Regarding applicability of potentials

Hi all,

I want to simulate Ni-Ti alloy to get mechanical properties ( upto very high temperature and pressure by applying some impact or load ) by using eam/fs (e.g- Ni-Ti, NiTi.eam.fs) potential. My questions are–

  1. How can I decide initially that this particular potential is really good to describe the lattice and also interaction at such very high pressure and temperature ? Is there any way to find out the range upto which it may work ? Is there any energy continuity criterion type of things to test it? I have gone through the reference paper related to the potential but confused. As I am in beginner-level in MD, so really sorry, if this is a very basic question.

  2. I have also gone through the paper Phys. Rev. B. 69, 144113 (2004), by Zhao et al which contain the EAM parameters of several metals, if I use EAM parameterization to get mixed alloys for Ni-Ti, would that be use-full to model at such extreme condition (high T & P) ?

Any reference regarding this and your comments would be a great help for me.

Thanks,

Batista Cheng

Hi all,

I want to simulate Ni-Ti alloy to get mechanical properties ( upto very high
temperature and pressure by applying some impact or load ) by using eam/fs
(e.g- Ni-Ti, NiTi.eam.fs) potential. My questions are--

1. How can I decide initially that this particular potential is really good
to describe the lattice and also interaction at such very high pressure and
temperature ? Is there any way to find out the range upto which it may work

this is the million dollar question. if you could say up front if a
potential is suitable for a specific area of phase space, things would
be a lot easier. anyway, there are two aspects to look into:
1) what is the strength or weakness of a particular model (e.g. EAM)
for your systems of interest? for that you can search the literature
and look for studies on other materials. it is probably best to start
by looking for review articles.
2) how transferable is the specific set of parameters? for that you
need to look for experimental data to compare to and do test
calculations. again, a literature search can help. perhaps somebody
has already looked into this.

this gets complicated by using an alloy, so you have another degree of
freedom in the composition of the alloy. again, you have to check how
well EAM works for high or low or equal concentrations of A in B.

? Is there any energy continuity criterion type of things to test it? I
have gone through the reference paper related to the potential but confused.
As I am in beginner-level in MD, so really sorry, if this is a very basic
question.

validation of a potential is a good way to start. you are asking the
right questions. it can become a lot of work though...

axel.

For high pressures I would look at some of the shock work done previously.
Beware that NiTi behaves in an unconventional way.

see:-
JCF Millet et al
http://dx.doi.org/10.1063/1.1498877
Meziere
http://dx.doi.org/10.1063/1.2219083
and
Nemat-Nasser et al
Mechanics of Materials 38 (2006) 463–474

Then compare some initial results to the experimental data

Nigel

Hi,

Well, Usually the semi-empirical potential is compared Ab-initio Calculations and Experimental Data . For example : Comparing the Melting Curve or the phase stability by means of the enthalpy differences at T = 0 , for various phases is a good Start …

Oscar G.

Hi ,

many many thanks to Axel, Nigel and Oscar for your valuable insight.
So far I realize that validation is the correct way to confirm the range upto which the potential is capable to describe.
I have tested some stress-strain behavior (upto ~42 GPa, with small strain rate) initially looks like a match with experimental findings, but needs to see what happen in case of high strain-rate/ shock loading condition.

But Oscar would you be so pleased to elaborate " phase stability by means of the enthalpy differences at T = 0 , for various phases".

  1. How can I calculate phase stability ?
  2. Regarding different phase, are you indicating bcc/fcc type of initial configuration ?

In this context, I want to acknowledge and appreciate Oscar’s Shock simulation manual, which I found in archive of this forum, is really a help. I found the 1st version in archive, Oscar have you add or modify it now ? Regarding the T and P calculation there is some issue to subtract com velocity bias, if I’m not wrong.

Thanks,

Batista Cheng

  1. How can I calculate phase stability ?
    Usually the phase change is realated with a change in the Thermodymic potential. More precisely a change in the Gibbs Free Energy G(p,T) = U + pVTS (in order to see 2nd order phase transition), but at T=0 notice G = H = U + P V . For example : its well know BCC change to HCP under pressure thus an Analysis of the enthalpy differences Hbcc -Hhcp = 0 , its a good way to determine a threshold in pressure in which the potential is good describre the BCC phase . Also comparation of the EOS vs AB-initio data …

  2. Regarding different phase, are you indicating bcc/fcc type of initial configuration ?
    YES

  3. Oscar have you add or modify it now ?
    No …But more probably I will add more stuff as time permits (I am a lazy person), but there are also a few things I must publish before releasing the info …

A Salute
Oscar G.