Performance evaluation of an ANN–GA aided experimental modeling and optimization procedure for enhanced synthesis of marine biosurfactant in a stirred tank reactor

BACKGROUND: An improved resilient back‐propagation neural network modeling coupled with genetic algorithm aided optimization technique was employed for optimizing the process variables to maximize lipopeptide biosurfactant production by marine Bacillus circulans. RESULTS: An artificial neural network (ANN) was used to develop a non‐linear model based on a 2⁴ full factorial central composite design involving four independent parameters, agitation, aeration, temperature and pH with biosurfactant concentration as the process output. The polynomial model was optimized to maximize lipopeptide biosurfactants concentration using a genetic algorithm (GA). The ranges and levels of these critical process parameters were determined through single‐factor‐at‐a‐time experimental strategy. Improved ANN‐GA modeling and optimization were performed using MATLAB v.7.6 and the experimental design was obtained using Design Expert v.7.0. The ANN model was developed using the advanced neural network architecture called resilient back‐propagation algorithm. CONCLUSION: Process optimization for maximum production of marine microbial surfactant involving ANN‐GA aided experimental modeling and optimization was successfully carried out as the predicted optimal conditions were well validated by performing actual fermentation experiments. Approximately 52% enhancement in biosurfactant concentration was achieved using the above‐mentioned optimization strategy. © 2012 Society of Chemical Industry     

Neural Network Modeling Applied to Polyacrylamide Based Hydrogels Synthetized By Single Step Process

This paper presents a series of experimental data obtained from the synthesis of polyacrylamide-based hydrogels and a general neural network methodology that accomplishes the modeling and optimization of the polymerization process.

Using direct neural network modeling, the variation of the main parameters in the synthesis of polyacrylamide-based hydrogels (polymerization yield and maximum swelling degree) was modeled in correlation with reactant concentrations, temperature, and reaction time. The predictions of the network, verified against initial training data and other testing data in the domain of the reaction conditions, were quite precise.

Inverse neural modeling determines, in a facile manner and with good results, the initial reaction conditions, which lead to a preestablished reaction yield and maximum swelling degree. This optimization method is more advantageous compared to a difficult classical procedure that requires a good mathematical model and an optimization solving technique.     

Computational study on reaction enthalpies of urethane‐forming reactions

Computational calculations were performed on urethane‐forming reactions using Gaussian 09 software (i.e. molecular modeling) toward the goal of providing thermodynamic parameters. Total electronic and thermal enthalpies and zero‐point vibrational energies of reactants and products were computed by the software and then reaction enthalpies were calculated based on these results. The location of functional groups has the most significant impact on reaction enthalpies while molecular size, chain length and solvent effect have relatively less impact on reaction enthalpies. By comparison to new experimental studies and values reported in literature, better‐informed recommendations on which values of reaction enthalpies to use for urethane foam process simulation were provided. The utility of computational chemistry results succeeded in being an enabling technology to improve foam process simulation. In turn, simulation of urethane‐forming reactions is useful to bridge the gap between fundamental computational chemistry calculations and practical applications. POLYM. ENG. SCI., 55:1420–1428, 2015. © 2015 Society of Plastics Engineers     

Modeling of glycolysis of flexible polyurethane foam wastes by artificial neural network methodology

The glycolysis process as a useful approach to recycling flexible polyurethane foam wastes is modeled in this work. To obtain high quality recycled polyol, the effects of influential processing and material parameters, i.e. process time, process temperature, catalyst‐to‐solvent (Cat/Sol) and solvent‐to‐foam (Sol/Foam) ratios, on the efficiency of the glycolysis reaction were investigated individually and simultaneously. For the continuous prediction of process behavior and interactive effects of parameters, an artificial neural network (ANN) model as an efficient statistical‐mathematical method has been developed. The results of modeling for the criteria that determine the glycolysis process efficiency including the hydroxyl value of the recycled polyol and isocyanate functional group conversion prove that the adopted ANN model successfully anticipates the recycling process responses over the whole range of experimental conditions. The Cat/Sol ratio showed the strongest influence on the quality of the recycled polyol among the studied parameters, where the minimum hydroxyl value was obtained at a medium amount of the assigned ratio. For the consumed polyurethane foam, a higher value of this ratio led to an increase in the hydroxyl value and isocyanate conversion. © 2015 Society of Chemical Industry     

