Prediction Of Viscosity Research Articles

Rheological behavior predictions of non-Newtonian nanofluids via correlations and artificial neural network for thermal applications

Nanofluids possess enhanced viscous and thermal features that can be utilized to improve the heat transfer performance of several applications involving sustainable manufacturing and industrial ecology, such as in heating/cooling systems, electronics, transportation etc. Therefore, it is important to understand and optimize the flow pattern of these fluids. This research emphasizes the predictions of viscosity of water/ethylene-glycol (EG) based non-Newtonian nanofluids. Four experiment-based data sets are used to predict and validate the effective viscosity, i.e., Fe3O4-Ag/EG, MWCNT-alumina/water-EG, Fe3O4-Ag/water-EG, and MWCNT-SiO2/EG-water, via existing correlations and artificial neural network (ANN). The modeling is based on three input parameters, i.e., particle concentrations, temperatures, and shear rates, and one output parameter, i.e., viscosity. The predicted outcomes are compared to the three existing correlation structures. The error matrix consists of the coefficient of determination (R2), average absolute deviation (AAD %), the sum of squared error (SSE), that are employed to evaluate the performance of the model. Results from ANN are found to be more precise, with R2 values greater than 0.99 for all datasets, compared to data fitting into existing correlations, in which Fe3O4-Ag/water-EG resulted in an R2 value as low as 0.72, to determine the nanofluids’ effective viscosity.

Bimodal suspensions: semi-empirical prediction of viscosity and linear viscoelastic response

Bimodal suspensions: semi-empirical prediction of viscosity and linear viscoelastic response

Machine learning prediction and optimization of CO2 foam performance for enhanced oil recovery and carbon sequestration: Effect of surfactant type and operating conditions

Foam flooding handles the challenges of gas injection constituting a high viscosity medium that sweeps oil more efficiently meanwhile creating a barrier that prevents leakage of gas phase. Therefore, it is a promising enhanced oil recovery (EOR) method, simultaneously an efficient carbon dioxide (CO2) storage technology. The foam's performance displays susceptibility to various reservoir and operational parameters. Numerous laboratory experiments would need to be carried out to study foams and find the optimum foam formulation for the given core conditions. Data-driven approaches can be an alternative method to the time-consuming experimental and conventional modeling techniques, which often face challenges to incorporate the effect of important related parameters such as temperature and type of surfactant to the performance of foam. In the current study, machine learning (ML) models were constructed to predict the CO2 foam apparent viscosity in carbonate and sandstone formations focusing on experimental data from the literature for CO2 foams stabilized by different surfactants. Various anionic, cationic, and nonionic surfactants were considered and characterized based on the hydrophilic-lipophilic balance (HLB) number. Predictive models were developed using ML algorithms based on the data available for eight key process parameters. Artificial neural networks and extra randomized tree algorithms provided the most precise prediction among the applied algorithms. Various predictions for foam apparent viscosity on hypothetical case studies were made on altered input parameters, enabling CO2 mobility to co-optimize EOR and CO2 sequestration while selecting the optimum surfactant for each case.

Fluid viscosity prediction leveraging computer vision and robot interaction

Accurately determining fluid viscosity is crucial for various industrial and scientific applications. Traditional methods of viscosity measurement, though reliable, often require manual intervention and cannot easily adapt to real-time monitoring. With advancements in machine learning and computer vision, this work explores the feasibility of predicting fluid viscosity by analyzing fluid oscillations captured in video data. The pipeline employs a 3D convolutional autoencoder pretrained in a self-supervised manner to extract and learn features from semantic segmentation masks of oscillating fluids. Then, the latent representations of the input data, produced from the pretrained autoencoder, are processed with a distinct inference head to infer either the fluid category (classification) or the fluid viscosity (regression) in a time-resolved manner. When the latent representations generated by the pre-trained autoencoder are used for classification, the system achieves a 97.1% accuracy across a total of 4140 test datapoints. Similarly, for regression tasks, employing an additional fully-connected network as a regression head allows the pipeline to achieve a mean absolute error of 0.258 cP over 4416 test datapoints. This study represents an innovative contribution to both fluid characterization and the evolving landscape of Artificial Intelligence, demonstrating the potential of deep learning in achieving near real-time viscosity estimation and addressing practical challenges in fluid dynamics through the analysis of video data capturing oscillating fluid dynamics.

