Reassessing the transport properties of fluids: A symbolic regression approach.

The viscosity and thermal conductivity coefficients of the Lennard-Jones fluid are extracted through symbolic regression (SR) techniques from data derived from simulations at the atomic scale. This data-oriented approach provides closed form relations that achieve fine accuracy when compared to well-established theoretical, empirical, or approximate equations, fully transparent, with small complexity and high interpretability. The novelty is further outlined by suggesting analytical expressions for estimating fluid transport properties across the whole phase space, from a dilute gas to a dense liquid, by considering only two macroscopic properties (density and temperature). In such expressions, the underlying physical mechanisms are reflected, while, at the same time, it can be a computationally efficient alternative to costly in time and size first principle and/or molecular dynamics simulations.

Fluid Properties Extraction in Confined Nanochannels with Molecular Dynamics and Symbolic Regression Methods.

In this paper, we propose an alternative road to calculate the transport coefficients of fluids and the slip length inside nano-conduits in a Poiseuille-like geometry. These are all computationally demanding properties that depend on dynamic, thermal, and geometrical characteristics of the implied fluid and the wall material. By introducing the genetic programming-based method of symbolic regression, we are able to derive interpretable data-based mathematical expressions based on previous molecular dynamics simulation data. Emphasis is placed on the physical interpretability of the symbolic expressions. The outcome is a set of mathematical equations, with reduced complexity and increased accuracy, that adhere to existing domain knowledge and can be exploited in fluid property interpolation and extrapolation, bypassing timely simulations when possible.

Artificial Intelligence in Physical Sciences: Symbolic Regression Trends and Perspectives.

Symbolic regression (SR) is a machine learning-based regression method based on genetic programming principles that integrates techniques and processes from heterogeneous scientific fields and is capable of providing analytical equations purely from data. This remarkable characteristic diminishes the need to incorporate prior knowledge about the investigated system. SR can spot profound and elucidate ambiguous relations that can be generalizable, applicable, explainable and span over most scientific, technological, economical, and social principles. In this review, current state of the art is documented, technical and physical characteristics of SR are presented, the available programming techniques are investigated, fields of application are explored, and future perspectives are discussed. Supplementary Information: The online version contains supplementary material available at 10.1007/s11831-023-09922-z.

Influence of carbon nanotube suspensions on Casson fluid flow over a permeable shrinking membrane: an analytical approach.

The present work employs the single-wall carbon nanotube (SWCNT) and multiwall carbon nanotube (MWCNT) models on axisymmetric Casson fluid flow over a permeable shrinking sheet in the presence of an inclined magnetic field and thermal radiation. By exploiting the similarity variable, the leading nonlinear partial differential equations (PDEs) are converted into dimensionless ordinary differential equations (ODEs). The derived equations are solved analytically, and a dual solution is obtained as a result of the shrinking sheet. The dual solutions for the associated model are found to be numerically stable once the stability analysis is conducted, and the upper branch solution is more stable compared to lower branch solutions. The impact of various physical parameters on velocity and temperature distribution is graphically depicted and discussed in detail. The single wall carbon nanotubes have been found to achieve higher temperatures compared to multiwall carbon nanotubes. According to our findings, adding carbon nanotubes volume fractions to convectional fluids can significantly improve thermal conductivity, and this can find applicability in real world applications such as lubricant technology, allowing for efficient heat dissipation in high-temperatures, enhancing the load-carrying capacity and wear resistance of the machinery.

Analytical investigation of an incompressible viscous laminar Casson fluid flow past a stretching/shrinking sheet.

This paper presents an analytical approach on capturing the effect of incompressible, non-Newtonian, viscous, Casson nanofluid flow past a stretching/shrinking surface, under the influence of heat radiation and mass transfer parameter. The governing nonlinear partial differential equations are first transformed into a series of associated nonlinear ordinary differential equations with aid of predictable transformation, while numerical computations follow. The implied nanofluid here is aluminum oxide ([Formula: see text]). The analytical solution is exploited to reveal the accompanying non-dimensional boundary value problem and output results are employed to verify the method's reliability, where it is shown that they agree with current findings in the field. An incomplete gamma function is used to solve temperature equation analytically. We present various instances of the solution, depicting effects of the essential flow factor, the stretching/shrinking parameter, the mass transfer parameter, radiation parameter, and Prandtl number.

