Search

This study demonstrates the application of artificial neural networks (ANNs) and Bayesian support vector regression (BSVR) models for predicting the relative viscosity of nanofluids. The study examined 19 nanofluids comprising 1425 experimental datasets that were randomly split in a ratio of 70:30 as a training dataset and a testing dataset, respectively. To establish the inputs that will yield the best model prediction, we conducted a systematic analysis of the influence of volume fraction of nanoparticles, the density of nanoparticles, fluid temperature, size of nanoparticles, and viscosity of base fluids on the relative viscosity of the nanofluids. Also, we analyzed the results of all possible input combinations by developing 31 support vector regression models based on all possible input combinations. The results revealed that the exclusion of the viscosity of the base fluids (as a model input) leads to a significant improvement in the model result. To further validate our findings, we used the four inputs—volume fraction of nanoparticles, the density of nanoparticles, fluid temperature, and size of nanoparticles to build an ANN model. Based on the 428 testing datasets, the BSVR and ANN predicted the relative viscosity of nanofluids with an average absolute relative deviation of 3.22 and 6.64, respectively. This indicates that the BSVR model exhibits superior prediction results compared to the ANN model and existing empirical models. This study shows that the BSVR model is a reliable approach for the estimation of the viscosity of nanofluids. It also offers a generalization ability that is much better than ANN for predicting the relative viscosity of nanofluids. [ABSTRACT FROM AUTHOR]

The density of nanofluids is an important thermophysical property whose value is required to evaluate various heat-transfer parameters such as the Reynolds number, the Nusselt number, pressure loss, and the Darcy friction factor. The determination of these parameters is central to the design of many heat-transfer applications. Notably, the density of nanofluids has received relatively little research attention compared with other thermophysical properties. The present study thus focuses on the development of a support vector regression model to estimate the densities of aluminum nitride, titanium nitride, and silicon nitride nanoparticles dispersed in ethylene glycol solution. As inputs, the proposed model uses the mass fraction, temperature, nanoparticle size, and the molecular weight of the nanoparticles. The proposed model predicts the nanofluid densities with high accuracy, as determined by a correlation coefficient of 99.87% and an absolute average relative deviation of 0.0701. To further highlight the accuracy of the proposed model, we compare its results with those of the model of Pak and Cho. The Pak and Cho results deviate considerably from the experimental data except at 298 K. Overall, the proposed support vector regression model is much more accurate than the Pak and Cho model. We thus conclude that the machine learning approach is more reliable for obtaining rapid estimates of the density of nanofluids. [ABSTRACT FROM AUTHOR]

Lattice constant mismatch between materials affects the quality of thin film fabrication. For this reason, lattice constants information is vital in the design of materials for technological applications. The determination of lattice constants via experimental analysis is relatively expensive and laborious. As a result, several linear empirical models have been proposed to predict the lattice constant of crystal structures. However, the accuracies of these models are limited partly due to their failure to account for nonlinearity in the atomic parameters-lattice constant relationship. Machine learning techniques have shown excellent ability to deal with nonlinear problems in many areas of materials science; hence, they are considered suitable computation tools to study the crystal structure of materials. In this contribution, we developed a support vector regression (SVR) model to predict the lattice constant of cubic crystals of the form A2XY6 (A = K, Cs, Rb, TI; X = tetravalent cation; and Y = F, Cl, Br, I). The SVR algorithm uses the ionic radii and electronegativities data of the constituent elements of A2XY6 cubic crystals as model inputs. The robustness of the proposed model was demonstrated by comparing our result with an existing linear model based on 26 cubic crystal samples. The result revealed a total relative deviation of 1.757 and 2.704 for the SVR model and the existing linear equation, respectively. This result proves that the SVR model has a huge potential in the search for new materials for different applications. [ABSTRACT FROM AUTHOR]

Human exhaled breath has been utilized to identify biomarkers for diseases such as diabetes and cancer. The existence of these illnesses is indicated by a rise in the level of acetone in the breath. The development of sensing devices capable of identifying the onset of lung cancer or diabetes is critical for the successful monitoring and treatment of these diseases. The goal of this research is to prepare a novel breath acetone sensor made of Ag NPs/V2O5 thin film/Au NPs by combining DC/RF sputtering and post-annealing as synthesis methods. The produced material was characterized using X-ray diffraction (XRD), UV-Vis, Raman, and atomic force microscopy (AFM). The results revealed that the sensitivity to 50 ppm acetone of the Ag NPs/V2O5 thin film/Au NPs sensor was 96%, which is nearly twice and four times greater than the sensitivity of Ag NPs/V2O5 and pristine V2O5, respectively. This increase in sensitivity can be attributed to the engineering of the depletion layer of V2O5 through the double activation of the V2O5 thin films with uniform distribution of Au and Ag NPs that have different work function values. [ABSTRACT FROM AUTHOR]

The suspension of the nanoparticles in the conventional heat transfer fluids results in the increment of the fluids' thermal conductivity, viscosity, and density of the fluid. These improvements in thermophysical properties have enormous technological benefits. The thermal fluid properties depend upon a host of parameters such as volume fraction of the nanoparticles, temperature, density of fluid base and nanoparticles, nanoparticles size, nanolayer, the thermal conductivity of base fluid and particles, and pH. Studying the properties of the thermal fluids via experiments is a laborious task that can be simplified using predictive modeling. In this study, a machine learning regressive model named Gradient Boost Regression (GBR) is developed to predict the density of the aluminum nitride (Al2N3), silicon nitride (Si3N4), and titanium nitride (TiN) nanoparticles suspended in ethylene glycol (EG) solution. The predictive model developed is used to predict the density of the nanofluid based on four parameters; the size of nanoparticles, the molecular weights of nanoparticles, the volume concentration of the particle, and temperature. The proposed predictive model predicted the density of nanofluids for the training dataset with 99.99 percent accuracy whereas the testing dataset was validated at 99.91% accuracy and a standard deviation of 0.0000783 from the experimental data. To highlight the accuracy of the proposed model, its predictive performance was compared with an existing model formulated by Pak and Cho. The Pak & Cho model has a mean absolute deviation (MAD) of 0.0397 whereas the proposed GBR model has a MAD value of 0.0000783 for the estimated density of the nanofluids. Finally, the developed GBR model was compared with experimental density values at different temperatures and mass concentrations (0–5%). The proposed (GBR) model agrees well with experimental results under the different experimental conditions. [ABSTRACT FROM AUTHOR]

The lattice constants of spinel compounds AM2X4 are correlated with the constituent elemental properties using support vector regression (SVR) optimized with Bayesian optimization. A total of 185 spinel compounds of different groups were used for model development and validation. The ionic radii (RA, RX, RM) and electronegativity values χA, χM and χX) of constituent elements were considered as inputs for the model development. A systematic analysis of the impact of each of the inputs revealed that the set of inputs with the optimum prediction performance are RA, RX, RM, χA, χM. The generalization of the model was tested on 52 binary spinel compound datasets. The proposed model exhibited impressive accuracy as measured by high correction and low root mean square error of 99.50 percent and 0.1128 Armstrong, respectively. The extrapolating ability of the model was extended to new tertiary spinel compounds that comprise cation mixing at A and M sites, and ionic substitution on stable parent spinel structures/lattices. The results obtained are very promising which shows that in the search for new materials for different applications, the developed SVR model could be used for screening potential spinel compounds, consequently speeding up the process of material development and applications. [ABSTRACT FROM AUTHOR]