Simulation of liquid physical blowing agents for forming rigid urethane foams

A computer‐based simulation for rigid polyurethane foam‐forming reactions was compared with experimental data for six blowing agents including methyl formate and C5‐C6 hydrocarbons. Evaporation of blowing agent was modeled as an overall mass transfer coefficient times the difference in activity of the blowing agent in the gas foam cells versus the resin walls of the cells. Successful modeling hinged upon use of a mass transfer coefficient that decreased to near zero as the foam resin approached its gel point. Modeling on density agreed with experimental measurements. The fitted parameters allowed for interpretations of the final disposition of the blowing agent, especially, if the blowing agent successfully led to larger foam cells versus being entrapped in the resin. The only component‐specific fitted parameters used in the modeling was the activity coefficient that was lower for methyl formate than the value used for hydrocarbons. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42454.     

A Multi‐Ionic Continuum‐Based Model for C3S Hydration

A computationally efficient strategy for modeling tricalcium silicate hydration based on through‐solution‐phase kinetics is demonstrated. This study extends a recently introduced advanced continuum‐based single particle model by including rigorous multi‐ionic transport, nonlinear reversible reaction kinetics and portlandite precipitation. Model parameters were either fixed based on known values, estimated using experimental measures or extracted by model fitting to benchmark experimental datasets. The model is now able to generate calorimetric hydration and evolution of pore solution chemistry responses that are in good agreement with available experimental results and predictions of other multiphysical modeling platforms. Once calibrated, the model was tested to see if it could predict the effect of water to cement ratio (w/c) and particle size on hydration outcomes. The findings support the need for a mechanism that limits the volume into which product can form.     

Artificial Neural Network Modeling of Mixing Efficiency in a Split-Cylinder Gas-Lift Bioreactor for Yarrowia lipolytica Suspensions

The mixing efficiency in a split-cylinder gas-lift bioreactor has been analyzed for Yarrowia lipolytica cells suspensions. Based on the experimental results, three different approaches for modeling have been applied to predict the mixing time depending on yeast concentration, aeration rate, as well as position on the riser or downcomer regions height. These approaches are represented by: an algorithm mixing differential evolution (DE) with artificial neural networks (ANNs), named hSADE-NN, regression, and the Multilayer Perceptron module from IBM SPSS. In the hSADE-NN, ANN models the process, while DE simultaneously optimizes the topology and the internal parameters of the ANN, so that an optimal model is obtained. It was found from simulations that ANNs are able to model the targeted process with a high degree of efficiency (average absolute relative error less than 8.5%), a small difference among the two ANN-based approaches being observed. Additionally, a sensitivity analysis was performed for determining the model inputs influence on the mixing time.     

Thermal degradation kinetics of a commercial epoxy resin—Comparative analysis of parameter estimation methods

The thermal degradation behavior of a commercial epoxy resin, EpoFix® (Struers), has been investigated by thermogravimetry (TG), differential thermal gravimetry (DTG), and differential thermal analysis (DTA) under nonisothermal conditions in an argon atmosphere. Different methods (Kissinger, Flynn–Wall–Ozawa (FWO), Friedman isoconversion methods, and nonlinear least‐squares (NLSQ) estimation method) have been used to analyze the thermal degradation process and determine the apparent kinetic parameters. The methods produce similar results in terms of activation energy estimations. Nevertheless, the NLSQ method has several advantages over the other methods in terms of both characterizing the activation energy and modeling the thermal degradation—i.e., including this model in a resin degradation process simulation. However, it is interesting to combine the NLSQ method with other isoconversion methods: they can reflect the dependence and variability of the activation energies during pyrolysis processes, while providing a good starting point for a nonlinear procedure, especially with respect to the activation energy E. This work is the first step (apparent kinetic reaction) of complete simulation of experimental oven of degradation of epoxy resin coating of impregnate nuclear fuel sample. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42201.     

