From: Alfredo Ferrari (alfredo.ferrari@cern.ch)
Date: Wed Oct 11 2006 - 09:29:44 CEST
Hi all
a respin version of Fluka2006.3 is available since monday on the Fluka web
site.
It contains only some (rare) crash fix, a feature enhancement requested by
Alice, an increase of the max. number of allowed regions to 20000, and
(the only item with physics impact) the rolling back of a new feature
in Dpmjet-3 which we believe should be better tested before being exposed
to the users.
In particular:
a) the 232-Th issue (crash while generating beta+/- spectra) is solved.
It turned out to be a bug in the Brookhaven database for beta
emissions, where some of the end-points are incorrectly listed as
"0" energy. For all users, please if you find in your .err files
messages like
*** Slgmbt: A xxx Z yyy E_end kkkkkk
report them to me or to this list. I fixed the 232-Th specific one but
I suspect there could a few other similar quirks in the database. I
protected the code from crashing if another of these issues is met at
run time: in this case you get the above message and the emission is
simply ignored. However in the long term I would like to solve all of
them fixing the database, hence the need for reporting them.
b) Primary ionizations (if activated) are now recorded together with the
position where they occurred (Alice folks, please let me know if it
works as advertised)
c) The maximum number of region prototypes is now 20000 (still to be made
active thru a GLOBAL card). As usual, lattices can be used to replicate
identical regions as many times as you like. replicas are not subject
to this limit
d) A rare "array out of boundary" crash when the new high energy model
generator is activated is fixed
e) The threshold region for photon -> mu+ mu- is now better protected,
hence rare crashes when an interaction was called below threshold
because of the finite grid of xsec tabulations should be solved
(SLAC guys, please check)
f) A possible issue with FLUGG when photonuclear interactions are
activated has been solved
g) A change in the p_t distribution of hadrons for DPMJET-3 events has
been momentarily withdrawn. The change is beneficial when comparing to
exp. p_t data, however it has some minor unwanted effect on particle
multiplicity. It requires some global retuning of hadronization
parameters before being inserted again.
I would like to issue a warning which can be of interest for some users.
When BOXes or other deprecated shapes (ie ARB's) are used in the geometry
it is mandatory to write the geometry in the high precision fixed format
(the one with 3 numbers each 22 field long on each line for a body), or
in the new name based free format PROVIDING ALL DIGITS up to the 16th one.
This because, ie for BOXes, faces must be EXACTLY perpendicular to allow
the code run properly. Failure to do so, or geometry inputs given in the
standard "low" precision format (six numbers per line) will always result
in geometry errors. Those errors will be dramatically more frequent when
particles come at grazing incidence to one of the face and can result in
nonsense results. As usual, the safest line of action is not to use the
"deprecated" shapes (ARB BOX RAW WED).
Ciao
Alfredo
-- +----------------------------------------------------------------------------+ | Alfredo Ferrari || Tel.: +41.22.767.6119 | | CERN-AB || Fax.: +41.22.767.7555 | | 1211 Geneva 23 || e-mail: Alfredo.Ferrari@cern.ch | | Switzerland || Alfredo.Ferrari@mi.infn.it | +----------------------------------------------------------------------------+
This archive was generated by hypermail 2.1.6 : Wed Oct 11 2006 - 10:19:02 CEST