Search

A multi-block progressive modelling approach is proposed for enhanced fault isolation in batch processes. The unfolding of batch data typically leads to matrices with a large number of columns and this complicates contribution analysis. In order to rapidly focus fault isolation in batch processes, it would be desirable to employ multi-block modelling techniques. Multi-block model such as consensus principal component analysis (CPCA) can produce multiple monitoring charts for sub-blocks and block loadings and block scores can be obtained which can represent unique behaviour of each sub-block. CPCA model uses super score which is the same as score from normal principal component analysis (PCA) model and it does not produce enhanced monitoring performance. Multi-block PCA (MBPCA) model using block score for model calculation can represent sub-blocks’ character but block scores are obtained from super loading so it may not be the best way to describe sub-blocks. A new MBPCA model is proposed for better expression of each sub-block. Through progressive modelling and contribution analysis, variables related to or affected by the fault, as well as the associated time information, are gradually identified. This enables a fault propagation path being established. The proposed method is applied to a benchmark simulated penicillin production process, PenSim.

Atechnique forfaultlocalizationinbatchprocesses using progressiveprincipalcomponentanalysis(PCA)modeling isproposed inthis paper. APCA model isdeveloped from normal process operation dataand is used foronline process monitoring. Once a fault is detected by the PCA model, process variables that are related to the fault are identified using contribution analysis. The time information on when abnormalities occurred in these variables is identified using a time series plot of the squared prediction errors (SPE) on these variables. These variables are then removed and another PCA model is developed using the remainingvariables.IfthefaultybatchcannotbedetectedbythenewPCAmodel,thentheremainingvariablesarenotrelatedtothe fault.IfthefaultybatchcanstillbedetectedbythenewPCAmodel,thenfurthervariablesassociatedwiththefaultareidentifiedfrom SPE contribution analysis. The procedure is repeated until the faulty batch can no longer be detected using the remaining variables. Usingthetimeinformationonwhenabnormalitiespresentedinthevariablesassociatedwiththefault,faultpropagationpathscanbe establishedandtheoriginofthefaultcouldbetraced.Theproposedmethodistestedonabenchmarksimulatedfed-batchpenicillin production process, PenSim. The results demonstrate that the proposed method is particularly effective in isolating faults that have occurred on measured variables. For more complex faults that have occurred on unmeasured variables, the method can identify variables affected by the fault, and process knowledge is required to determine the fault.

A colloidal dispersion of indium tin oxide (ITO) nanoparticles in an organic solvent was achieved using a milling process in which β-diketones or titanate coupling agents were used as a dispersing agent. Isopropyl tri(N-ethylenediamino)ethyl titanate and 2,4-pentanedione were found to be the most suitable dispersants for stabilizing ITO nanoparticles in a mixed organic solvent. The secondary particle size of the colloidal ITO dispersion was markedly affected by operational parameters in the milling process. Stable dispersions of antimony-doped tin oxide (ATO) nanoparticles in organic solvent, using β-diketones as dispersants, were also prepared. The stable dispersions of ITO and ATO nanoparticles were used as conductive transparent film coatings on cathode-ray tube panels. We found that the electrical conductivity of the film coating was affected by the boiling point of the dispersion stabilizers.

Combinatory responses of proinflamamtory cytokines have been examined on the nitric oxide-mediated function in cultured mouse calvarial osteoblasts. Interleukin-1beta (IL-1beta) and tumor necrosis factor-alpha (TNF-alpha) induced iNOS gene expression and NO production, although these actions were inhibited by L-NG-monomethylarginine (L-NMMA) and decreased alkaline phosphatase (ALPase) activity. Furthermore, NO donors, sodium nitroprusside (SNP) and NONOate dose-dependently elevated ALPase activity. In contrast, transforming-growth factor-beta (TGF-beta) decreased NO production stimulated by IL-1beta, TNF-alpha and interferon-gamma (IFN-gamma). iNOS was expressed by mouse calvarial osteoblast cells after stimulation with IL-1beta, TNF-alpha, and IFN-gamma. Incubation of mouse calvarial osteoblast cells with the cytokines inhibited growth and ALPase activity. However, TGF-beta-treatment abolished these effects of IL-1beta, TNF-alpha and IFN-gamma on growth inhibition and stimulation of ALPase in mouse calvarial osteoblast cells. In contrast, IL-1beta, TNF-alpha, and IFN-gamma exerted growth-inhibiting effects on mouse calvarial osteoblast cells which were partly NO-dependent. The results suggest that NO may act predominantly as a modulator of cytokine-induced effects on mouse calvarial osteoblast cells and TGF-beta is a negative regulator of the NO production stimulated by IL-1beta, TNF-alpha and IFN-gamma.

Nanosized colloidal indium tin oxide (ITO) dispersion was prepared for electrically conductive and transparent coating materials. A titanate coupling agent, isopropyl tri(N-ethylenediamino)ethyl titanate, was chosen as a dispersant for the stabilization of ITO nanoparticles in organic solvent. ITO sol was deposited on a cathode ray tube panel for antistatic or electromagnetic shielding purposes, and alkyl silicate was used for the formation of an antireflective over-coat layer. The resulting double-layered coating showed low sheet resistance, which satisfied semi-TCO regulation and low reflectance of visible light. To control the electrical and optical properties of the coating layer, the effects of secondary particle size of ITO aggregates and the dispersant concentration of ITO sol were studied. The stability of ITO sol was estimated by measuring the particle size as a function of the storage days and the aggregation of colloidal ITO dispersion with storage day was explained by depletion flocculation.

