Radial distribution function and the Verlet integration timestep (dt)

Dear lammps user,

My lammps version is 16 March 2018.

I am simulating the radial distribution function (RDF) for some fluid system. According to my best knowledge, the RDF does not depend on the timestep (dt) used in Verlet integration. At some point, I had to check this for several timesteps. I run my lammps script for the same system at different timesteps and then plotted the RDFs at the corresponding timesteps. I noticed that RDF changes dramatically when choosing a very small timestep (dt=0.000001) as it is clear from the attached figure.

My first guess is that this could be due to the precision used in lammps.

If so, could you please tell me how to run my script using higher precision ?. I could not find a direct command about it.

Below is my script. Please, for my work I have been directed by our group not to consider time averaging RDF at this moment. I just take take the average over 100 runs. Also, I have been directed to follow the equilibration scheme mentioned in the script.

rdf.png

Dear lammps user,

My lammps version is 16 March 2018.

I am simulating the radial distribution function (RDF) for some fluid system. According to my best knowledge, the RDF does not depend on the timestep (dt) used in Verlet integration. At some point, I had to check this for several timesteps. I run my lammps script for the same system at different timesteps and then plotted the RDFs at the corresponding timesteps. I noticed that RDF changes dramatically when choosing a very small timestep (dt=0.000001) as it is clear from the attached figure.

My first guess is that this could be due to the precision used in lammps.

your guess is wrong.

If so, could you please tell me how to run my script using higher precision ?. I could not find a direct command about it.

you cannot and you need not.

Below is my script. Please, for my work I have been directed by our group not to consider time averaging RDF at this moment. I just take take the average over 100 runs. Also, I have been directed to follow the equilibration scheme mentioned in the script.

well, you should discuss with your group then. I suspect they want you to figure out by yourself, why you see the difference, so please think about this.

you should think about the following: did the people in your group specifically ask you to do the exact same number of steps for equilibration or do the equilibration for the same amount of time?
there is a difference and that matters here.

axel.