Particles occur in almost all processes in chemical and life sciences. The particle size and shape influence the process performance and product quality, and in turn they are influenced by the flow behavior of the par...Particles occur in almost all processes in chemical and life sciences. The particle size and shape influence the process performance and product quality, and in turn they are influenced by the flow behavior of the particles during production. Monitoring and controlling such characteristics in multiphase systems to obtain sufficient qualities will greatly facilitate the achievement of reproducible and defined distributions. So far, obtaining this information inline has been challenging, because existing instruments lack measurement precision, being unable to process overlapping signals from different particle phases in highly concentrated multiphase systems. However, recent advances in photo-optics made it possible to monitor such features(particle size distribution(PSD), aspect ratio and particle concentration) with advanced image analysis(IA) in real-time. New analysis workflows as well as single feature extractions from the images using multiple image analysis algorithms allowed the precise real-time measurements of size, shape and concentration of particle collectives even separated from each other in three phase systems. The performances, advantages and drawbacks with other non-photo-optical methods for assessing the particle size distribution are compared and discussed.展开更多
An approach based on artificial neural network （ANN） is used to develop predictive relations between hydrodynamic inline force on a vertical cylinder and some effective parameters. The data used to calibrate and val...An approach based on artificial neural network （ANN） is used to develop predictive relations between hydrodynamic inline force on a vertical cylinder and some effective parameters. The data used to calibrate and validate the ANN models are obtained from an experiment. Multilayer feed-forward neural networks that are trained with the back-propagation algorithm are constructed by use of three design parameters （i.e. wave surface height, horizontal and vertical velocities） as network inputs and the ultimate inline force as the only output. A sensitivity analysis is conducted on the ANN models to investigate the generalization ability （robustness） of the developed models, and predictions from the ANN models are compared to those obtained from Morison equation which is usually used to determine inline force as a computational method. With the existing data, it is found that least square method （LSM） gives less error in determining drag and inertia coefficients of Morison equation. With regard to the predicted results agreeing with calculations achieved from Morison equation that used LSM method, neural network has high efficiency considering its convenience, simplicity and promptitude. The outcome of this study can contribute to reducing the errors in predicting hydrodynamic inline force by use of ANN and to improve the reliability of that in comparison with the more practical state of Morison equation. Therefore, this method can be applied to relevant engineering projects with satisfactory results展开更多
This study focuses on the transitions in flow states around two-, three-and four-inline square cylinders under the effect of Reynolds numbers at two different gap spacing values using the lattice Boltzmann method. For...This study focuses on the transitions in flow states around two-, three-and four-inline square cylinders under the effect of Reynolds numbers at two different gap spacing values using the lattice Boltzmann method. For this purpose, Reynolds number is varied in the range 1–130 while two different values of spacing taken into account are gap spacing =2 and 5. Before going to actual problem, the code is tested for flow around a single square cylinder by comparing the results with experimental and numerical results of other researchers, and good agreement is found.The current numerical computations yield that for both spacing values and all combinations of cylinders there exist three different sates of flow depending on Reynolds numbers: steady state, transitional state and unsteady state. It is found that the range of Reynolds numbers for these flow states is different for both spacing values. At gap spacing =2 the range of Reynolds numbers for each flow state decreases by increasing the number of cylinders while at gap spacing =5 opposite trend is observed. The results also show that at gap spacing =2 the reduction in drag force is greater than the corresponding reduction at gap spacing =5. The maximum reduction in drag force is observed at Reynolds numbers =1 at both spacing values. Similarly, at both spacing values and all Reynolds numbers, the maximum reduction in drag force is observed for the case of four-inline square cylinders.展开更多
