The number of molecules used in a typical Molecular Dynamics (MD) simulations is orders of magnitude lower than in the thermodynamic limit. It is therefore essential to correct diffusivities computed from Molecular Dynamics simulations for finite-size effects. We present a comprehensive review on finite-size effects of diffusion coefficients by considering self-, Maxwell–Stefan, and Fick diffusion coefficients in pure liquids, as well as binary, ternary, and quaternary mixtures. All finite-size corrections, both analytical and empirical, are discussed in detail. The finite-size effects of rotational and confined diffusion are also briefly discussed.

We present a new plugin for LAMMPS for on-the-fly computation of transport properties (OCTP) in equilibrium molecular dynamics. OCTP computes the self- and Maxwell-Stefan diffusivities, bulk and shear viscosities, and thermal conductivities of pure fluids and mixtures in a single simulation. OCTP is the first implementation in LAMMPS that uses the Einstein relations combined with the order-n algorithm for the efficient sampling of dynamic variables. OCTP has low computational requirements and is easy to use because it follows the native input file format of LAMMPS. A tool for calculating the radial distribution function (RDF) of the fluid beyond the cutoff radius, while taking into account the system size effects, is also part of the new plugin. The RDFs computed from OCTP are needed to obtain the thermodynamic factor, which relates Maxwell-Stefan and Fick diffusivities. To demonstrate the efficiency of the new plugin, the transport properties of an equimolar mixture of water-methanol were computed at 298 K and 1 bar.

The conventional design of organic Rankine cycle (ORC) power systems starts with the selection of the working fluid and the subsequent optimization of the corresponding thermodynamic cycle. More recently, systematic methods have been proposed integrating the selection of the working fluid into the optimization of the thermodynamic cycle. However, in both cases, the turbine is designed subsequently. This procedure can lead to a suboptimal design, especially in the case of mini- and small-scale ORC systems, since the preselected combination of working fluid and operating conditions may lead to infeasible turbine designs. The resulting iterative design procedure may end in conservative solutions after multiple trial-and-error attempts due to the strong interdependence of the many design variables and constraints involved. In this work, we therefore present a new design and optimization method integrating working fluid selection, thermodynamic cycle design, and preliminary turbine design. To this purpose, our recent 1-stage continuous-molecular targeting (CoMT)-computer-aided molecular design (CAMD) method for the integrated design of the ORC process and working fluid is expanded by a turbine meanline design procedure. Thereby, the search space of the optimization is bounded to regions where the design of the turbine is feasible. The resulting method has been tested for the design of a small-scale high-temperature ORC unit adopting a radial-inflow turbo-expander. The results confirm the potential of the proposed method over the conventional iterative design practice for the design of small-scale ORC turbogenerators.

Mutual diffusion coefficients can be successfully predicted with models based on the Darken equation. However, Darken-based models require composition-dependent self-diffusion coefficients which are rarely available. In this work, we present a predictive model for composition-dependent self-diffusion coefficients (also called tracer diffusion coefficients or intradiffusion coefficients) in nonideal binary liquid mixtures. The model is derived from molecular dynamics simulation data of Lennard-Jones systems. A strong correlation between nonideal diffusion effects and the thermodynamic factor is observed. We extend the model by McCarty and Mason (Phys. Fluids 1960, 3, 908-922) for ideal binary gas mixtures to predict the composition-dependent self-diffusion coefficients in nonideal binary liquid mixtures. Our new model is a function of the thermodynamic factor, the self-diffusion coefficients at infinite dilution, and the self-diffusion coefficients of the pure substances, which are readily available. We validate our model with experimental data of 9 systems. For both Lennard-Jones systems and experimental data, the accuracy of the predicted self-diffusion coefficients is improved by a factor of 2 compared to the correlation of McCarty and Mason. Thus, our new model significantly expands the practical applicability of Darken-based models for the prediction of mutual diffusion coefficients.