Viscosity Calculation for Al‐Si‐Mg‐Fe System through CALPHAD Method

Viscosity is a crucial parameter affecting the fluidity of metal melts, which directly influences founding properties of Al alloys. However, obtaining viscosity measurement data is difficult for metal melts, and the viscosity prediction for ternary and multi‐component alloys would provide the significant data for selection of process parameters. This study compares the applicability of the Hirai model, SDS model, and R‐K function for viscosity calculations in the sub‐binary systems of Al‐Si‐Mg‐Fe alloys. R‐K function shows the best agreement with experimental data. However, the SDS model shows lower relative error than Hirai model which would play an important role in those system lacking experimental data to predict the viscosities. Ultimately, a database capable of predicting viscosity values for the entire composition and temperature range of the Al‐Si‐Mg‐Fe system was established using the CALPHAD method. Viscosity parameters for Al‐Si, Al‐Mg, Al‐Fe, Mg‐Si, and Mg‐Fe were evaluated and optimized through the R‐K derivation, corroborated with existing experimental data. Using the R‐K function, successful extrapolation and prediction of viscosity for the Al‐Si‐Mg‐Fe ternary and quaternary systems were achieved, with a mean square error between predicted and experimental values of only 0.7%, demonstrating the successful application of the Al‐Si‐Mg‐Fe database for viscosity prediction.This article is protected by copyright. All rights reserved.

Dynamic viscosity prediction of nanofluids using artificial neural network (ANN) and genetic algorithm (GA)

Dynamic viscosity prediction of nanofluids using artificial neural network (ANN) and genetic algorithm (GA)

Sequence-Based Viscosity Prediction for Rapid Antibody Engineering.

Through machine learning, identifying correlations between amino acid sequences of antibodies and their observed characteristics, we developed an internal viscosity prediction model to empower the rapid engineering of therapeutic antibody candidates. For a highly viscous anti-IL-13 monoclonal antibody, we used a structure-based rational design strategy to generate a list of variants that were hypothesized to mitigate viscosity. Our viscosity prediction tool was then used as a screen to cull virtually engineered variants with a probability of high viscosity while advancing those with a probability of low viscosity to production and testing. By combining the rational design engineering strategy with the in silico viscosity prediction screening step, we were able to efficiently improve the highly viscous anti-IL-13 candidate, successfully decreasing the viscosity at 150 mg/mL from 34 cP to 13 cP in a panel of 16 variants.

Viscosity and melting temperature prediction of mold fluxes based on explainable machine learning and SHapley additive exPlanations

Viscosity and melting temperature are indispensable properties for mold flux design and evaluation, significant investment is required for measurement. In this paper, viscosity and melting temperature predictive model for mold fluxes were established and trained based on 3300 groups of data and four representative machine learning algorithms. The gradient boosting regression tree (GBRT) algorithm-based model performed best prediction with the determination coefficient R2 of 0.969 and 0.900 for viscosity and melting temperature. It also far outperforms the widely used viscosity and melting temperature prediction models, with a least mean deviation of 7.06% and 0.8%. SHapley Additive exPlanations (SHAP) analysis revealed that Al2O3, followed by SiO2 showed the strongest positive correlation with viscosity, Na2O has the greatest negative contribution to melting temperature. Ternary viscosity and melting temperature distribution diagrams were constructed to visualize the prediction results. Insights from this study will highly benefit computer-aided design of mold fluxes with desired properties.

Development of a viscosity prediction model and investigation into drag reduction in weakly magnetized waxy crude oil

The present study investigates the low magnetization of marine waxy crude oil. Through variance analysis and nonlinear regression analysis, a four-parameter and five-factor model was established to predict the magnetized viscosity of waxy crude oil. The mean absolute relative error of this model is 4.97 %, indicating its suitability for predicting the viscosity of magnetized waxy crude oil with an intensity below 100 mT. The prediction model for magnetized viscosity provides a theoretical basis for predicting key parameters in magnetization transport pressure drop and establishing the wax deposition model under magnetization conditions. By coupling the flow field and magnetic field, a pressure drop model for transporting magnetized waxy crude oil was established to predict the drag reduction effect in magnetized transportation. It is anticipated that magnetization can potentially reduce the drag rate of an oceanic oil pipeline by 8 %–10 %.