The objective of present work is to find out the sources of fluid-borne noise in vertical inline pump for various flow rates. The three-dimensional unsteady Reynolds Average Navier Stokes equation was solved using com...The objective of present work is to find out the sources of fluid-borne noise in vertical inline pump for various flow rates. The three-dimensional unsteady Reynolds Average Navier Stokes equation was solved using computational fluid dynamics code to predict the acoustic distribution. The pump chosen for study was of low specific speed and the experimental performance characteristic was very well matched with computational head developed. PROUDMAN sound power contour analysis showed the critical zone of noise in inlet pipe,impeller,and volute. Based on this,the variations of acoustic power were depicted over the cross section of inlet pipe,along the mean streamline of inlet pipe,as well along the volute circumference. The result concludes that the predominant flow noise is at tongue region and followed by noise generated due to turbulence in inlet pipe which occurs by the sudden variation in flow passage as well it depends on the operating condition of pump. The frequency analysis gives a glimpse of understanding about the broadband noise distribution due to flow phenomenon over a frequency range.展开更多
The total inline wave forces, the irregular wave forces in particular, on an isolated pile are investigated by experiment. The relationships between force coefficients Cd and CM including in Morison's Eq. . and KC...The total inline wave forces, the irregular wave forces in particular, on an isolated pile are investigated by experiment. The relationships between force coefficients Cd and CM including in Morison's Eq. . and KC number or Reynolds number Re, and the variation of Cd and Cm in frequency domain are analysed with the method of least-squares in time domain and that of cross-spectral analysis. The plots of C4and Cmversus KCare given for both regular and irregular waves and those for irregular waves are used for numerical simulation of the irregular wave forces on the vertical pile and the results are in fairly good agreement with the test data. Based on the experimental results , the applicability of the spectral analysis method for calculating irregular wave forces on an isolated pile is investigated with the coherency γ between wave and wave forces and with KC number.展开更多
A novel low-coherence digital inline holographic microscope for accurate three-dimensional(3D)position estimation and nanoparticle classification is proposed and validated.Two low-coherence digital inline holograms of...A novel low-coherence digital inline holographic microscope for accurate three-dimensional(3D)position estimation and nanoparticle classification is proposed and validated.Two low-coherence digital inline holograms of a sample containing numerous nanoparticles,generated by two illumination light beams forming a small angle with each other from a low-coherence light source,are employed to determine the nanoparticles’actual 3D positions.Each nanoparticle’s sub-holograms,extracted from the holograms of the sample,are used to reconstruct the intensity scattering image at its respective actual position using the Rayleigh–Sommerfeld backpropagation method.The intensity scattering image of each nanoparticle is then used to classify particles with similar sizes and shapes.The advantages of the proposed system include rapid and highly accurate 3D nanoparticle position determination and nanoparticle classification without the need to pre-prepare patterns or have prior knowledge of the nanoparticle characteristics.展开更多
Aflatoxin B1(AFB1)is a toxic fungal metabolite that contaminates almonds from cultivation to harvesting.It leads to chronic health problems and significant economic loss to the producers.Therefore,a fast and non-invas...Aflatoxin B1(AFB1)is a toxic fungal metabolite that contaminates almonds from cultivation to harvesting.It leads to chronic health problems and significant economic loss to the producers.Therefore,a fast and non-invasive detection technique is crucial for safeguarding food safety by swiftly identifying and eliminating contaminated almonds from the supply chain.Hyperspectral imaging has been explored as a potential non-destructive technology for detecting AFB1.However,the diverse geometries of almonds present a significant challenge on acquired images,thereby impacting the accuracy of the developed prediction and classification models.This study investigates the effectiveness of short-wave infrared(SwIR)hyperspectral imaging combined with deep learning for detecting AFB1 in almonds of varying geometries.Initially,partial least squares regression(PLSR)and support vector machine(SvM)regression models were evaluated for quantification,while SVM and quadratic discriminant analysis(QDA)classifiers were applied for classification.The results indicated that spectral responses varied with almond thickness,making quantification models unreliable for industrial applications.The Competitive Adaptive Reweighted Sampling(CARS)algorithm was employed to identify key spectral features for developing multi-spectral AFB1 classification models to evaluate the feasibility of high-speed,accurate in-line detection.The deep learning approach significantly outperformed traditional machine learning models,with the pre-trained Inception V3 network achieving a cross-validation accuracy of 84.82%,an F1-score of 0.8522,and an area under curve of 0.893.These findings highlight the superiority of deep learning-based hyperspectral imaging for accurate and reliable AFB1 detection in almonds with diverse shapes and thicknesses.展开更多
