Forum Navigation
You need to log in to create posts and topics.

Strain analysis

Dear OVITO experts,

As I remmebr in the previous lammps versions like 3.0.0 or 2.9.0 , it was possible to measure "Inelastic" atomic strains (X, Y, Shear strain) and removing the elastic strains. Am i right?!

Yours Sincerely,

Bahman

Dear Bahman,

you probably meant to type ovito instead of lammps, am I right? What you're asking for can be done by using the Atomic Strain modifier and activating the option affine mapping -> to reference.

This is explained in detail in the section Affine Mapping of the simulation cell in the following manual entry:
https://www.ovito.org/docs/current/particles.modifiers.atomic_strain.php

-Constanze

Dear Constanze,

Thank you for your response. Yes, sorry, I meant OVITO! also sorry for my question, I came back to this analysis after several month and I did not expect that I would forget some of the details!

Yours Sincerely,

Bahman

Dear Constanze,

My model is 2 dimension, non- periodic and the simulation cell shrinks NON-uniformly in Y direction .  Shall I remain the "affine mapping " off  OR use "affine mapping of simulation cell to reference " ?! I am going to measure the inelastic strain tensors by using histogram and making average at some specified time steps.

According to OVITIO manual, the affine mapping "should" be remained off for NON-PERIODIC systems. How about for cells with non-uniform deformations like my case?!

Yours Sincerely,

Bahman

Bahman

Hi Bahman,

Could you please post two pictures showing the initial and final states of your simulation? It doesn't become clear from your description what is happening. Note that the simulation cell can, by definition, only deform uniformly. Here, "uniformly" means that the deformation is affine and doesn't depend on the spatial position.

In contrast, the particles/atoms can be deformed non-uniformly, because each atom can move independently from the others. The Atomic Strain function of OVITO allows you to calculate this non-uniform deformation field described by the moving particles. In the default mode, when the "affine mapping" option is turned off, the calculation only takes into account the particle motion, ignoring any deformation of the surrounding simulation cell. Sometimes, however, it is desirable to subtract the uniform deformation described by the simulation cell from the (non-uniform) deformation described by the particles. One such scenario is a deformation test, which leads to both elastic and plastic deformation of a material. Plastic deformation typically is accompanied by localised, relative displacements of atoms (slip), whereas elastic deformation happens more uniformly. In such a situation, the "affine mapping" option of the Atomic Strain modifier may provide a way to "filter out" the overall deformation described by the surrounding simulation cell (corresponding, in first approximation, to the elastic deformation the material undergoes). The atomic strains are then calculated as if the overall shape of the simulation cell did not change.

Hi Alexander,

Thank you very much for your email. Please find in the attachment the initial model and the final model (at time t). the model is a nanoparticle collides to a substrate with an initial impact velocity of V.

there is no periodic boundary  in X and Y directions and Simulation cell size is fixed in my simulations. However, OVITO reports a shrinkage in cell during the simulation when the nanoparticle comes down and I guess strain analysis considers this cell shrinkage in strain analysis too.

Yours Sincerely,

Bahman

 

Uploaded files:
  • You need to login to have access to uploads.

In a simulation without periodic boundary conditions, the simulation box is typically resized (either by OVITO or the MD code) in an ad-hoc fashion to completely contain all atoms in their current configuration (bounding box). Some MD codes do not update the simulation box at all during the course of a non-periodic simulation.

That means the simulation box has no physical meaning in this case and should be ignored, because it is not representative of the elastic strain in the material. In other words, you should leave the "affine mapping" option of the Atomic Strain modifier turned off.

There is not really a way to separate elastic and plastic deformation in such a situation. One option you still have though is to use OVITO's Calculate Elastic Strain function, either instead of or in addition to the Atomic Strain function. The Calculate Elastic Strain modifier lets you compute purely elastic lattice strains, but only at non-defective atom sites of the crystal.

Dear Alexander,

thank you very much for your valuable clarification!

Yours Sincerely,

Bahman