TWIST Magnetic Field in Software

Field Types

GEANT field settings

1) ghelix() and mofia tracktoz() are roughly equivalent. ghelx3() is a special case for uniform fields 
   parallel to the Z axis.
2) "BFLD N M" rescales the magnetic field to set Bz at coordinates (0,0,0) to M kGauss.
3) To make a GEANT run that matches a particular data run, first analyze one event of the data run in mofia. 

From the logfile 
grep "magnet_mod: Field at origin is" 
Take the value XXXXX in Telsa and in the GEANT ffcard file set the BFLD card (in kGauss) to (remeber 10kGauss = 1 Tesla):
BFLD 1 XXXXX ...

Processing of MC with MOFIA

Default behavior:

The bank contains:
  Entry   Value
  -----   -----
   1     JFIELD
   2     FIELD_MAX
   3     B_THETA
   4     B_PHI
   5,6,7 DETECT_P0S(3)

1) Namelist unpackmc must be set, i.e. "name global unpackmc=t". (default:unpackmc=f)
2) By default, Namelist IgnoreMcBFLD is set to "false".
   Make sure the following is not set in the kcm file. "name global ignoreMcBFLD=t". 
3) In the case where JFIELD setting of 1 is used, the fieldmap file to be used from CFM is set by 
   the environment variable OPERAMAP i.e. setenv OPERAMAP $CAL_DB/bfld_map.00012 (default)

Over-riding the default settings:

1) Namelist unpackmc must be set prior to processing the MC data. i.e."name global unpackmc=t"
   (default:unpackmc=f)
2) Namelist IgnoreMcBFLD should be set, i.e."name global ignoreMcBFLD=t" (default: ignoreMcBFLD=f)
3) A calibration file, BFLD_SET must be set in the kcm file. This file determinesets whether
   a "uniform field" or a "filed map" should be used.  It also contains info on the B-field value 
   to rescales the magnitude of the magnetic field to at a given u,v,z coordinates (at the NMR location, 
   at the center (0,0,0), ...). e.g. USE BFLD_SET "~e614/e614soft/caldb_ascii/bfld_set.00007"

The bfld_set calibration file contains the following information:
Bswitch: Whether to use a "uniform field" (Bswitch=0) or a "field map" (Bswitch=1).
u,v,z (cm): Position at which the B field will be scaled.
B: The magnitude of the B field to which the field will be scaled to.

The bfld_set calibration file has the following format:
C bfld_set.00007 => Field Map scaled to 1.5T at NMR probe
C
C
C Solenoid field setting
C Bswitch: 0 = uniform field
C          1 = field map
C
C To use uniform field map, then use bswitch=1, and
C put a use statement in the kcm file that points
C bfld map to the right map.
C
C Bswitch   u(cm)     v(cm)     z(cm)     B (T)
    1     -6.200000  16.60000  12.70000   1.5

In the case when Bswitch is set to "1", then a bfld_map calibration file should also be set by using
the environment variable OPERAMAP. e.g. setenv OPERAMAP $CAL_DB/bfld_map.00012
This can be put in the e614.com file.


At the time of writing this document there were 15 different "bfld set" files that can be used
for different data (or with MC if you are so inclined).  The files are:

bfld_set.00001 => Uniform field of 2.T
bfld_set.00002 => Field map scaled to 2.T at Zero
bfld_set.00003 => Field map scaled to 1.998573T at NMR probe as for "2T" 2002 data
bfld_set.00004 => Field Map scaled to 1.95967T at NMR probe as for "1.96T" 2002 data
bfld_set.00005 => Field Map scaled to 2.03850T at NMR probe as for "2.04T" 2002 data
bfld_set.00006 => Field Map scaled to 2.012T at NMR probe as for 2001 engineering run data
bfld_set.00007 => Field Map scaled to 1.5T at NMR probe as for a few runs in 2002 (field not checked)
bfld_set.00008 => Field Map scaled to 1.0T at NMR probe as for a few runs in 2002 and 2004 (field not checked)
bfld_set.00009 => Field Map scaled to 0.5T at NMR probe as for a few runs in 2002 (field not checked)
bfld_set.00010 => Field Map scaled to 1.99863T at NMR probe as for set 14 of 2003 data
bfld_set.00011 => Field Map scaled to 1.99862T at NMR probe as for sets 12,13,15 of 2003 data
bfld_set.00012 => Field Map scaled to 1.998749T at NMR probe as for sets 30,31 of 2004 data
bfld_set.00013 => Field Map scaled to 1.998739T at NMR probe as for sets 32,33,34  of 2004 data
bfld_set.00014 => Field Map scaled to 1.998734T at NMR probe as for sets 35 to 39 of 2004 data
bfld_set.00015 => Uniform field of 0.T as for alignment data, or any other field off data



Processing of Data with MOFIA
Two calibration files are used.  The two calibration files are:
 
bfld_set: Used to sets the B field value to which the magnitude of the field in 
          the "field map" is scaled to at a particular position.  For data the bfld_set is
          chosen automatically using the CFM database.
bfld_map: The field map to be used. The default values in MOFIA are chosen to use the best available 
          OPERA map $CAL_DB/field_map.00012 rescaled to the 2002 average value of the DS NMR probe.



NOTE on MOFIA tracking: The same tracktoz() code is used for uniform and OPERA
(non-uniform) magnetic fields. The tracktoz() algorithm with "k=0" (no energy
loss correction) is supposed to be identical to GEANT's ghelix() and ghelx3().
For non-uniform fields, the step size inside tracktoz() is reduced to limit the
change in the magnetic field to less than 1E-10 Tesla. This parameter can be
adjusted to increase the integration accuracy at the expense of round-off error
and CPU consumption (see trackswim_mod::tracktosk()). For non-uniform field
tracking, tracktoz() uses a modified stopping algorithm to stop the track within
100 um from the requested Z. With existing field maps (almost uniform) and
normal tracks, it stops much closer than 100um (but how much closer?) 
(see trackswim_mod::tracktozks()).



Last updated Apr.14,2005. Mina Nozar.