Hi
Well fix NVT/NPT damping parameters are used to specify how rapidly pressure and temperature should be relaxed. And so goes following example e.g in fix NVT
The Tdamp parameter is specified in time units and determines how rapidly the temperature is relaxed. For example, a value of 100.0 means to relax the temperature in a timespan of (roughly) 100 time units* (tau or fmsec or psec - see the units command).
But when I read run command section it gives following example
fix 1 all nvt 200.0 300.0 1.0
run 1000 start 0 stop 10000
run 1000 start 0 stop 10000
...
run 1000 start 0 stop 10000
The NVT fix ramps the target temperature from 200.0 to 300.0 during a run. If the run commands did not have the start/stop keywords (just “run 1000”), then the temperature would ramp from 200.0 to 300.0 during the 1000 steps of each run. With the start/stop keywords, the ramping takes place over the 10000 steps of all runs together
Here it says temperature would ramp in 10000 steps of total run, but the damping parameter is 1.0 time unit and so according to * above, there’s a contradiction?
Regards
Chetan
Hi, all. A general MD question. I've always sort of taken it as accepted
wisdom that using thermostats, especially Langevin thermostats, allow one
to use larger timesteps than can be used in NVE integration. My vague
notion of why this is is that the random noise forces somehow "stabilize"
the system. But some non-MD people in my group who know a lot more about
numerical analysis than I do say that this cannot be. They are interested
because they also solve equations with a Langevin-like noise term.
Does anyone know why Langevin allows larger timesteps, or a book/article
that examines the issue in detail? For example, is it connected
with integration scheme? Someone told me that the Langevin term can
affect the order-of-accuracy of time integration.
Thanks,
Rob
Ramping and damping are 2 different things. If your start and stop T are the
same, the damping parameter determines what kind of (hopefully) small
oscillations
you will get in T. Try it and plot T and you should see some differences
as you change the damping parameter. Ramping just sets the target temperature
at any point during the run.
Steve
My view would be that a Langevin thermostat lets you take a bit
longer timestep than NVE, but by hiding the energy error that
the long timestep is creating. So the thermostat is just hiding
a problem.
Steve
Take a look at the “Molecular Modeling and Simulation” book by Tamar Schlick. See page 435. She argues that Langevin thermostatting is a good way to go because it allows you to take much longer timesteps.
Paul
Hi Steve
Thanks for quick reply
Well I actually used the word damp, since it's what's used in lammps docs
such as Tdamp/Pdamp, otherwise I mean ramping only So, if this meaning, we
look into the two examples I have given, the meanings contradict each other,
don't they?
In the first case we mean that Tdamp is a time during which temperature gets
relaxed, so if we interpret second case according to this, then it should be
ramped from 200 to 300 in 1 timeunit and not 10000 steps as document is
saying?(here I am assuming that timestep takes default value of 1 femtosec)
Thanks
Chetan
2008/6/30 Chetan Mahajan <cmahajan@…670…>:
Hi Steve
Thanks for quick reply
Well I actually used the word damp, since it’s what’s used in lammps docs
such as Tdamp/Pdamp, otherwise I mean ramping only So, if this meaning, we
look into the two examples I have given, the meanings contradict each other,
don’t they?
In the first case we mean that Tdamp is a time during which temperature gets
relaxed, so if we interpret second case according to this, then it should be
ramped from 200 to 300 in 1 timeunit and not 10000 steps as document is
saying?(here I am assuming that timestep takes default value of 1 femtosec)
You’re confusing the ramping and relaxation processes.
The ramping of the temperature from the beginning to the end of the run is a strictly linear process. Using the “start” and “stop” keywords here makes it one continuous process starting at 200 and going to 300 instead of (presumably) 10 separate ramps each starting at 200 and going to 300.
The relaxation (damping) process explains how quickly oscillations get smoothed out by the NVT integrator (the smaller the constant, the “sharper” the temperature corrections). The value of this parameter has nothing to do with how quickly the temperature adjusts from 200 to 300–only with how quickly deviations from the linear temperature progression are “wiped out.”
There’s no contradiction.
–AEI
To add to what Ahmed said, if you choose 200 and 300 as
your start/stop Ts, and run for 1M steps, then it will take 1M steps
for the T of the system to go from 200 to 300.
The damp parameter has to do with how long it takes
the system to go from its current T to the target. If you stated
that simulation with the T at 150, then it would take about "damp" time
for the system to go to T = 200. Then it would slowly climb to 300
for the remaining 1M steps.
Steve