Bug Fixes for Elements

• The CWIGGLER element had an error in the ordering of rotation and displacement removal operations. Simulations that contained both rotational and displacement errors for the same CWIGGLER element should be repeated.
• There was an error in an exponent for synchrotron radiation expressions, resulting in incorrect dependence of the quantum fluctuations on transverse position in a dipole, quadrupole, or sextupole. Tests with the APS ring showed small effects on results even for orbits with several percent momentum deviation.
• The correction to the response matrix when there was dispersion at the corrector was being added with the wrong sign. This would affect response matrix output in the bending plane. This is a minor effect in large rings but quite significant in small rings. Reported by Michael Abo-Bakr (BESSY).
• When particles were lost on an ECOL element with non-zero length, the loss locations were recorded incorrectly. Reported by Feng Zhou (SLAC).
• The FTRFMODE element, which allows simulating multiple tranverse rf modes, had an unitialized variable that would result in incorrect accumulation of voltages over many passes.
• Previously, the transverse space-charge element SCMULT could be inserted by hand into the lattice, but it wouldn't actually run. Now, the user is required to provide a insert_sceffects command to use SCMULT.
• The longitudinal space charge algorithm (LSCDRIFT and RFCW) would give give invalid results if the number of bins was odd. This odd choice of bin number is no longer allowed.