How to apply stress to a box in LAMMPS and draw a curve of strain with time?
Hi,
I highly recommend you go over all of Mark Tschopp’s LAMMPS Tutorials (Link below). You can find your answer as you learn more about LAMMPS.
Also, here’s a friendly reminder: performing LAMMPS simulations to get accurate results is dependent on many parameters, and you need to read a good number of papers on your material and system to understand them.
1 Like
I have gone through them.I want to study creep in LAMMPS and in Literature it mentions “A constant stress was applied oer 30 ps.”. How to do that?
There is no way you have gone through a tutorial and still have such a question 
In tutorials it tells about fix deform where we have to mention strain rate. But I want to apply stress and understand how strain varies with time. For that I shall use fix aveforce or fix setforce which one?
Without details about your geometry and system, nobody will be able to help you. Depending on whether your system is fully periodical in all directions or if you have a block of solid in vacuum, the solution will differ.
Also note that if you don’t make the minimum effort to explain your problem, nobody will want to help you.
I am taking a Silica-Cr-Au system. The silica is substrate. I have deposited firstly Cr on SiO2 and then on Cr, I have deposited Au.It is fully periodic in all directions.Now I want to study creep behavioir of this system. I want to understand how to apply constant stress to the system so that I can study how the applied stress causes strain in system with respect to time. Basically I want to understand time dependent strain behavioir to study mechanism of primary, secondary, tertiary creep. For this I need output of strain vs time at a constant stress. How shall I apply stress to study creep behavior using LAMMPS? Which command shall I use to apply constant stress?
I suppose using fix NPT would do the job, but I’ve never seen it used for that purpose, have a look at one of the previous post were this was discussed.
What you are looking for is a one-to-one tutoring service, which is not the purpose of this forum, nor the recommended way to learn molecular modelling.
If nobody has the right expertise in your institution, your second best bet is to study and figure it out yourself. Instead of waiting for someone to set up a simulation for you (which won’t happen), I suggest you open the LAMMPS manual, preferably the PDF version which you find with the source code, and head to section 8.3.5 Calculate elastic constants, following the references to various examples on how determine stress-strain curves. Use the change_box command to stress your sample and then you are good to simulate it over time, as requested.
I am very happy to talk about Radiohead’s hit song Creep any time, with no preparation. If you are talking about a scientific study of creep, however, that does take a bit of homework.
Ideally you would have access to a mentor who can help you one on one, as others here have said – but if not, you should start by searching the literature and attempting to replicate prior papers that have studied similar phenomena with molecular dynamics.
Please note that if you cannot find prior MD papers similar to your proposed study, this is also meaningful information. Molecular dynamics is a scientific technique, and like any other scientific technique, it has restrictive assumptions and limited domains of application. You cannot study the Sun with a microscope, a bacterium with a telescope, or steel’s Young modulus with a kitchen scale and a 15cm ruler.
Similarly, molecular dynamics is usually restricted to phenomena where simulating a nanometer-sized region over nanoseconds of time, at atomistic (but not electronic) detail, will give you useful results. If there are no prior MD studies similar to your proposal, this is a strong hint that your phenomenon cannot be usefully modelled over MD length and time scales. If you proceed, you do so at your own risk, and if you do so without an experienced supervisor you have a very high chance of failing.