Hi All
I am using LAMMPS to calculate the critical buckling load for double walled carbon nanotubes. In order to validate my LAMMPS code I used it to calculate the critical load values for a few double walled carbon nanotubes and compared the results with the results published in journal articles and found my values are totally different. I am new to Molecular Dynamics simulation and have no clue where I went wrong. In some of the literatures an older version of LAMMPS is used (some version in year 2010), and when I compare these results with my simulation I got critical load values underestimated by ~15%. I am using the version on LAMMPS published on 7th of April 2011. My input file and data files are given below. I would be very obliged if someone can answer the following queries:
1. Is it expected that in a newer version of LAMMPS results will be different.
2. Is there anything wrong in my input file.
I briefly summarize the problem:
Inner tube type = armchair, chirality = 10
Outer tube type = armchair, chirality = 10
Total length = 72.3001 A
Free length = 60.0458 A (Length excluding the constrained atoms)
Type of potential used = AIREBO
Velocity of pushing the nanotube = 0.05 A/ps
Simulation temperature = 300 K
Procedure followed:
step 1: I am minimizing the system
step 2: Then I perform nvt analysis to compress the nanotube and record the force displacement
========LAMMPS INPUT FILE===============================================
1. Is it expected that in a newer version of LAMMPS results will be different.
No, though there have been recent changes to the AIREBO potential.
I suggest you contact the authors, get their scripts and try them out.
I note that calculating most things to within 15% with MD is pretty good.
Steve
Hi Steve/All
Thanks for your reply, I will contact the authors regarding AIREBO potential. But, I was wondering if I my input file is wrong correct, may be my approach to the simulation is not correct. Is there any mistake you observe in mu LAMMPS input file. I am increasing the temperature using velocity scaling then I am minimizing the initial double walled carbon nanotube assembly. After that I run the simulation for a number of steps, and in each step the system is displaced , and a nose hoover thermostat to keep the temperature constant. While assigning the temperature by velocity scaling I choose a random number and have also tried with
different random numbers but the total energy does not change by much.
1.Do you observe any mistake in the procedure that I employed to investigate the buckling of nanotube.
2.I would also like to mention that I use a bigger simulation box to contain the nanotube for my simulation, to my opinion it should not have any effect on the result, do you feel the same.
I will really appreciate your kind cooperation, and sorry for troubling with trivial questions.
Thanking You
With Regards
Amar Nath Roy Chowdhury
Graduate Research Student
Department of Civil Engineering
National University of Singapore
h.p. 8524 - 1880
I meant check with the authors of the paper that used LAMMPS. They should
be able to give you their LAMMPS script.
Sorry, I don't typically have time to look at complicated input scripts. Maybe
someone else on the list can offer you advice.
Steve
Hi Steve
Thank you very much for your valuable time.
With Regards
Amar Nath Roy Chowdhury
Graduate Research Student
Department of Civil Engineering
National University of Singapore
h.p. 8524 - 1880