The precipitation phase boundary of trisodium salt of 2-(2-carboxyethyl)-3-decyl maleic anhydride (DCMA-3Na) was determined with various metal ions, such as Ca +2 , Mg +2 , Zn +2 , Cu +2 , and Al +3 . The tricarboxylic-acid-type biosurfactant DCMA-3Na was synthesized from spiculisporic acid. Solid precipitates were analyzed by FT-IR and an optical microscope. The precipitation boundary diagrams were prepared through both the measurement of the supernatant solution concentration and the visual inspection of precipitate formation. The multivalent ion concentration necessary to cause the precipitation of the surfactant (i.e., tolerance) decreased as the DCMA-3Na concentration increased, up to a certain critical concentration in the same way as with chemically synthesized commercial surfactants such as SDS. However, unlike the commercial surfactants, the tolerance of DCMA-3Na toward hardness ions did not increase when the concentration of surfactant increased further above the critical point. This interesting phenomenon is due to the fact that excess amounts of sodium ions dissociated from the three carboxylate groups in the head group of DCMA-3Na compete with the multivalent ions over binding onto the micelle surface.

This paper considers capture of small particles on the surface of a much larger bubble by Brownian diffusion and interception induced by flow. The bubble is assumed to rise under the action of gravity, and the terminal rising velocity is determined for the bubble size typical of bubble flotation processes. The mechanism of particle transport to the surface of bubble is controlled by the ratio of particle (a) to bubble radius (A). The dominant transport mechanism changes from diffusion to interception as the size ratio δ(≡a/A) increases above a certain critical value. Although the Marangoni effect caused by the adsorbed particles reduces the collection efficiency, the general features of transport mechanism are preserved when the tangential velocity is nonzero on the bubble surface. Specifically, the analysis shows that for a bubble collector with slip boundary the critical value follows δcr ∝ A-4/3, in contrast to the result δcr ∝ A-1 for a no-slip collector. In the diffusion-control region, the collection efficiency of a collector with slip boundary decreases with increasing collector radius as A-3/2, which is much slower than A-2 for the case of a no-slip collector. On the other hand, in the interception-control region, the collection efficiency increases in proportion to the size ratio δ. In this limit, the efficiency of a collector with slip boundary is very large compared to the result for the case of a no-slip collector where the efficiency is only O (δ2).

The conventional process monitoring procedure using principal component analysis (PCA) can show which variable is highly related with the fault by looking at the contribution plots for the monitoring statistics, SPE (squared prediction errors) and T2. However, this procedure is not able to determine if the variable is just affected by the fault or the variable is the cause of the fault. In addition, it is not able to show fault propagation through the process variables during the process time. The proposed progressive PCA modeling procedure can identify all variables related to the fault through progressively removing the identified variables and PCA modeling with the remaining variables. It can also provide timing information of when abnormal behaviors are observed for the identified variables by using time series SPE plots with control limits estimated by weighted chi-squared distribution. Based on the timing information, it is able to build a flow chart showing the fault propagation paths. The proposed method is demonstrated on a benchmark fed-batch penicillin process simulator.

Stable dispersion of colloidal indium tin oxide nanoparticles was prepared by using indium tin oxide nanopowder, organic solvent, and suitable dispersants through attrition process. Various comminution parameters during the attrition step were studied to optimize the process for the stable dispersion of indium tin oxide sol. The transparent and conductive films were fabricated on glass substrate using the indium tin oxide sol by spin coating process. To obtain antireflective function, partially hydrolyzed alkyl silicate was deposited as over-coat layer on the pre-fabricated indium tin oxide film by spin coating technique. This double-layered structure of the nanostructured film was characterized by measuring the surface resistance and reflectance spectrum in the visible wavelength region. The final film structure was enough to satisfy the TCO regulations for EMI shielding purposes.

Quality prediction is usually required for product quality monitoring and setting up control strategy can reduce operating cost and improve production efficiency. Partial least square (PLS) regression is a popular statistical method for predictive modelling. The amount of data measured and stored in a typical industrial process is dramatically increased due to the fast development of computer and measuring system. It is hard to analyse all measured data using one matrix for its complexity. Multi-Block PLS model allows the data to be separated into sub-blocks and the sub-blocks can be analysed independently. Data from the fed-batch fermentation process is used to build models. Data is divided by different modes and different phases and model parameters are used to select variables that can be used as good predictors. The new set of data after variable selections is used to build a new model again. In most cases, new models show improved prediction performances compared with results from the conventional method.

The present paper is concerned with the motion of fluid layer between two parallel concentric circular plates, when the inertia of fluid is not negligible. We consider the two specific problems to examine the inertia effects; one in which an incompressible Newtonian fluid is injected into the gap through the hole located at the center of each plate and the other in which two parallel plates rotate coaxially with arbitrary angular velocities. The method of solution is an asymptotic expansion which is usually employed for the thin-film lubrication problem in the limit of small but finite Reynolds number based on the gap height. The asymptotic solutions for the two problems considered here provide the inertia-induced secondary flow patterns, which in turn determine the inertia contributions to the flow parameters such as the pressure drop, injection flow rate, and torque required to sustain the rotation of each plate.

Multi block PLS (MBPLS) models have been reported to handle the complexity of data which is difficult to anlayse with the conventional PLS models. However, the conventional MBPLS models do not offer improved predictive modeling in terms of prediction performance. A data partitioning method for enhanced predictive process modeling is proposed in this paper and is called Time-MBPLS. It enables data to be separated into blocks by different measuring time. Model parameters can be used to express contributions for all variables in a certain time period and it is possible to determine data blocking structures according to variable impacts on the quality variables. The proposed method is applied to the inferential estimation of a quality variable in a bio-production process. The results demonstrate that the proposed method can improve model prediction performance on the unseen testing data