Garfield news
2008
- January: Excitation and ionisation rates from Magboltz are transferred to Garfield.
The Townsend coefficients can be adjusted on the basis of user-provided Penning transfers
probabilities.
- February: Ansys interface completed with material map, weighting fields and examples.
2007
- All along the year: Introduction of a SRIM derived model with an energy loss fluctuation
distribution chosen according to the thickness of the material.
- January: Algebraic formula parsing made more efficient (1500 times faster !).
- March: Field map mesh windows have been moved from the FIELD-MAP command to become
a command of its own.
This command works on all field maps.
2006
There have been numerous minor corrections over 2006.
Main developments to be noted:
- introduction of a version running in the BOINC framework;
- interface with the SRIM program;
- weighting fields for truely periodic read-out structures;
- update of the finite-element interfaces: new elements, new formats;
- correction of the sign of the E×B term in the Langevin velocity.
2005
As of January 2005, CERN assumes responsability for Garfield maintenance.
I plan to spend the first half of 2005 on recovering the backlog in known
problems: interfaces with finite element programs that no longer work, the
longstanding error in the Monte-Carlo treatment of diffusion, updates of
gas cross sections etc.
- 21/9: Magboltz version 7.1 interfaced with Garfield version 9.
- 20/4: Added a CGI script which displays the
dates
of the latest modifications in the Garfield source files.
- 8/4: New cross section
display page.
This pages contains the cross sections used by Magboltz 7.1, which is not yet
interfaced with Garfield.
The new page can show cross sections for several gases at the time,
thus simplifying comparisons.
It runs as a CGI script, not anymore as JavaScript, and should be
more browser tolerant.
- 4/4: Switching between 2 of the 3 modes of Magboltz calculations (E parallel
with B and the general case) occurred, as a result of a typing error in Magboltz,
at an angle of 1 degree rather than at 0.001 degree.
This has been corrected.
- 3/4: B fields entered via the FIELD-MAP command in the cell section
were not recognised when computing Magboltz transport tables.
This has been corrected.
- 13/3: SAVE-FIELD-MAP now saves the COMPUTE-ELECTRIC-FIELD flag.
- 13/3: Increase of the maximum number of E-B angles to 20.
- 12/3: Increase of the maximum field map size to 250,000 elements.
- 12/3: Correction in the calculation of electron pulses.
As a result of a typing error, the y-component of the
v.Ew
term in the electron induced signal was effectively missing.
This only affects electron signals, not ion signals.
- 11/3: Derivatives and Jacobian of shape functions for 2nd order triangles
and tetrahedrons.
This is needed for the FEMLAB interfaces since FEMLAB, like Tosca, does not
write out electric fields separately for each element to which a node belongs.
Contrary to the potential, the electric field at a point on a boundary
between materials changes in value depending on whether one approaches the
point from within one medium or from within the other.
Computing the electric field on the fly is a perfectly reasonable approach
since this requires only a minor effort while it significantly reduces the
storage space required by the field map.
- 8/3: FEMLAB 2D interface
- 25/2: FEMLAB 3D interface
- 16/1: Mac OS/X compilation
- 15/1: update of the Tosca interface: derivatives and Jacobian for hexahedrons and
pentahedrons
Last updated on 15/3/08.