reported temperature from fix pimd

Hello all:
I’m attempting to run a PIMD test calculation for water. Using the attached code and the magic of lammps multi-partition, all runs fine and the resulting trajectories all seem reasonable. Unfortunately, the temperature or the various replicas (beads) vary wildly, only approaching the requested temperature for the 1st bead and monotonically decreases for each subsequent bead.

Am I missing something in these simulations or is there a bug in this fix? My control and data files are attached.

Any advice is much appreciated,
Thanks!

–Mike

==temperature profile from log files
#log avg_temp stdev

qspcfwbox.cmd.298K.equil.lammps.log.0 306.633 9.71081
qspcfwbox.cmd.298K.equil.lammps.log.1 181.225 5.85922
qspcfwbox.cmd.298K.equil.lammps.log.10 8.33912 0.269676
qspcfwbox.cmd.298K.equil.lammps.log.11 6.17998 0.198728
qspcfwbox.cmd.298K.equil.lammps.log.12 6.18013 0.199805
qspcfwbox.cmd.298K.equil.lammps.log.13 4.91128 0.160577
qspcfwbox.cmd.298K.equil.lammps.log.14 4.9123 0.160077
qspcfwbox.cmd.298K.equil.lammps.log.15 4.12153 0.135043
qspcfwbox.cmd.298K.equil.lammps.log.16 4.11734 0.133088
qspcfwbox.cmd.298K.equil.lammps.log.17 3.62027 0.118836
qspcfwbox.cmd.298K.equil.lammps.log.18 3.62045 0.118452
qspcfwbox.cmd.298K.equil.lammps.log.19 3.31409 0.107186
qspcfwbox.cmd.298K.equil.lammps.log.2 181.173 5.91668
qspcfwbox.cmd.298K.equil.lammps.log.20 3.31271 0.107848
qspcfwbox.cmd.298K.equil.lammps.log.21 3.14375 0.102361
qspcfwbox.cmd.298K.equil.lammps.log.22 3.14444 0.10382
qspcfwbox.cmd.298K.equil.lammps.log.23 3.0902 0.101436
qspcfwbox.cmd.298K.equil.lammps.log.3 46.1212 1.50023
qspcfwbox.cmd.298K.equil.lammps.log.4 46.1122 1.49157
qspcfwbox.cmd.298K.equil.lammps.log.5 21.1089 0.689872
qspcfwbox.cmd.298K.equil.lammps.log.6 21.1079 0.692773
qspcfwbox.cmd.298K.equil.lammps.log.7 12.3635 0.399243
qspcfwbox.cmd.298K.equil.lammps.log.8 12.3619 0.398653
qspcfwbox.cmd.298K.equil.lammps.log.9 8.33873 0.272006

in.qspcfwbox.equil (2.31 KB)

data.qspcfwbox (81.1 KB)