Nanoscale slip length prediction with machine learning tools

Abstract

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.

Drought Forecasting: A Review and Assessment of the Hybrid Techniques and Data Pre-Processing

“…Screening continuous factors for normality is an essential early stage in the analysis, and the histograms are one of the most important ways of assessing normality [85]. In ML approaches, data curation is a necessary pre-processing step, beginning with data normali-sation to restrict the input value range [86]. As a result, converting continuous variables is essential for creating time series that are normally or close to normally distributed [87].…”

“…On the other hand, various approaches are employed when the strategy is multivariate, such as dimensionality reduction, which can be achieved by retaining input features that have large variances and discarding those terms that have small variances (i.e., Principal Component Analysis (PCA) is a feature selection approach that can reduce the model's dimensionality without impacting its performance.) [86]. Additionally, variance inflation factor and tolerance approaches are used to determine potential multicollinearity and exclude inputs from ML algorithms.)…”

Specific Heat Capacity of Confined Water in Extremely Narrow Graphene Nanochannels

“…As an effective and powerful alternative tool, molecular dynamics (MD) simulation has presented excellent performance in the study of thermophysical properties of water at nanoscales, such as viscosity (Neek-Amal et al, 2016;Zhou et al, 2021), thermal conductivity (Zhao et al, 2020a;Zhao et al, 2020b), diffusion coefficient (Zhao et al, 2020c), and dielectric constant (Hamid et al, 2021). And it becomes a common sense that molecular behaviors of water under nanoconfinement deviate from bulk behaviors (Sofos et al, 2013;Sofos and Karakasidis, 2021) because of the large surface-to-volume ratio and the enhanced effects of surface properties including surface wettability and surface morphology (Shadloo-Jahromi et al, 2020;Shadloo-Jahromi et al, 2021). Thus, the thermophysical properties generally depend on the size of nanochannels or pores (Sofos et al, 2013;Zhao et al, 2020a;Zhao et al, 2020b;Zhao et al, 2020c;Hamid et al, 2021;Sofos and Karakasidis, 2021;Zhou et al, 2021).…”

“…And it becomes a common sense that molecular behaviors of water under nanoconfinement deviate from bulk behaviors (Sofos et al, 2013;Sofos and Karakasidis, 2021) because of the large surface-to-volume ratio and the enhanced effects of surface properties including surface wettability and surface morphology (Shadloo-Jahromi et al, 2020;Shadloo-Jahromi et al, 2021). Thus, the thermophysical properties generally depend on the size of nanochannels or pores (Sofos et al, 2013;Zhao et al, 2020a;Zhao et al, 2020b;Zhao et al, 2020c;Hamid et al, 2021;Sofos and Karakasidis, 2021;Zhou et al, 2021). Studies on the specific heat capacity of water and other fluids using the MD method (Gautam et al, 2018;Alkhwaji et al, 2021) further demonstrated the potential of the MD method on the investigation of water at nanoscales.…”

A review on slip boundary conditions at the nanoscale: recent development and applications

“…Generally, the methods to investigate slip boundary conditions for nanoconfined liquids include theoretical analysis, physical experiments, and numerical simulations [8,[24][25][26][27][28][29][30][31][32][33][34]. In recent years, machine learning methods have also been applied in the study of dynamic properties of liquids including diffusion and slip flow behavior, and in the prediction of the slip length at the nanoscale [35][36][37]. Owing to the rapid development of technology, experimental studies of flow boundary conditions have been successfully extended to nanoscale systems [8,34,38,39].…”