Different Result with different version

Dear lammps user,

I have run a simulation using 3 different versions of lammps: 11 July 2013, 11 November 2013 and 14 March 2015 (windows executable). The input script for all version is exactly the same, but I got a totally different results between the 2013 version and the 2015 version. Moreover, the results from the 2013 version are more logical than from 2015 version. This different results confuse me, and I wonder whether this is problems with lammps or not. In addition, I attach my input file and its data file.Thank you

Best regards,

The result that I check is the water molecule file, the water molecules in the feed side should decrease over time. The results from 2013 version gave this phenomena, while the results from 2015 version gave a reversed phenomena where the water molecules in the feed side are increasing.

Data-SaltwaterGraph3_AfterRun.txt (338 KB)

InputCver12.txt (9.98 KB)

Who can say? You have a 250-line complicated input

script that you run for 50K timesteps. Do the 2 simulations

drift apart slowly. Is the process you are monitoring

randomly stochastic? If you do 10 runs with different

initial conditions with each, do you get different results

on average? Can you simply the script dramatically

to figure out what is different between the two?

No one is going to do that for you.