The influence of solution parameters on the electrospinning intensity from foamed surface

Electrospinning is an efficient process for producing polymeric and hybrid nanofibers. There is, however, a lack of understanding concerning scalability of the process and in particular the production rate optimization. The electrospinning mass transfer intensity depends predominately on solution parameters, process parameters and the design of the equipment. These parameters influence the deposition intensity of the spinning process differently, but it is not known which factors dominate. The e‐spinning deposition intensity of polyethylene oxide, polyvinyl alcohol and their mixtures was investigated using a bubble foamed polymer solution surface to promote high mass deposition. Based on the measured properties of the solutions, a mathematical criterion was developed which made it possible to predict the electrospinning intensity of a given polymer solution. The proposed formula agrees with the experimental data and confirms that spinning intensity can be predicted from pre‐determined solution parameters. Using computer modeling, the weighting coefficients of the solution parameters have been determined, showing which parameter is the most important for the process intensity. The criterion and the same weighting coefficients were applied to the analysis of published data and it was found that they can be applied not only for electrospinning from the foamed surface but also from the free surface. A physical explanation of the criterion is proposed. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42034.     

Nonlinear viscoelastic modeling of soft polymers

A one‐dimensional phenomenological constitutive model, representing the nonlinear viscoelastic behavior of polymers is developed in this study. The proposed model is based on a modification of the well‐known three element standard solid model. The linear dashpot is replaced by an Eyring type one, while the nonlinearity is enhanced by a nonlinear, strain dependent spring constant. The new constitutive model was proved to be capable of capturing the main aspects of nonlinear viscoelastic response, namely, monotonic and cyclic loading, creep and stress relaxation, with the same parameter values. Model validation was tested on the experimental results at various modes of deformation for two elastomeric type materials, performed elsewhere. A very good agreement between model simulations and experimental data was obtained in all cases. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42141.     

Modeling heterogeneous photocatalytic inactivation of E. coli using suspended and immobilized TiO2 reactors

A study was carried out to develop a kinetic model of the photocatalytic inactivation of Escherichia coli using different TiO₂ catalysts. The model developed is based on a reaction scheme that involves effectively coupling mass‐transfer fluxes between bacteria and catalyst surface on one hand and bacterial degradation reaction on the other. The photocatalytic results were derived from experiments led in a batch reactor under both dark and Ultra Violet (UV) irradiation conditions. Using a reference catalyst, the robustness of the developed model was tested under solar conditions. The experimental data validated the model as successfully able to reproduce evolutions in the viable bacteria concentration in the range of parameters studied without any further adjustment of the kinetic parameters. The model was used to simulate the bacterial degradation kinetics under different working conditions to describe the partitioning of both bacterial adhesion and photocatalytic reaction in the solution to be treated © 2015 American Institute of Chemical Engineers AIChE J, 61: 2532–2542, 2015     

Nonisothermal curing kinetics of epoxy resin composite utilizing Ga (III) xanthate as a latent catalyst

Cure kinetics and curing mechanism of epoxy resin composite utilizing gallium (III) xanthate as a latent catalyst was investigated and compared with the commercial latent catalyst UCAT3512T formulation. Nonisothermal differential scanning calorimetric technique at different heating rates was employed to investigate the kinetic parameters. Activation energy was determined using Kissinger's and Flynn‐Wall‐Ozawa methods. Ga (III) xanthate was found to possess superior latent properties compared with UCAT3512T since the activation energy value obtained was higher for epoxy resin composite consisting of Ga (III) xanthate than UCAT3512T. Friedman's isoconversional method was utlizied to for kinetic modeling. An autocatalytic model was found to be successful in describing the curing reaction for both of the formulations. The calculated conversion rate as a function of temperature obtained by solving the autocatalytic equation showed a very good fit with experimental values. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42149.     

