Empirical Likelihood for a Long Range Dependent Process Subordinated to a Gaussian Process

This article develops empirical likelihood methodology for a class of long range dependent processes driven by a stationary Gaussian process. We consider population parameters that are defined by estimating equations in the time domain. It is shown that the standard block empirical likelihood (BEL) method, with a suitable scaling, has a non‐standard limit distribution based on a multiple Wiener–Itô integral. Unlike the short memory time series case, the scaling constant involves unknown population quantities that may be difficult to estimate. Alternative versions of the empirical likelihood method, involving the expansive BEL (EBEL) methods are considered. It is shown that the EBEL renditions do not require an explicit scaling and, therefore, remove this undesirable feature of the standard BEL. However, the limit law involves the long memory parameter, which may be estimated from the data. Results from a moderately large simulation study on finite sample properties of tests and confidence intervals based on different empirical likelihood methods are also reported.     

Evaluating effects of applying different kinetic models to pyrolysis modeling of fiberglass‐reinforced polymer composites

This research evaluates the effects of applying different kinetic models (KMs), developed based on thermal analysis using thermogravimetric analysis data, when used in typical 1D pyrolysis models of fiberglass‐reinforced polymer (FRP) composites. The effect of different KMs is isolated from the FRP heating by conducting pyrolysis modeling based on measured temperature gradients. Mass loss rate simulations from this pyrolysis modeling with various KMs show changes in the simulations due to applying different KM approaches are minimal in general. Pyrolysis simulations with the most complex KM are conducted at several heat flux levels. Mass loss rate comparison shows there is good overlap between simulations and the experimental data at low incident heat fluxes. Comparison shows there is poor overlap at high incident heat fluxes. These results indicate that increasing complexity of KMs to be used in pyrolysis modeling is unnecessary for these FRP samples and that the basic assumption of considering thermal decomposition of each computational cell in comprehensive pyrolysis modeling as equivalent to that in a thermogravimetric analysis experiment becomes inapplicable at depth and higher heating rates. Copyright © 2014 John Wiley & Sons, Ltd.     

Inference for non‐stationary time‐series autoregression

The article considers simultaneous inference for a class of non‐stationary autoregressive models where the model parameters are allowed to vary smoothly over time. Simultaneous confidence tubes with asymptotically correct coverage probabilities are constructed to assess the overall patterns and magnitudes of the parameter functions over time. Simulation studies are conducted, and a real time‐series dataset is analyzed to demonstrate the usefulness of the proposed methodology.     

Evaluation of pyrolysis parameters for fiberglass reinforced polymer composites based on multi‐objective optimization

This study was conducted to investigate the ability of global, multi‐objective/variable optimization methods to estimate material parameters for comprehensive pyrolysis models—thermo‐physical and optical properties of two fiberglass reinforced polymer (FRP) composites that share the same fiberglass. With these optimization methods used in pair with a comprehensive pyrolysis model, parameter estimation was carefully conducted with considerations given to applying appropriate thermal decomposition kinetic models (three different models from simple to complex) and optimization targets (cone calorimeter data irradiated at 50 kW/m²). Estimation results are compared with independently measured effective properties—thermal conductivity, specific heat capacity, and emissivity of polymer resins and FRPs. Additionally, fiberglass properties estimated from the two FRPs are compared to analyze for consistency in optimized values. The results show that for a well‐configured parameter estimation exercise using the optimization method described earlier, (1) estimated results are within ±100% of the measurements in general and sometimes comparable to effective property values, (2) increasing complexity of the kinetic modeling for a single component system has insignificant effect on estimated values, and (3) increasing complexity of the kinetic modeling for a multiple component system with each element having different thermal characteristics has positive effect on estimated values. Copyright © 2014 John Wiley & Sons, Ltd.     

Integrating data‐based modeling and nonlinear control tools for batch process control

A data‐based multimodel approach is developed in this work for modeling batch systems in which multiple local linear models are identified using latent variable regression and combined using an appropriate weighting function that arises from fuzzy c‐means clustering. The resulting model is used to generate empirical reverse‐time reachability regions (RTRRs) (defined as the set of states from where the data‐based model can be driven inside a desired end‐point neighborhood of the system), which are subsequently incorporated in a predictive control design. Simulation results of a fed‐batch reactor system under proportional‐integral (PI) control and the proposed RTRR‐based design demonstrate the superior performance of the RTRR‐based design in both a fault‐free and faulty environment. The data‐based modeling methodology is then applied on a nylon‐6,6 batch polymerization process to design a trajectory tracking predictive controller. Closed‐loop simulation results illustrate the superior tracking performance of the proposed predictive controller over PI control. © 2011 American Institute of Chemical Engineers AIChE J, 2012     

Bioprocess scale‐up: quest for the parameters to be used as criterion to move from microreactors to lab‐scale

Advances in high‐throughput process development and optimization involve the rational use of miniaturized stirred bioreactors, instrumented shaken flasks and microtiter plates. As expected, each one provides different levels of control and monitoring, requiring a compromise between data quantity and quality. Despite recent advances, traditional shaken flasks with nominal volumes below 250 mL and microtiter plates are still widely used to assemble wide arrays of biotransformation/bioconversion data, because of their simplicity and low cost. These tools are key assets for faster process development and optimization, provided data are representative. Nonetheless, the design, development and implementation of bioprocesses can present variations depending on intrinsic characteristics of the overall process. For each particular process, an adequate and comprehensive approach has to be established, which includes pinpointing key issues required to ensure proper scale‐up. Recently, focus has been given to engineering characterization of systems in terms of mass transfer and hydrodynamics (through gaining insight into parameters such as k_La and P/V at shaken and microreactor scale), due to the widespread use of small‐scale reactors in the early developmental stages of bioconversion/biotransfomation processes. Within this review, engineering parameters used as criteria for scaling‐up fermentation/bioconversion processes are discussed. Particular focus is on the feasibility of the application of such parameters to small‐scale devices and concomitant use for scale‐up. Illustrative case studies are presented. Copyright © 2010 Society of Chemical Industry     

Towards a methodology for the characterization of fire resistive materials with respect to thermal performance models

A methodology is proposed for the characterization of fire resistive materials with respect to thermal performance models. Typically in these models, materials are characterized by their densities, heat capacities, thermal conductivities, and any enthalpies (of reaction or phase changes). For true performance modelling, these thermophysical properties need to be determined as a function of temperature for a wide temperature range from room temperature to over 1000°C. Here, a combined experimental/theoretical/modelling approach is proposed for providing these critical input parameters. Particularly, the relationship between the three‐dimensional microstructure of the fire resistive materials and their thermal conductivities is highlighted. Published in 2005 by John Wiley & Sons, Ltd.     

Optimization of industrial‐scale deodorization of high‐oleic sunflower oil via response surface methodology

Optimization of industrial‐scale deodorization of high‐oleic sunflower oil (HOSO) via response surface methodology is presented in this study. The results of an experimental program conducted on an industrial‐scale deodorizer were analyzed statistically. Predictive models were derived for each of the oil quality indicators (QI) in dependence on the studied variable deodorization process parameters. The deodorization behavior of some minor components was analyzed on a pilot‐scale deodorizer. For comparison, a similar experimental program was also performed on the laboratory‐scale. The results of this study demonstrate that optimization of the deodorization process requires a suitable compromise between often mutually opposing demands dictated by different oil QI. The production of HOSO with top‐quality organoleptic and nutritional values (high tocopherol and phytosterol contents and low free and trans fatty acid contents) and high oxidative stability demands deodorization temperatures in the range between 220 and 235 °C and a total sparge steam above 2.0% (wt/wt in oil). The response surface methodology provides the tools needed to identify the optimum deodorization process conditions. However, the laboratory‐scale experiments, while showing similar response characteristics of QI in dependence on the process parameters and thus helpful as a guide, are of limited value in the optimization of an industrial‐scale operation.     

Relationship between molecular structure and zero‐shear viscosity of polymers

The relationship between molecular structure and zero‐shear viscosity of polymers was studied. In this study we propose a new equation, which is based on Berry and Fox's equation. This new equation is constructed from some molecular parameters, such as mean square length and average molecular weight of statistical skeletal unit, characteristic ratio, entanglement molecular weight, glass‐transition temperature, free volume fraction at glass‐transition temperature, and thermal expansion coefficient of free volume. It is proposed that some of these molecular parameters could be predicted by group contribution methods, except for the free volume parameters. We also propose new empirical relations between free volume parameters and molecular structures of polymers, which make it possible for free volume parameters to be obtained from molecular structure. Using these relationships, it is possible that the zero‐shear viscosity and its temperature dependence are obtainable from the molecular structure of polymers. We applied this formula to some polymers, including both amorphous and semicrystalline polymers. Comparison between the measured and calculated zero‐shear viscosity showed quite good agreement. © 2001 John Wiley & Sons, Inc. J Appl Polym Sci 80: 1609–1618, 2001     

Experimental and numerical investigations on the thermal response of multilayer glass fibre/unsaturated polyester/organoclay composite

