[gmx-users] Analysis pull code simulation data to obtain Strain Stress curve

[Alex]

Since you are not providing any details on your boundary conditions, or 
whether you'd like to calculate the elastic modulus at a particular 
temperature, or in the static approach, here's a general advice.

The potential energy of the system (per unit of sample cross-section) 
associated with small strain is E = Y*eps^2/2, where E is the energy, 
eps is the relative strain along a given axis (orthogonal to the 
cross-section), and Y is the Young's modulus. The goal is to stretch or 
squeeze the structure at different values of eps and plot the value of E 
as a function of eps. If quadratic dependence is observed, you can 
easily fit a curve with a given value of Y. Note that E is not absolute, 
but the difference between what your simulation spits out and that 
corresponding to eps = 0, i.e. the energy shift. Whether E comes from 
energy minimization or as an average of a well-converged finite-T 
simulation, this is a fairly standard approach (see, for instance, Fig. 
S2 (a) of the supplement for 
http://pubs.acs.org/doi/abs/10.1021/acsnano.6b05274).

This works for solids. For softer polymers, you will have to try and 
see. One can possibly use slow pulling to get the curve in one shot, but 
I would use a scripted series of simulations, i.e. a loop like this:

1. simulate/relax the structure

2. save the final structure (coordinates and velocities, if dynamic)

3. use the structure from step 2, but rescale the box in a given 
direction by, say 0.25%

4. repeat step 1.

 From each simulation, assuming you properly extract your energy data, 
you can obtain an average and get the entire result automatically.

Alex


On 3/25/2017 12:44 AM, Md. Imrul Reza Shishir wrote:
> Dear Alex
>
> My initial target to obtain nanomechanical properties of cellulose
> nanofibril. I have an 18 chain Cellulose Iβ structure of 20 residues. I
> want to obtain the young modulus of the fibril my molecular dynamics.
>
> But I am confused how I can approach to obtain this target.
>
> Thanks for the response. And wish some guidance.
>
> Best Regards
> Md Imrul Reza Shishir
>
>
> On Fri, Mar 24, 2017 at 5:35 PM, Alex <nedomacho at gmail.com> wrote:
>
>> Hi there,
>>
>> Everyone is confused about how to get stress-strain curves from Justin's
>> tutorial. Because it has nothing to do with nanomechanics. Obtaining
>> Young's modulus from a particle-based simulation is quite elementary, if a
>> clear physical picture is present.
>>
>> Please describe your system and explain what you are trying to achieve.
>>
>> Alex
>>
>>
>> On 3/24/2017 2:15 AM, Md. Imrul Reza Shishir wrote:
>>
>>> Dear all
>>> My target is to calculate nanomechanical properties (i.e. young's
>>> modulus/elastic modulus) from a strain-stress curve.
>>> For this purpose, I study Beven lab umbrella sampling pull code
>>> simulation.
>>> After that, I also run the same procedure simulation on my desire system.
>>> But I am now confused how I can obtain strain-stress curve from this
>>> simulation data.
>>>
>>> I want some guidance about this purpose.
>>>
>>> Thank you very much
>>>
>>>
>> --
>> Gromacs Users mailing list
>>
>> * Please search the archive at http://www.gromacs.org/Support
>> /Mailing_Lists/GMX-Users_List before posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>> send a mail to gmx-users-request at gromacs.org.
>>
>
>