Quality control in the polypropylene manufacturing process: An efficient,data‐driven approach

In the propylene polymerization process, the melt index (MI), as a critical quality variable in determining the product specification, cannot be measured in real time. What we already know is that MI is influenced by a large number of process variables, such as the process temperature, pressure, and level of liquid, and a large amount of their data are routinely recorded by the distributed control system. An alternative data‐driven model was explored to online predict the MI, where the least squares support vector machine was responsible for establishing the complicated nonlinear relationship between the difficult‐to‐measure quality variable MI and those easy‐to‐measure process variables, whereas the independent component analysis and particle swarm optimization technique were structurally integrated into the model to tune the best values of the model parameters. Furthermore, an online correction strategy was specially devised to update the modeling data and adjust the model configuration parameters via adaptive behavior. The effectiveness of the designed data‐driven approach was illustrated by the inference of the MI in a real polypropylene manufacturing plant, and we achieved a root mean square error of 0.0320 and a standard deviation of 0.0288 on the testing dataset. This proved the good prediction accuracy and validity of the proposed data‐driven approach. © 2014 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 41312.     

Bifunctional cellulose derivatives for the removal of heavy‐metal ions and phenols: Synthesis and adsorption studies

In this article, we report a study of the design and synthesis of a bifunctional cellulose derivative on the removal of phenols and heavy‐metal ions in wastewater treatment. A radical polymerization was performed in an ionic liquid, 1‐allyl‐3‐methylimidazolium chloride, to graft two monomers, butyl methacrylate and 4‐vinyl pyridine, on the backbone of cellulose. The effects of the five reaction conditions on the yield of final products were evaluated. The grafted celluloses were characterized by means of Fourier transform infrared spectroscopy, scanning electron microscopy, and thermogravimetric analysis. Adsorption experiments were carried out on the cellulose‐g‐poly(butyl methacrylate‐co‐4‐vinyl pyridine) to evaluate the capacity of the removal of 2,4‐dichlorophenol (2,4‐DCP) and Cu(II) in water. The adsorption isotherms were measured at five temperatures and interpreted by a Langmuir model of adsorption. The thermodynamics of the adsorption suggested that the binding process was mildly exothermic for Cu(II) and endothermic for 2,4‐DCP. Kinetic studies were interpreted with a pseudo‐second‐order adsorption model. The process of the adsorption of 2,4‐DCP could be described overall by the model, whereas the adsorption of Cu(II) involved two processes. This was due to adsorption both on the surface and inside the adsorbent. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 41830.     

Detailed study on fusion characteristics of rigid poly(vinyl chloride) nanocomposites: The comparison of using multiple regression analysis and artificial neural network

Fusion behavior of poly(vinyl chloride) (PVC) compounds plays an important role in the development of physical properties of processed material. The fusion characteristics in PVC processing are governed by material variables that affect the fusion with some interactions. In this research, the aim was to characterize the effects of formulation ingredients on fusion characteristics of PVC. Four material parameters, including the contents of nanoclay (NC), azodicarbonamide, calcium stearate, and processing aid, are proposed as affecting variables. The fusion time (FT) as well as fusion factor (FF) are considered fusion indicators and are experimentally determined in some different levels of affecting parameters. The multivariable regression analysis (MRA) and the Artificial Neural Network (ANN) modeling are considered as two analytical methods. The regression analysis result for the FT denotes, in part, significant linear and quadratic effects of NC and also its significant interactions with azodicarbonamide and calcium stearate, whereas that of FF indicates only a linear effect of NC. ANN modeling is performed with a three‐layer (input, hidden, and output) neural network. The results of the comparison of the MRA and ANN predictions with experimental values are reported as the correlation coefficient (R²), mean‐square error, and mean absolute percentage error for both FF and FT parameters. The obtained values clearly denote that the ANN results are more precise and especially more general than those of MRA. However, in the case of FT, improvement of the ANN modeling is much greater than that of FF. J. VINYL ADDIT. TECHNOL., 21:147–155, 2015. © 2014 Society of Plastics Engineers     

A systematic methodology for kinetic modeling of chemical reactions applied to n‐hexane hydroisomerization

Kinetic modeling provides chemical engineers with a unique opportunity to better understand reaction kinetics in general and the underlying chemistry in particular. How to systematically approach a modeling assignment in chemical reaction kinetics is typically less clear, especially for novices in the field. The proposed modeling methodology pursues an adequate compromise between statistical significance and physical meaning of the kinetic model and the corresponding parameters and typically results in models of an appropriate complexity. It comprises the following activities: (1) data analysis, aiming at qualitative information on the reaction mechanism and corresponding rate equations, (2) model regression to quantify this information via optimal parameter values, and (3) validation of the statistical significance and physical meaning of the parameter estimates. This methodology is successfully applied to n‐hexane hydroisomerization on a bifunctional catalyst. © 2014 American Institute of Chemical Engineers AIChE J, 61: 880–892, 2015     

