A metallic glass has been taken for quenching from 300 to 2300 and then followed by quenching...and the model was simulated for the same in LAMMPS. Now can anybody of you please describe the best possible and easy way to compute the stress strain behaviour (sigma-e curve)?
Dear Amit:
the easiest way for obtaining a stress-strain curve of your materials by using LAMMPS is by performing an explicit deformation. To be more precise, the necessary steps are:
1. let your system equilibrate under its original box dimensions. Calculate the average stress along the direction of interest, e.g. the xx component of the stress tensor.
2. deform your box along the desired direction by applying a specific strain, e.g. 0.1 % along the xx direction. This can be accomplished by the "change_box" command of LAMMPS.
3. rescale the coordinates of all atoms in an affine way. This can be accompished by providing the "remap" keyword to the "change_box" command above.
4. run a new simulation of the deformed configuration and estimate the average xx component of the stress tensor.
5. apply a new deformation and repeat steps 2-4.
I should note that, depending on your material, the time needed for stress to equilibrate may be quit long. Moreover, in general, stress from molecular simulations exhibits extreme fluctuations, so you have to be careful with averaging.
Best wishes for success!,
Georgios
Dear Amit:
the easiest way for obtaining a stress-strain curve of your materials by using LAMMPS is by performing an explicit deformation. To be more precise, the necessary steps are:
1. let your system equilibrate under its original box dimensions. Calculate the average stress along the direction of interest, e.g. the xx component of the stress tensor.
2. deform your box along the desired direction by applying a specific strain, e.g. 0.1 % along the xx direction. This can be accomplished by the "change_box" command of LAMMPS.
3. rescale the coordinates of all atoms in an affine way. This can be accompished by providing the "remap" keyword to the "change_box" command above.
4. run a new simulation of the deformed configuration and estimate the average xx component of the stress tensor.
5. apply a new deformation and repeat steps 2-4.
I should note that, depending on your material, the time needed for stress to equilibrate may be quit long. Moreover, in general, stress from molecular simulations exhibits extreme fluctuations, so you have to be careful with averaging.
Best wishes for success!,
Georgios
Hi Amit,
You can use fix deform command of LAMMPS, which allows you to generate tensile or shear stress-strain curves. You can have a look at our paper that has recently been published.
http://pubs.rsc.org/en/content/articlelanding/2016/sm/c5sm02788h#!divAbstract
baris.
Sir (for all respected who replied above)
Actually the problem is such as "firstly i am processing the quenching the specimen afterwards i want to calculate the properties (by stress strain curve). So in a single coding...how to implement both the processes...???
You need to do them stepwise. First do your quenching sequence and equilibrate the structure then use fix deform to get the stess strain curve. You should also check the effect of strain rates on the stress-strain behavior you are getting.
Dear Mohammad Rafat Sir,
Thanx for the answer but After quenching, the fix deform command is not accepting. It is showing error.
Hi
To initialize the simulation, a relaxation state is defined and the atoms are left freely
with a random velocity using NVT. The temperature is going to be set around 0K during all the simulation and the volume and number of atoms too. Different simulations with
increasing strain rates will be performed to confirm that the Young Modulus does not depend on this variable.
Once the velocity is set in the upper layers and the lower layers are locked, the
simulation will be executed using NVT which performs time integration on non-Hamiltonian quations of motion, thus, generates positions and velocities sampled from canonical ensembles. On the middle atoms the position and velocity will be updated using NVE
integration which keeps the volum and the energy of the system constant. This creates a
system trajectory consistent with the microcanonical ensemble.
Please contact me if you need to have the input file by : [email protected]
Here I have demonstrated a procedure to calculate the stress of a graphene sample in LAMMPS: https://www.linkedin.com/pulse/how-perform-uniaxial-tensile-tests-graphene-sample-lammps-dewapriya/. Two MATLAB scripts are also provided to create the input files for LAMMPS and to extract data from the LAMMPS output file.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering