{"title":"Machine learning-aided tailoring of double-emulsions within double-T microchannel","authors":"Saeed Ghasemzade Bariki, Salman Movahedirad, Mohadeseh Babaei layaei","doi":"10.1007/s10404-024-02758-4","DOIUrl":null,"url":null,"abstract":"<div><p>The formation of double-emulsions or core/shell microdroplets in microchannels, essential for various chemical applications, traditionally relies on costly and time-consuming laboratory methods. In this regard, computational fluid dynamics (CFD) and artificial neural network (ANN) techniques were employed. The present study developed ANN models to predict the relationship between shell thickness and double-emulsion size in a double-T microchannel, using two datasets comprising 180 experimental and CFD data points. Assessing this relationship involved analyzing various input factors, including the Capillary, Weber (case A), and Reynolds numbers (case B) of the core, shell, and continuous phases. Among twelve training algorithms and four activation functions, the Levenberg–Marquardt (LM) algorithm with sigmoidal activation functions (Tansig and Logsig), in contrast to the linear activation functions (Poslin and Purelin), achieved the highest predictive accuracy. Additionally, the predictive accuracy of ANN models was found to be significantly improved when trained using a combination of capillary and Weber numbers, as opposed to models trained only using capillary, Weber, and Reynolds numbers. The optimal neural network architectures were [10 5] neurons for case A (tansig and logsig) and [8] neurons for case B (tansig), yielding coefficients of determination (R<sup>2</sup>) of 0.99 and 0.98, respectively. These models demonstrated high precision and effective generalization, evidenced by statistical measures such as R<sup>2</sup>, MSE, RMSE, AAD, %AARD, and computational time. Moreover, their ability to generalize within the training dataset further substantiates their predictive capacity.</p></div>","PeriodicalId":706,"journal":{"name":"Microfluidics and Nanofluidics","volume":null,"pages":null},"PeriodicalIF":2.3000,"publicationDate":"2024-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Microfluidics and Nanofluidics","FirstCategoryId":"5","ListUrlMain":"https://link.springer.com/article/10.1007/s10404-024-02758-4","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"INSTRUMENTS & INSTRUMENTATION","Score":null,"Total":0}
引用次数: 0
Abstract
The formation of double-emulsions or core/shell microdroplets in microchannels, essential for various chemical applications, traditionally relies on costly and time-consuming laboratory methods. In this regard, computational fluid dynamics (CFD) and artificial neural network (ANN) techniques were employed. The present study developed ANN models to predict the relationship between shell thickness and double-emulsion size in a double-T microchannel, using two datasets comprising 180 experimental and CFD data points. Assessing this relationship involved analyzing various input factors, including the Capillary, Weber (case A), and Reynolds numbers (case B) of the core, shell, and continuous phases. Among twelve training algorithms and four activation functions, the Levenberg–Marquardt (LM) algorithm with sigmoidal activation functions (Tansig and Logsig), in contrast to the linear activation functions (Poslin and Purelin), achieved the highest predictive accuracy. Additionally, the predictive accuracy of ANN models was found to be significantly improved when trained using a combination of capillary and Weber numbers, as opposed to models trained only using capillary, Weber, and Reynolds numbers. The optimal neural network architectures were [10 5] neurons for case A (tansig and logsig) and [8] neurons for case B (tansig), yielding coefficients of determination (R2) of 0.99 and 0.98, respectively. These models demonstrated high precision and effective generalization, evidenced by statistical measures such as R2, MSE, RMSE, AAD, %AARD, and computational time. Moreover, their ability to generalize within the training dataset further substantiates their predictive capacity.
期刊介绍:
Microfluidics and Nanofluidics is an international peer-reviewed journal that aims to publish papers in all aspects of microfluidics, nanofluidics and lab-on-a-chip science and technology. The objectives of the journal are to (1) provide an overview of the current state of the research and development in microfluidics, nanofluidics and lab-on-a-chip devices, (2) improve the fundamental understanding of microfluidic and nanofluidic phenomena, and (3) discuss applications of microfluidics, nanofluidics and lab-on-a-chip devices. Topics covered in this journal include:
1.000 Fundamental principles of micro- and nanoscale phenomena like,
flow, mass transport and reactions
3.000 Theoretical models and numerical simulation with experimental and/or analytical proof
4.000 Novel measurement & characterization technologies
5.000 Devices (actuators and sensors)
6.000 New unit-operations for dedicated microfluidic platforms
7.000 Lab-on-a-Chip applications
8.000 Microfabrication technologies and materials
Please note, Microfluidics and Nanofluidics does not publish manuscripts studying pure microscale heat transfer since there are many journals that cover this field of research (Journal of Heat Transfer, Journal of Heat and Mass Transfer, Journal of Heat and Fluid Flow, etc.).