Exploring unseen 3D scenarios of physics variables using machine learning-based synthetic data: An application to wave energy converters

This work uses machine learning to produce synthetic data of wave energy converters from time-expensive 3D simulations based on computational fluid dynamics models. The simulations to analyse the response of these systems to incoming waves are lengthy and computationally expensive to obtain. Here, we explore the use of a beta-VAE and a Principal Components-based adversarial autoencoder for generating new synthetic data. The compression plus the generation of synthetic data introduces an exceptionally fast surrogate model of the original simulation and delivers more samples of either dynamic viscosity and velocity fields, enlarging the design space. The new generated synthetic samples can have a speed up from 5 to 6 orders of magnitude. The new design space can be used to improve the prediction of dynamic viscosity given the velocity fields. The generative model has the potential to capture the transition and the new physical phenomena under extreme initial conditions.

A Machine Learning Approach to Predict Fluid Viscosity Based on Droplet Dynamics Features

In recent years, machine learning has made significant progress in the field of micro-fluids, and viscosity prediction has become one of the hotspots of research. Due to the specificity of the application direction, the input datasets required for machine learning models are diverse, which limits the generalisation ability of the models. This paper starts by analysing the most obvious kinetic feature induced by viscosity during flow—the variation in droplet neck contraction with time (hmin/R∼τ). The kinetic processes of aqueous glycerol solutions of different viscosities when dropped in air were investigated by high-speed camera experiments, and the kinetic characteristics of the contraction of the liquid neck during droplet falling were extracted, using the Ohnesorge number (Oh=μ/(ρRσ)1/2) to represent the change in viscosity. Subsequently, the liquid neck contraction data were used as the original dataset, and three models, namely, random forest, multiple linear regression, and neural network, were used for training. The final results showed superior results for all three models, with the multivariate linear regression model having the best predictive ability with a correlation coefficient R2 of 0.98.

High-Throughput Screening and Accurate Prediction of Ionic Liquid Viscosities Using Interpretable Machine Learning

High-Throughput Screening and Accurate Prediction of Ionic Liquid Viscosities Using Interpretable Machine Learning

Simulated phase state and viscosity of secondary organic aerosols over China

Abstract. Secondary organic aerosols (SOAs) can exist in liquid, semi-solid, or amorphous solid states. Chemical transport models (CTMs), however, usually assume that SOA particles are homogeneous and well-mixed liquids, with rapid establishment of gas–particle equilibrium for simulations of SOA formation and partitioning. Missing the information of SOA phase state and viscosity in CTMs impedes accurate representation of SOA formation and evolution, affecting the predictions of aerosol effects on air quality and climate. We have previously developed a parameterization to estimate the glass transition temperature (Tg) of an organic compound based on volatility and to predict viscosity of SOA. In this study, we apply this method to predict the phase state of SOA particles over China in summer of 2018 using the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem). The simulated Tg of dry SOA (Tg,org) agrees well with the value estimated from ambient volatility measurements at an urban site in Beijing. For the spatial distributions of Tg,org, simulations show that at the surface the values of Tg,org range from ∼287 to 305 K, with higher values in northwestern China, where SOA particles have larger mass fractions of low-volatility compounds. Considering water uptake by SOA particles, the SOA viscosity shows a prominent geospatial gradient in which highly viscous or solid SOA particles are mainly predicted in northwestern China. The lowest and highest SOA viscosity values both occur over the Qinghai–Tibet Plateau, where the solid phase state is predicted over dry and high-altitude areas and the liquid phase state is predicted mainly in the south of the plateau with high relative humidity during the summer monsoon season. Sensitivity simulations show that, including the formation of extremely low-volatility organic compounds, the percent time that a SOA particle is in the liquid phase state decreases by up to 12 % in southeastern China during the simulated period. With an assumption that the organic and inorganic compounds are internally mixed in one phase, we show that the water absorbed by inorganic species can significantly lower the simulated viscosity over southeastern China. This indicates that constraining the uncertainties in simulated SOA volatility distributions and the mixing state of the organic and inorganic compounds would improve prediction of viscosity in multicomponent particles in southeastern China. We also calculate the characteristic mixing timescale of organic molecules in 200 m SOA particles to evaluate kinetic limitations in SOA partitioning. Calculations show that during the simulated period the percent time of the mixing timescale longer than 1 h is >70 % at the surface and at 500 hPa in most areas of northern China, indicating that kinetic partitioning considering the bulk diffusion in viscous particles may be required for more accurate prediction of SOA mass concentrations and size distributions over these areas.