Multiphase transport modeling for vacuum drying of pharmaceutical products

Vacuum drying of active pharmaceutical ingredients (API) is an energy‐intensive process that is often a manufacturing bottleneck. A multiphase transport model to predict drying performance under the assumption that boiling is the dominant mechanism is developed. Laboratory scale drying experiments were performed over a range of temperatures and pressures using acetone as the solvent and glass beads of three different particle sizes to mimic APIs. A two‐phase transport model with the vapor and solid considered as one phase and the liquid treated as the second phase was capable of qualitatively reproducing the drying dynamics is found. Adjustable model parameters estimated from experimental data collected over a range of operating conditions exhibited trends that provided further insight into drying behavior. Boiling is the dominant mechanism in vacuum drying and our transport model captured the key physics of the process. © 2015 American Institute of Chemical Engineers AIChE J, 61: 3639–3655, 2015     

Numerical simulation on residual thickness of pipes with curved sections in water‐assisted co‐injection molding

The residual thicknesses of the skin and the inner layers are important quality indicators of water‐assisted co‐injection molding (WACIM) process or overflow WACIM (O‐WACIM) parts. At the curved section, the residual thicknesses change significantly. A numerical simulation program based on the computational fluid dynamics method was developed to simulate the O‐WACIM process. After the numerical simulation program was validated with the experimental results, it was used to study the effects of the bending radii and bending angles on the residual thicknesses of the skin and inner layers of O‐WACIM parts. The results showed that the penetration of the inner melt and water was always close to the inner concave side due to the higher local pressure gradient and temperature. The effects of processing parameters on the residual thicknesses of the skin and inner layers were investigated using the orthogonal simulation method. It was found that the residual thicknesses of the skin/inner layer at the inner concave/outer convex side are mainly influenced by different parameters. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 42468.     

Industrial melt index prediction with the ensemble anti‐outlier just‐in‐time Gaussian process regression modeling method

Several data‐driven soft sensors have been applied for online quality prediction in polymerization processes. However, industrial data samples often follow a non‐Gaussian distribution and contain some outliers. Additionally, a single model is insufficient to capture all of the characteristics in multiple grades. In this study, the support vector clustering (SVC)‐based outlier detection method was first used to better handle the nonlinearity and non‐Gaussianity in data samples. Then, SVC was integrated into the just‐in‐time Gaussian process regression (JGPR) modeling method to enhance the prediction reliability. A similar data set with fewer outliers was constructed to build a more reliable local SVC–JGPR prediction model. Moreover, an ensemble strategy was proposed to combine several local SVC–JGPR models with the prediction uncertainty. Finally, the historical data set was updated repetitively in a reasonable way. The prediction results in the industrial polymerization process show the superiority of the proposed method in terms of prediction accuracy and reliability. © 2015 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2015 , 132, 41958.     

Kinetic modeling of lipase‐mediated one‐pot chemo‐bio cascade synthesis of R‐1‐phenyl ethyl acetate starting from acetophenone

BACKGROUND: A systematic investigation of mutual interference between a hydrogenation catalyst, Pd/Al₂O₃, and an immobilized lipase in a one‐pot synthesis of R‐1‐phenyl ethyl acetate at 70 °C has been undertaken. This paper reports the kinetic modeling of lipase‐mediated chemo‐bio cascade synthesis of R‐1‐phenyl ethyl acetate starting from acetophenone. RESULTS: The kinetic results revealed that these catalysts were not acting independently but in concert. A mechanism which predicts the experimental observations for this reaction is proposed. CONCLUSION: The parameters of the kinetic model, which are in good agreement with the experimental data, were estimated through numerical data fitting. The reliability of the estimated parameters was analyzed using the Markov Chain Monte Carlo (MCMC) method. Copyright © 2009 Society of Chemical Industry