3 Gigabyte
     memory and 300 Gigabytes storage memory???> for one or 2 thousand 
     dollars or less.  A Fortran compiler <Compaq/Intel>, for those of us
     not making the leap to other programming languages, can be had for 
     something like 1000 dollars or so.  Such a compiler can include the 
IMSL
     <International Mathematical Statistical Library> software library, 
which is
     very handy to have.  MCNP <Monte Carlo Neutral Particle> program can 
be 
     gotten from RSICC <Radiation Shielding Information Center> at Oak Ridge
     or possibly from Los Alamos.  This computer code can be used to model
     neutron and/or x-ray/gamma problems.  EGS <Electron Gamma Shower>
     can be tracked down on the web and ordered for use in electron 
accelerator
     problems.  CASIM, MARS, LAHET, FLUKA, MCNPX, etc. can be ordered
     from RSICC or other sources.

           Such Monte Carlo codes have a  non-trivial learning curve, and 
it takes a
     while to set up problem geometries and to do other things.  Not all
     workplaces can afford such a devotion of time and/or financial 
resources.
     Still, I think many problems can be set up computationally in a day or 
so.
     And with today's computers, many problems can be run in under a day or 
     less.  Do what you will --- I don't make money on sale of these
     computer codes.

           For neutron data unfolding, look at the computer codes:  
BON4/BON5,
     LOUHI, MAXED, BUNKI, etc.  See Patterson and Thomas's book 
     Accelerator Health Physics <if you can find a copy> for solution 
details.
     See also Cossairt's Course notes on Accelerator Health Physics.

           One can use the repeated structure capability of MCNP to
      model nuclear reactor design problems.

           As for small nuclear reactors soon??? to be available to provide
      power for small towns etc., how cool is that.  The NRC will probably 
have
      to hire more NRC inspectors, if the sale of such reactors takes off.
      I doubt the NRC paperwork for such a reactor will be too complex, 
because
      such a reactor is simpler than a power reactor <PWR/BWR>.  We'll see.

           A local cable TV program keeps suggesting that the power utility 
that
      runs the nuclear reactors in South New Jersey will be looking to
      construct one more new reactor in South Jersey <not Oyster Creek>.
      Again, we'll see what happens.

           Too bad Fermi never had a copy of MCNP to run.  I heard a story 
that
      he used to do Monte Carlo calculations in his head at night, because 
he
      had trouble sleeping.  Too much expresso????!!!!

          The matter/antimatter mass difference in nature is being explained
      via the difference in muon versus anti-muon decay rates <see Fermilab,
      Brookhaven, CERN, etc. websites>.

           You all have a good week!!!!!


            Regards,    Joseph R. <Joe> Preisig, Ph.D.


  </HTML>