Prediction of Viscosity of the Oil–Surfactant–Brine Microemulsion Phase Using Molecular Dynamics Simulations

Prediction of Viscosity of the Oil–Surfactant–Brine Microemulsion Phase Using Molecular Dynamics Simulations

Applying knowledge management in optimal modeling of viscosity of nanofluids by response surface methodology for use in automobiles engine

In this study, using knowledge management and RSM, modeling and prediction of viscosity of oil-based hybrid nanofluid (HNF) with Multi-walled carbon nanotube (MWCNT) and Al2O3 nanoparticles in temperature conditions of 25–50 °C and solid volume fraction of SVF=0.0625–1% have been addressed. To find the most accurate model for predicting viscosity, Cubic, Quartic, and Fifth models were examined in terms of various model accuracy evaluation parameters. Based on the obtained results, the Fifth model has the most accuracy in terms of coefficient of determination and coefficient of variance with values of 0.9997 and 1.10, respectively. Based on the correlation presented to predict viscosity due to the dependence of viscosity on shear rate, the behavior of HNF is non-Newtonian. The graphs of temperature in terms of SVF for viscosity in all three models show that viscosity decreases with increasing temperature and viscosity increases with increasing SVF. For hot lubrication conditions, the best condition (255.28 mPa.sec) is at 35 °C and SVF is 0.161%. Also, knowledge management is considered as a safe and optimal method to increase efficiency, productivity and effectiveness in the field of nanofluids (NFs), which brings high added value to this sector.

Nuclear magnetic resonance experiments on the time-varying law of oil viscosity and wettability in high-multiple waterflooding sandstone cores

A simulated oil viscosity prediction model is established according to the relationship between simulated oil viscosity and geometric mean value of T2 spectrum, and the time-varying law of simulated oil viscosity in porous media is quantitatively characterized by nuclear magnetic resonance (NMR) experiments of high multiple waterflooding. A new NMR wettability index formula is derived based on NMR relaxation theory to quantitatively characterize the time-varying law of rock wettability during waterflooding combined with high-multiple waterflooding experiment in sandstone cores. The remaining oil viscosity in the core is positively correlated with the displacing water multiple. The remaining oil viscosity increases rapidly when the displacing water multiple is low, and increases slowly when the displacing water multiple is high. The variation of remaining oil viscosity is related to the reservoir heterogeneity. The stronger the reservoir homogeneity, the higher the content of heavy components in the remaining oil and the higher the viscosity. The reservoir wettability changes after water injection: the oil-wet reservoir changes into water-wet reservoir, while the water-wet reservoir becomes more hydrophilic; the degree of change enhances with the increase of displacing water multiple. There is a high correlation between the time-varying oil viscosity and the time-varying wettability, and the change of oil viscosity cannot be ignored. The NMR wettability index calculated by considering the change of oil viscosity is more consistent with the tested Amott (spontaneous imbibition) wettability index, which agrees more with the time-varying law of reservoir wettability.

Deterioration Mechanism and Status Prediction of Hydrocarbon Lubricants under High Temperatures and Humid Environments

In practical engineering applications, high temperatures and water ingress seriously affect the service life of hydrocarbon lubricants. In this study, the deterioration process of hydrocarbon lubricants under high temperatures and humid environments was investigated, and a new health state prediction model was proposed. Simulation of hydrocarbon lubricant Polyalpha−olefin (PAO) molecules used the ReaxFF force field to analyse the high temperature thermal oxidation process of lubricants. The rheological properties of oil−water emulsions were determined by observing the morphology of oil−water two−phase mixtures with different water contents. A multiparameter fusion viscosity prediction model was proposed using a linear model of the viscosity of aqueous fluids, as affected by temperature and water content, and was fitted with the Andrade viscosity−temperature equation to predict lubricant viscosity changes under multiple parameters. Online validation tests were carried out on a compound planetary transmission system, and the surface topographical parameters of the transmission components were further discussed. Experimental results show that the linear correlation with the improved lubricant viscosity prediction model is 0.966, and the surface wear of transmission components is consistent with the trend of lubricant quality change. These findings provide a fundamental basis for the assessment of lubricant service life in high temperatures and humid environments.

