Search

Smart at‐ or online process sensors, which employ machine learning (ML) to process data, have been the subject of extensive research in recent years, due to their potential for real‐time process control. In this paper, a passive acoustic emission process sensor has been used to detect gas–liquid regimes within a stirred, aerated vessel using novel ML approaches. Pressure fluctuations (acoustic emissions) in an air‐water system were recorded using a piezoelectric sensor installed on the external wall of three identical cylindrical tanks of diameter, T = 160 mm, filled to a volume of 5 L (height, H = 1.5 T). The tanks were made of either glass, steel, or aluminium, and each tank was equipped with a Rushton turbine of diameter, D = 0.35 T. The investigated flow regimes, flooding, loading, complete dispersion, and un‐gassed, were obtained by changing the air feed flow rates and by varying the impeller speed. The acoustic spectra obtained were processed to select an optimal number of features characterizing each of the regimes, and these were used to train three different ML algorithms. The pre‐processing includes a principal component analysis (PCA) step, which reduces the volume of data fed to the ML algorithms, saving computational time up to a factor of 5. The algorithms (decision tree, k‐nearest neighbour, and support vector machines) were challenged to use these features to identify the correct flow regime. Accurate predictions of the three gas–liquid regimes of interest have been achieved. The accuracy of the prediction ranges from 90% to 99%, and this difference is related to the material used for the vessel. [ABSTRACT FROM AUTHOR]

The increasing need to improve the sustainability of industrial processes requires more flexible and intensified solutions. For this purpose, nowadays lots of efforts are made to switch from batch to continuous processes, the latter being able to ensure the same processing history to all fluid elements, with a consequent better control of the operating conditions and product quality. The present work aims at developing a continuous flow reactor for the production of several fine chemicals, including medical‐surgical aids, but also other substances for specific industrial sectors. The plant is basically an inline reactor equipped with various static mixers and side inlets, and it is conceived to ensure on‐site production. This is an important feature also in light of the recent COVID‐19 pandemic, which asked for flexible and distributed production of chemicals. Numerical simulations based on computational fluid dynamics are employed to study the performance, in terms of pressure drops and degree of mixing, of different static mixers, that is, the Lightnin Inliner Series 50 and Ross low pressure drop (LPD), combining various elements of mixing and injections in different operating conditions in both laminar and turbulent regimes. The results highlighted how numerical simulations may represent a valid tool for supporting the detailed design of such flow reactors by allowing the evaluation of the optimal design solutions. [ABSTRACT FROM AUTHOR]

Planar laser‐induced fluorescence (PLIF) and electrical resistance tomography (ERT) were applied simultaneously to monitor the mixing performance of a KM static mixer for the blending of non‐Newtonian fluids of dissimilar rheologies in the laminar regime. The areal distribution method was used to obtain quantitative information from the ERT tomograms and the PLIF images. Comparison of the ERT and PLIF results demonstrates the ability of ERT to detect the mixing performance in cases of poor mixing within the resolution of the measurement, though the accuracy decreases as the condition of perfect mixing is approached. Thus, ERT has the potential to detect poor mixing within the confines of its resolution limit and the required conductivity contrast, providing potential rapid at‐line measurement for industrial practitioners. [ABSTRACT FROM AUTHOR]

Computational fluid dynamics (CFD) is very appealing to investigate mixing and reaction in microdevices, as it allows easily investigating different operating conditions as well as mixer geometries. This latter aspect is very important as the flow in microdevices is laminar so the mixing between reactants should be promoted by a clever mixer design, aimed at breaking the flow symmetries. Recently time periodic motions that improve mixing have been observed to take place in a T‐junction at low Reynolds numbers. In this case the numerical modelling should be based on direct numerical simulations (DNS), thus involving high computational resources. In this work, two different CFD approaches, i.e., finite volume and spectral element methods, are applied and compared for the analysis of the mixing process in the well known T‐shaped micromixer. Spectral elements methods are particularly suited for DNS; however, they have been scarcely applied to study micromixers, while plenty of works can be found with finite volume methods. The analysis is carried out using both ideal and non‐ideal liquid binary mixtures, the latter presenting a negative fluidity of mixing (i.e., the viscosity of the mixture is higher than that of the pure components). Moreover the numerical results are validated with simple flow visualization experiments. [ABSTRACT FROM AUTHOR]