[Fwd: [ROOTCVS] geant3 brun]

The VMC system in the CVS head contains now an important contribution from Andrea Fontana
and Alberto Rotondi presented at the ROOT workshop. See the news below.

Rene Brun

attached mail follows:

CVSROOT:	/user/cvs
Module name:	geant3
Changes by:	brun	07/05/18 10:44:16

Modified files:
	TGeant3        : TGeant3.cxx TGeant3.h geant3LinkDef.h 
	comad          : gcomad.F 
	erdecks        : erland.F eustep.F 
	geant321       : gccuts.inc 
Added files:
	geant321       : gcmore.inc gtmore.inc 

Log message:
A major update of GEANTE by Andrea Fontana and Alberto Rotondi

1. update of the Coulomb multiple scattering parametrization;
2. update of the straggling of energy loss for thin materials;
3. new options to extrapolate the track parameters to the point of closest approach to a point or to a wire (straight line).

Details on the physical motivation behind this work can be found in our report for the Panda Collaboration, available at:

http://www.pv.infn.it/~fontana/tracking.pdf

Feel free to contact us for questions and discussions about these topics by using the following email addresses:

alberto.rotondi_at_pv.infn.it
andrea.fontana_at_pv.infn.it

---

List of changes in the fortran and C++ routines of the geant3
VMC directory:



- gcmore.inc

gtmore.inc
geant3LinkDef.h
gcomad.F

Added a new common that contains all the new variables:
COMMON/GCMORE/GCALPHA,ICLOSE,PFINAL(3),DSTRT,WIRE1(3),WIRE2(3),
+              P1(3),P2(3),P3(3),CLENG(3)

input to ERLAND:
GCALPHA: energy cut parameter for energy loss fluctuations

input to EUSTEP:
ICLOSE: = 1 the use of the common is enabled for the closest
approach to a point PFINAL(3)
= 2 the use of the common is enabled for the closest
approach to a wire of extremes WIRE1(3) and WIRE2(3)
= 0 the common is empty and disabled
PFINAL(3): assigned point
DSTRT: assigned distance between initial point in ERTRAK
and PFINAL along straight line (currently noy used)
WIRE1(3): first point of a wire
WIRE2(3): second point of a wire

output from EUSTEP:
P1(3): point previous to the point of closest approach to
PFINAL() or wire
P2(3): point of closest approach to PFINAL() or wire
P3(3): point next to the point of closest approach to
PFINAL() or wire
CLENG(3): track length to the previous 3 points

Important note: the calculated points of closest approach are
depending on the GEANE steps. For calculating the true point
of closest approach the last 3 points of the extrapolation, i.e.
the previous to closest, the closest and the next to closest are
returned to the user. Different algorithms can be implemented, but
we decided to leave this to the users in the C++ interface to GEANE.



- ermcsc.F

new expression for the variance of the Coulomb multiple scattering
according to Fruhwirth and Regler, NIM A 456 (2001) 369



- ertrch.F

added DESTEP in the calling string of ERLAND for calculation with
Urban model. Added and saved previous step PRSTEP.



- erland.F

added new calculation for sigma of straggling in energy loss
to include in Geane the Urban/Landau approximation, as explained
in the Geant manual and related papers.
The model parametrization can be controlled with a user variable (GCALPHA)
in the new GCMORE common block: 1 is for old gaussian model valid
for dense materials, other values (see the report) are for gaseous
materials.



- eustep.F

added the calculation to the distance of closest approach to a point
or to a wire.



- TGeant3.h

- TGeant3.cxx

added the possibility to define user cuts (already present in the gccuts
struct but not in the TGeant3::SetCUTS method) and to define the new
variables of the GCMORE common with two new methods SetECut() and
SetClose().
Added new method InitGEANE() to initialize GEANE to the old behaviour
(default) for backward compatibility. Only the multiple scattering has
been updated to a more correct formula.
Corrected a typo in the call to the routine Trscsd().