CCL: chilling effect for andersen thermostat with shake algorithm

From: "conor douglas parks" <coparks2012-x-gmail.com>
Subject: CCL: chilling effect for andersen thermostat with shake algorithm
Date: Fri, 8 May 2015 23:40:15 -0400

 Sent to CCL by: "conor douglas parks" [coparks2012---gmail.com]
 Hello,
 I have recently ran some simulations of a bulk solution of glycine in water
 using a custom code. I
 have noticed that when constraints are present, the andersen thermostat will
 equilibrate my liquid
 at about 20 degrees below the set point temperature. I was using the shake
 algorithm found in
 lammps to constrain all the hydrogen bonds in water and glycine. I did some
 digging around, and
 found this referred potentially as the chilling effect
 https://accelrys.hosted.jivesoftware.com/message/9669
 has anyone else experienced this with the andersen thermostat? If so, is it
 possible to correct
 outside of switching thermostats? Is this type of effect also observed when
 using a langevin
 thermostat?