Re: [lammps-users] r-RESPA for pressure calculation
Axel Kohlmeyer <akohlmey@...24...>
Mon, 16 Apr 2018 12:21:40 -0400
you also forgot to mention the LAMMPS version you are using and
whether you get the same issue, when you run r-RESPA with 0.25fs time
step and only one level or a multiplier of 1 (so that all levels run
with the same size timestep).
finally, you mention a multi-component system so, you should make
tests with a homogeneous system first.
On Mon, Apr 16, 2018 at 9:40 AM, Pengyu Huang <hughh.py@...24...> wrote:
> Just forgot to mention in my previous email, the outermost timestep is 2 fs
> in r-RESPA, so that the bond and angle forces are calculated every 0.25 fs,
> pair forces at 1 fs and kspace at 2 fs.
> On Mon, Apr 16, 2018 at 11:37 PM, Pengyu Huang <hughh.py@...24...> wrote:
>> Dear LAMMPS developers,
>> I am running the simulation for a system with flexible spc and flexible
>> co2, and silica surfaces. I found the different pressures in the bulk water
>> region between the Verlet integrator and r-RESPA integrator. I used a
>> timestep of 0.25 fs for Verlet and the following set up for r-RESPA:
>> run_style respa 3 4 2 bond 1 angle 1 dihedral 1 improper 1 pair 2 kspace
>> so that the bond and angle forces are calculated every 0.25 fs, pair
>> forces at 1 fs and kspace at 2 fs. For kspace calculation (pair_style
>> lj/cut/coul/long), pppm with accuracy 1e-6 was used.
>> The pressure in bulk water region (far away from the interfaces) with
>> Verlet is around 12 MPa, while one with r-RESPA gives 115 MPa, which is 10x
>> larger. The pressures in bulk co2 region from the two systems are
>> statistically similar. I am aware when using large timestep for flexible
>> water with Verlet integrator, the calculation of pressure would be
>> incorrect, as I have tested something similar in
>> http://lammps.sandia.gov/threads/msg66029.html (Although I got correct
>> pressures for rigid spc at 1 fs...). So, would this means that the way of
>> the pressure calculation in r-RESPA (at a larger timestep?) cause the
>> Thank you for your time.
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> lammps-users mailing list
Dr. Axel Kohlmeyer akohlmey@...24... http://goo.gl/1wk0
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.