Effects of channel size, wall wettability, and electric field strength on ion removal from water in nanochannels.

Molecular dynamics simulations are employed to estimate the effect of nanopore size, wall wettability, and the external field strength on successful ion removal from water solutions. It is demonstrated that the presence of ions, along with the additive effect of an external electric field, constitute a multivariate environment that affect fluidic interactions and facilitate, or block, ion drift to the walls. The potential energy is calculated across every channel case investigated, indicating possible ion localization, while electric field lines are presented, to reveal ion routing throughout the channel. The electric field strength is the dominant ion separation factor, while wall wettability strength, which characterizes if the walls are hydrophobic or hydrophilic has not been found to affect ion movement significantly at the scale studied here. Moreover, the diffusion coefficient values along the three dimensions are reported. Diffusion coefficients have shown a decreasing tendency as the external electric field increases, and do not seem to be affected by the degree of wall wettability at the scale investigated here.

Nanoscale slip length prediction with machine learning tools.

This work incorporates machine learning (ML) techniques, such as multivariate regression, the multi-layer perceptron, and random forest to predict the slip length at the nanoscale. Data points are collected both from our simulation data and data from the literature, and comprise Molecular Dynamics simulations of simple monoatomic, polar, and molecular liquids. Training and test points cover a wide range of input parameters which have been found to affect the slip length value, concerning dynamical and geometrical characteristics of the model, along with simulation parameters that constitute the simulation conditions. The aim of this work is to suggest an accurate and efficient procedure capable of reproducing physical properties, such as the slip length, acting parallel to simulation methods. Non-linear models, based on neural networks and decision trees, have been found to achieve better performance compared to linear regression methods. After the model is trained on representative simulation data, it is capable of accurately predicting the slip length values in regions between or in close proximity to the input data range, at the nanoscale. Results also reveal that, as channel dimensions increase, the slip length turns into a size-independent material property, affected mainly by wall roughness and wettability.

Molecular Dynamics Simulations of Ion Drift in Nanochannel Water Flow.

The present paper employs Molecular Dynamics (MD) simulations to reveal nanoscale ion separation from water/ion flows under an external electric field in Poiseuille-like nanochannels. Ions are drifted to the sidewalls due to the effect of wall-normal applied electric fields while flowing inside the channel. Fresh water is obtained from the channel centerline, while ions are rejected near the walls, similar to the Capacitive DeIonization (CDI) principles. Parameters affecting the separation process, i.e., simulation duration, percentage of the removal, volumetric flow rate, and the length of the nanochannel incorporated, are affected by the electric field magnitude, ion correlations, and channel height. For the range of channels investigated here, an ion removal percentage near 100% is achieved in most cases in less than 20 ns for an electric field magnitude of E = 2.0 V/Å. In the nutshell, the ion drift is found satisfactory in the proposed nanoscale method, and it is exploited in a practical, small-scale system. Theoretical investigation from this work can be projected for systems at larger scales to perform fundamental yet elusive studies on water/ion separation issues at the nanoscale and, one step further, for designing real devices as well. The advantages over existing methods refer to the ease of implementation, low cost, and energy consumption, without the need to confront membrane fouling problems and complex electrode material fabrication employed in CDI.

Effects of wall roughness on flow in nanochannels.

Nonequilibrium molecular dynamics simulation is applied to investigate the effect of periodic wall roughness on the flow of liquid argon through krypton nanochannels. The effect of the length of a rectangular protrusion on the flow is investigated and compared to the case of nanochannels with flat walls. The results show a clear trapping of fluid atoms inside the rectangular cavities that are formed between successive protrusions. The size of the cavities affects the potential energy map and, consequently, fluid atom localization. This localization results in a reduction of velocity values inside the cavities, as well as a reduction of the slip length near the rough wall.