基金financially supported by the grants for the project "Smart Process Inspection" (funding code ZF4184501CR5) from the "Zentrales Innovationsprogramm Mittelstand" (ZIM)
文摘Particles occur in almost all processes in chemical and life sciences. The particle size and shape influence the process performance and product quality, and in turn they are influenced by the flow behavior of the particles during production. Monitoring and controlling such characteristics in multiphase systems to obtain sufficient qualities will greatly facilitate the achievement of reproducible and defined distributions. So far, obtaining this information inline has been challenging, because existing instruments lack measurement precision, being unable to process overlapping signals from different particle phases in highly concentrated multiphase systems. However, recent advances in photo-optics made it possible to monitor such features(particle size distribution(PSD), aspect ratio and particle concentration) with advanced image analysis(IA) in real-time. New analysis workflows as well as single feature extractions from the images using multiple image analysis algorithms allowed the precise real-time measurements of size, shape and concentration of particle collectives even separated from each other in three phase systems. The performances, advantages and drawbacks with other non-photo-optical methods for assessing the particle size distribution are compared and discussed.
文摘An approach based on artificial neural network （ANN） is used to develop predictive relations between hydrodynamic inline force on a vertical cylinder and some effective parameters. The data used to calibrate and validate the ANN models are obtained from an experiment. Multilayer feed-forward neural networks that are trained with the back-propagation algorithm are constructed by use of three design parameters （i.e. wave surface height, horizontal and vertical velocities） as network inputs and the ultimate inline force as the only output. A sensitivity analysis is conducted on the ANN models to investigate the generalization ability （robustness） of the developed models, and predictions from the ANN models are compared to those obtained from Morison equation which is usually used to determine inline force as a computational method. With the existing data, it is found that least square method （LSM） gives less error in determining drag and inertia coefficients of Morison equation. With regard to the predicted results agreeing with calculations achieved from Morison equation that used LSM method, neural network has high efficiency considering its convenience, simplicity and promptitude. The outcome of this study can contribute to reducing the errors in predicting hydrodynamic inline force by use of ANN and to improve the reliability of that in comparison with the more practical state of Morison equation. Therefore, this method can be applied to relevant engineering projects with satisfactory results
文摘This study focuses on the transitions in flow states around two-, three-and four-inline square cylinders under the effect of Reynolds numbers at two different gap spacing values using the lattice Boltzmann method. For this purpose, Reynolds number is varied in the range 1–130 while two different values of spacing taken into account are gap spacing =2 and 5. Before going to actual problem, the code is tested for flow around a single square cylinder by comparing the results with experimental and numerical results of other researchers, and good agreement is found.The current numerical computations yield that for both spacing values and all combinations of cylinders there exist three different sates of flow depending on Reynolds numbers: steady state, transitional state and unsteady state. It is found that the range of Reynolds numbers for these flow states is different for both spacing values. At gap spacing =2 the range of Reynolds numbers for each flow state decreases by increasing the number of cylinders while at gap spacing =5 opposite trend is observed. The results also show that at gap spacing =2 the reduction in drag force is greater than the corresponding reduction at gap spacing =5. The maximum reduction in drag force is observed at Reynolds numbers =1 at both spacing values. Similarly, at both spacing values and all Reynolds numbers, the maximum reduction in drag force is observed for the case of four-inline square cylinders.
