Reliable prediction of transport coefficients for fluids, such as the viscosity, thermal conductivity, and diffusion coefficients, is an important prerequisite for process design. Besides experimental measurements and semiempirical correlations, molecular simulations are a promising method to estimate transport properties of fluids over wide ranges of temperatures and pressures. Transport properties are sensitive to the underlying intermolecular potentials. In this work we assess the Transferable Anisotropic united-atom Mie (TAMie) force field regarding the calculation of transport properties. The force field was parametrized for thermodynamic properties with emphasis on vapor− liquid coexistence properties. Equilibrium molecular dynamic simulations are used to calculate all transport properties in a single simulation, using the corresponding Green−Kubo methods. The simulated state points were distributed in the temperature and pressure based on an entropy scaling approach, where the PC-SAFT equation of state is used to calculate residual entropy. Utilizing the favorable behavior of dynamic properties when plotted over the residual entropy, only few simulations are needed to parametrize a correlation function that furthermore enables comparison with experimental data over a wide range of temperatures and pressures. TAMie yields good results for all transport properties and substances investigated (with average absolute deviations of 13% for viscosity, 18% for diffusion, and 10% for thermal conductivity), given that only static properties were considered in the parametrization of the force field. Combining few simulations with the entropy scaling method enables very efficient prediction of transport properties for a large temperature and pressure region.
A method is presented that allows to combine the effective potential between two nano crystals, the potential of mean force (PMF), as obtained from all-atomistic Molecular Dynamics simulations with perturbation theory. In this way, a functional dependence of the PMF on temperature is derived, that enables the prediction of the PMF in a wide temperature range. We applied the method for systems of capped gold nano crystals of different size. They show very good agreement with data from atomistic simulations.
In this work, we
present an open-source software package, referred
to as FeOsFramework for Equations of State and
Classical Density Functional Theory. FeOs is a collection
of interfaces and data types that can be used (1) to implement thermodynamic
equations of state and Helmholtz energy functionals for classical
density functional theory, and (2) to compute thermodynamic properties
of pure substances and mixtures, phase equilibria, and interfacial
properties such as surface tensions and adsorption isotherms. The
framework is written in the Rust programming language with a complete
Python interface and is designed with a focus on usability and extensibility.
It is openly available on GitHub (). Equations of state can be implemented in Rust, yielding performant
code, or as a Python class, which is useful for prototyping and with
less emphasis on execution speed. In both cases, the user has to implement
a single function: the Helmholtz energy. FeOs then uses
generalized (hyper-) dual numbers to evaluate the Helmholtz energy
as well as the required exact partial (higher-order) derivatives.
Using this type of automatic differentiation delivers performance
without the need for implementing any analytical derivatives. The
performance is further enhanced by a caching mechanism that avoids
duplicate model evaluations. Together with the core interfaces and
functionalities for equations of state and classical density functional
theory, we provide implementations for multiple models such as the
PC-SAFT equation of state (with homo- and heterosegmented group contribution
methods) and Helmholtz energy functionals (including segment-based
functionals). To showcase a selection of FeOs’ features,
an example study of the adsorption of biogas in porous media using
the PC-SAFT functional is provided.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.