Dear Colleague,

A couple of bugs have been spotted in the Dissipative Particle Dynamics (DPD) 
code of DL_MESO version 2.6.

When the DPD thermostat with DPD Velocity Verlet integration is used in the 
version of the code without OpenMP, two arrays for determining particle pairs 
when recalculating dissipative forces are incorrectly allocated. This can lead 
to segmentation faults and prevent the code from running. Many thanks are due to 
Tom Rodgers, University of Manchester, for spotting this bug.

To fix this bug, change line 649 in field_module.f90 to:

  ALLOCATE (lpart (mxpcell, 27), tload (27), pairs (3, pairbuf), pairdata (4, pairbuf), STAT=fail)

(No change is necessary for the OpenMP version, field_module_omp.f90.)

A bug has also been found for the Lowe-Andersen thermostat: the pairwise mass 
for velocity corrections between pairs of particles on different processors is 
currently being calculated incorrectly. To fix this bug, change line 226 in 
integrate_lowe.f90 to:

  rmassij = rmassi + rmassj

and insert the following line after lines 226 and 255:

  massij = 1.0_dp / rmassij

We strongly recommend that all registered users of DL_MESO version 2.6 should 
either make the above code changes or re-download the package using the 
instructions received on registration to get the corrected version for the 
software. Re-registration is not required and the same password is used for 
decryption.

Michael Seaton
2016-01-25