基金State Key Program of National Natural Science Foundation of China(51239005)
文摘The objective of present work is to find out the sources of fluid-borne noise in vertical inline pump for various flow rates. The three-dimensional unsteady Reynolds Average Navier Stokes equation was solved using computational fluid dynamics code to predict the acoustic distribution. The pump chosen for study was of low specific speed and the experimental performance characteristic was very well matched with computational head developed. PROUDMAN sound power contour analysis showed the critical zone of noise in inlet pipe,impeller,and volute. Based on this,the variations of acoustic power were depicted over the cross section of inlet pipe,along the mean streamline of inlet pipe,as well along the volute circumference. The result concludes that the predominant flow noise is at tongue region and followed by noise generated due to turbulence in inlet pipe which occurs by the sudden variation in flow passage as well it depends on the operating condition of pump. The frequency analysis gives a glimpse of understanding about the broadband noise distribution due to flow phenomenon over a frequency range.
文摘The total inline wave forces, the irregular wave forces in particular, on an isolated pile are investigated by experiment. The relationships between force coefficients Cd and CM including in Morison's Eq. . and KC number or Reynolds number Re, and the variation of Cd and Cm in frequency domain are analysed with the method of least-squares in time domain and that of cross-spectral analysis. The plots of C4and Cmversus KCare given for both regular and irregular waves and those for irregular waves are used for numerical simulation of the irregular wave forces on the vertical pile and the results are in fairly good agreement with the test data. Based on the experimental results , the applicability of the spectral analysis method for calculating irregular wave forces on an isolated pile is investigated with the coherency γ between wave and wave forces and with KC number.
基金funded by the Vietnam Ministry of Education and Training(Project No.B2025 BKA-11).
文摘A novel low-coherence digital inline holographic microscope for accurate three-dimensional(3D)position estimation and nanoparticle classification is proposed and validated.Two low-coherence digital inline holograms of a sample containing numerous nanoparticles,generated by two illumination light beams forming a small angle with each other from a low-coherence light source,are employed to determine the nanoparticles’actual 3D positions.Each nanoparticle’s sub-holograms,extracted from the holograms of the sample,are used to reconstruct the intensity scattering image at its respective actual position using the Rayleigh–Sommerfeld backpropagation method.The intensity scattering image of each nanoparticle is then used to classify particles with similar sizes and shapes.The advantages of the proposed system include rapid and highly accurate 3D nanoparticle position determination and nanoparticle classification without the need to pre-prepare patterns or have prior knowledge of the nanoparticle characteristics.
基金the Research Training Program International(RTPi)scholarship from Commonwealth Australiathe top-up scholarship provided by SureNut Ltd.SureNut Ltd.for supplying all the almonds used in this study.
文摘Aflatoxin B1(AFB1)is a toxic fungal metabolite that contaminates almonds from cultivation to harvesting.It leads to chronic health problems and significant economic loss to the producers.Therefore,a fast and non-invasive detection technique is crucial for safeguarding food safety by swiftly identifying and eliminating contaminated almonds from the supply chain.Hyperspectral imaging has been explored as a potential non-destructive technology for detecting AFB1.However,the diverse geometries of almonds present a significant challenge on acquired images,thereby impacting the accuracy of the developed prediction and classification models.This study investigates the effectiveness of short-wave infrared(SwIR)hyperspectral imaging combined with deep learning for detecting AFB1 in almonds of varying geometries.Initially,partial least squares regression(PLSR)and support vector machine(SvM)regression models were evaluated for quantification,while SVM and quadratic discriminant analysis(QDA)classifiers were applied for classification.The results indicated that spectral responses varied with almond thickness,making quantification models unreliable for industrial applications.The Competitive Adaptive Reweighted Sampling(CARS)algorithm was employed to identify key spectral features for developing multi-spectral AFB1 classification models to evaluate the feasibility of high-speed,accurate in-line detection.The deep learning approach significantly outperformed traditional machine learning models,with the pre-trained Inception V3 network achieving a cross-validation accuracy of 84.82%,an F1-score of 0.8522,and an area under curve of 0.893.These findings highlight the superiority of deep learning-based hyperspectral imaging for accurate and reliable AFB1 detection in almonds with diverse shapes and thicknesses.
