Lin FChen P.-CHuang T.-CLin G.-STung K.-L.KUO-LUN TUNG2022-03-222022-03-22202103767388https://www.scopus.com/inward/record.uri?eid=2-s2.0-85104915070&doi=10.1016%2fj.memsci.2021.119380&partnerID=40&md5=d8876c6f54b083eaaff2f70e1c732422https://scholars.lib.ntu.edu.tw/handle/123456789/598245For the first time, a 3D Digital Twin virtual model of a hollow fiber membrane was established based on real data. The filtration process of ultrafine particles by using a hollow fiber membrane was also successfully analyzed with multiscale simulation. In this study, we investigated the difference between scanning electron microscopy (SEM) and X-ray computed tomography (X-CT) with regard to membrane analysis. In addition, the pore size distribution was calculated based on two different modules: the Granulometry module and the Porosimetry module. The results of the latter showed greater accuracy than the results of the experimental measurements. In the flow analysis, the velocity and pressure distribution were also obtained through simulations to investigate transport phenomena in membrane filtration. Moreover, the unprecedented technique of particle tracing was applied and successfully provided detailed information regarding how the particles were captured. Evaluations of the membrane performance, pressure drop, and filtration efficiency were conducted and compared to the experimental results with 2.25% and 0.01% errors. Our work showed the potential of the construction of 3D virtual models. These models not only exhibited high accuracy but also provided detailed information that is difficult to obtain via experiments. Through a combination of image analysis and multiscale simulation, a platform to evaluate membrane performance with high reliability and efficiency was established. ? 2021 Elsevier B.V.Image analysisMultiscale simulationParticle tracingPM2.5Ultrafine particles air filtrationComputerized tomographyEfficiencyMicrofiltrationParticle size analysisPore sizeReliability analysisScanning electron microscopyFiltration processGas filtrationHollow-fibre membraneImage-analysisMembrane performanceMulti-scale simulationPM$-2.5$Ultrafine particleUltrafine particle air filtrationMembranesaccuracyArticlecontrolled studydigital twinfiltrationfiltration efficiencyflow measurementflow rategasimage analysisparticulate matter 2.5pore size distributionpriority journalreliabilityscanning electron microscopyultrafine particulate matterx-ray computed tomographyFibersFiltrationImage AnalysisParticlesPore SizeScanning Electron Microscopyjournal article10.1016/j.memsci.2021.1193802-s2.0-85104915070