Equation of State (EOS) Analysis¶

Provided pressure/energy and volume data, MechElastic can perform EOS analysis using the following models:

Vinet
Birch
Murnaghan
Birch-Murnaghan

Please find examples for EOS analysis in the examples/EOS/ directory.

plot_eos()¶

INPUT: Text file with two columns containing volume ( $Å^{3}$ ) and energy (eV) (eostype='energy') or volume and pressure (GPa) (eostype='pressure'), respectively, should be passed into the MechElastic package.

With this, an initial second order parabolic polyfit is performed to obtain the initial fitting parameters, $E_{0}$ or $P_{0}$ , $B_{0}$ , $B_{p}$ and $V_{0}$ . Afterwards, a more accurate fitting for the energy or pressure is performed using the least square method against each EOS model. A central difference scheme is used to obtain the pressure from this fitted energy. If pressure is to be obtained from energy, an integrating scheme is used instead. Finally, a plot of Energy vs Volume and Pressure vs Volume is obtained comparing the values obtained from the different EOS models with the originally provided raw data. With these plots, an analysis of the phase boundaries could be performed. The energy or pressure is normalized with the number of atoms in the unit-cell set by natoms. If the output units are to be in atomic units, set au=True.

In addition to the fitting, MechElastic also performs a regression analysis for fitting using the Mean-Squared Error (MSE) of the residuals. This allows users to decide the best EOS model to use for their dataset.

Usage:

from mechelastic import EOS
eos_object = EOS()
eos_object.plot_eos(inputfile='EvsV.dat', eostype='energy', natoms=1, au=False, model="Birch-Murnaghan")

To set a initial and final value to the volume range use the flag vlim=[v_initial, v_final]. If not set, the minimum and maximum values of the provided dataset would be used.

If model is not set it will plot all the models on the same plot along with the raw data points. If raw_data=False the plot will omit the raw data points and will only plot the fitted model. The options for model are Vinet, Birch, Murnaghan and Birch-Murnaghan.

plot_enthalpy_curves()¶

This function plots Enthalpy vs. Pressure curves for multiple datasets of different phases and calculates the phase transition pressures. The pressure is calculated through a central difference scheme similar to plot_eos().

INPUT: Multiple input files containing two columns of volume ( $Å^{3}$ ) and energy (eV) for each phase.

Usage:

from mechelastic import EOS

eos_object = EOS()

infiles = ["Rm-3m", "Pm-3m", "Im-3m", "I4_mcm", "C12:m1"]
natoms = [6, 1, 2, 8, 4]
eos_object.plot_enthalpy_curves(infiles, natoms, au=True)

The flags vlim and au work similar to the plot_eos() section. If the volume ranges are to be set individually to each phase, one can use the vlim_list option. This is a list that contains elements of the minimum and maximum value of volume for each phase. For instance, to define the volume ranges to each of the five phases here set, vlim_list = [ [180,260], [25,41], [50,75], [270,372], [110,160] ].

This outputs enthalpy curves and if phase transitions are detected, it will calculate the phase transition pressures and the possible transition paths.

Additionally, one may plot the Enthalpy differences with respect to a certain phase by setting deltaH_index=<phase index> where <phase index> is the index of the phase in the infiles list starting from 0. For example, in the above list of phases, to plot the Enthalpy with respect to Im-3m, one should set deltaH_index=2.

If the data is required in a .json and .xml format, set export=True in either of these two functions.