Predicting the density and viscosity of deep eutectic solvents at atmospheric and elevated pressures

Deep eutectic solvents (DESs) are promising and sustainable substitutes for organic solvents in various applications. The knowledge of their density and viscosity at different temperatures and pressures is crucial for the corresponding process design. However, most literature primarily focuses on predicting these properties at atmospheric pressure. In this study, we employed the perturbed chain polar statistical associating fluid theory (PCP-SAFT) model and entropy scaling method to predict DESs’ density and viscosity at atmospheric and elevated pressures. 2831 density data points of 38 DESs and 1018 viscosity data points of 31 DESs at different temperatures and pressures were used to evaluate the accuracy of the applied methods. The pure component PCP-SAFT parameters of choline chloride (ChCl) were regressed using density data of ChCl-based DES at various temperatures and pressures. It was found that the PCP-SAFT model can provide satisfactory density predictions for DESs at atmospheric pressure, even without using the system-specific binary interaction parameter (kij=0). However, accurate viscosity prediction at atmospheric pressure for DESs necessitated the inclusion of the binary interaction parameter (kij=aijT+bij) in the PCP-SAFT model. The system-specific coefficients aij andbij, derived from experimental data at atmospheric pressure, were used to predict the density or viscosity of DES at elevated pressures. Accurate predictions for density were observed across the entire pressure range, whereas effective viscosity predictions were confined to a specific pressure range.

Review of modeling schemes and machine learning algorithms for fluid rheological behavior analysis

Abstract Machine learning’s prowess in extracting insights from data has significantly advanced fluid rheological behavior prediction. This machine-learning-based approach, adaptable and precise, is effective when the strategy is appropriately selected. However, a comprehensive review of machine learning applications for predicting fluid rheology across various fields is rare. This article aims to identify and overview effective machine learning strategies for analyzing and predicting fluid rheology. Covering flow curve identification, yield stress characterization, and viscosity prediction, it compares machine learning techniques in these areas. The study finds common objectives across fluid models: flow curve correlation, rheological behavior dependency on variables, soft sensor applications, and spatial–temporal analysis. It is noted that models for one type can often adapt to similar behaviors in other fluids, especially in the first two categories. Simpler algorithms, such as feedforward neural networks and support vector regression, are usually sufficient for cases with narrow range variability and small datasets. Advanced methods, like hybrid approaches combining metaheuristic optimization with machine learning, are suitable for complex scenarios with multiple variables and large datasets. The article also proposes a reproducibility checklist, ensuring consistent research outcomes. This review serves as a guide for future exploration in machine learning for fluid rheology prediction.

Advancing material property prediction: using physics-informed machine learning models for viscosity

In materials science, accurately computing properties like viscosity, melting point, and glass transition temperatures solely through physics-based models is challenging. Data-driven machine learning (ML) also poses challenges in constructing ML models, especially in the material science domain where data is limited. To address this, we integrate physics-informed descriptors from molecular dynamics (MD) simulations to enhance the accuracy and interpretability of ML models. Our current study focuses on accurately predicting viscosity in liquid systems using MD descriptors. In this work, we curated a comprehensive dataset of over 4000 small organic molecules’ viscosities from scientific literature, publications, and online databases. This dataset enabled us to develop quantitative structure–property relationships (QSPR) consisting of descriptor-based and graph neural network models to predict temperature-dependent viscosities for a wide range of viscosities. The QSPR models reveal that including MD descriptors improves the prediction of experimental viscosities, particularly at the small data set scale of fewer than a thousand data points. Furthermore, feature importance tools reveal that intermolecular interactions captured by MD descriptors are most important for viscosity predictions. Finally, the QSPR models can accurately capture the inverse relationship between viscosity and temperature for six battery-relevant solvents, some of which were not included in the original data set. Our research highlights the effectiveness of incorporating MD descriptors into QSPR models, which leads to improved accuracy for properties that are difficult to predict when using physics-based models alone or when limited data is available.Graphical