Organoclay glass fibre reinforced polymer (GFRP) nanocomposites are fabricated using the vacuum assisted resin transfer moulding. The unsaturated polyester resin is prepared with and without organoclay involving mechanical mixing, sonication, dilution solvent and heat treatment. Three levels of organophilic clay content are added, and its influences on the fire performance of composite samples are investigated. A novel numerical procedure combining pyrolysis analysis of the organoclay‐composites and the fire dynamic simulation of the combustion process are developed to validate the thermal responses obtained from the cone calorimetry experiments. Kinetic parameters obtained from the TGA tests and pyrolysis analyses are used as inputs for the models measuring the fire growth index and total heat release. To account for multilayer composite structure and organoclay distribution, three numerical models are proposed including composite (CPS), component (CPN) and CPN‐layer models. While CPS model assumes the homogeneity of the composite, later models consider multilayer effects with uniform (CPN model) or concentrated (CPN‐layer model) distribution of organoclay. Numerical results are compared with experimental ones in terms of total heat release, fire growth index. Finally, the fire resistance and total smoke release of the polyester/glass composites with the addition of organoclay will be evaluated taking into account influences of the fabrication processes.     

Experimental investigation and numerical modelling of the fire performance for epoxy resin carbon fibre composites of variable thicknesses

This paper applies a unique integrated approach to determine the flammability properties of a composite material (epoxy with carbon fibre) and compares its fire behaviour at two different thicknesses (2.1 and 4.2 mm) by performing small scale (thermo‐gravimetric analysis (TGA)/Fourier transform infrared radiation) and meso‐scale tests (cone calorimeter). For small‐scale tests, experiments were conducted in nitrogen using TGA coupled to gas analysis by Fourier transform infrared radiation. These results allow the determination of thermal stability, main degradation temperature and main gaseous emissions released during the thermal degradation. For meso‐scale tests, experiments were carried out using a cone calorimeter with sample dimensions of 100 × 100 mm at five heat fluxes (30, 40, 50, 60 and 70 kW/m²). The results show that the ignition time increases with an increase in the thickness of the material. Relative hazard classification of the fire performance of the current composites has also been compared with other materials using parameters obtained elsewhere. In addition, the effective ignition, thermal and pyrolysis properties obtained from the ignition and mass loss rate experiments for the 4.2‐mm thick samples were used in a numerical model for pyrolysis to predict well ignition times, back‐surface temperatures and mass pyrolysis rates for all heat fluxes as well as for the 2.1‐mm thick samples. Note that the ignition temperature obtained in the cone agrees with the main degradation temperature in the TGA. The flammability properties deduced here can be used to predict the heat release rate for real fire situations using CFD modelling. Copyright © 2016 John Wiley & Sons, Ltd.     

Thermal stability and pyrolysis kinetics of lignin‐phenol‐formaldehyde resins

A thermal stability and kinetic study from non‐isothermal experiments of a commercial and a lignin‐novolac resin mixed with two amounts of curing agent has been done employing thermogravimetric analysis technique. Three kinetic models have been tested: a single heating rate method, such as Coats‐Redfern, employing several mechanistic functions and contrasted with Van Krevelen—it is the first time that this method has been employed in polymer degradation. Finally, the Ozawa method allowed the obtaining of the activation energy by the multiple‐heating‐rate without knowing the mechanism. Results show that commercial mixtures of resins lose less weight than lignin‐novolac resins. The calculated kinetic parameters showed that Coats‐Redfern gives similar results to Van Krevelen, which means that these methods are adequate for novolac pyrolysis, and Ozawa shows activation energies in accordance with the last mentioned models. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

Flammability hazards of typical fuels used in wind turbine nacelle

This study aims to develop a complete methodology for assessing flammability hazards of typical fuels (ie, transformer oil, hydraulic oil, gear oil, and lubricating grease) used in a wind turbine nacelle by combining different experimental techniques such as thermogravimetric analysis and cone calorimetry. Pyrolysis properties (onset temperature, temperature of maximum mass loss rate, and mass residue) and reaction‐to‐fire properties (ignition time, heat release rate, mass loss rate, and smoke release rate) were determined and used for a preliminary assessment of thermal stability and flammability hazards. Additional indices, for ignition and thermal behavior (effective heat of combustion, average smoke yield, and smoke point height, heat release capacity, fire hazard parameter, and smoke parameter, were calculated to provide a more advanced assessment of the hazards in a wind turbine. Results show that pyrolysis of transformer oil, lubricating grease, hydraulic oil, and gear oil occur in the range of 150°C to 550°C. Lubricating grease and transformer oil show the higher and lower thermal stabilities with maximum pyrolysis rate temperatures of 471°C and 282°C, respectively. The measured relation between ignition time and radiant heat flux agrees well with Janssens method (a power of 0.55). The aforementioned indices appear to provide a reasonable prediction of performance under real fire conditions according to a full‐scale fire test documented by Declercq and Van Schevensteen. The results of the study indicate that transformer oil is the easiest to ignite while lubricating grease is the most difficult to ignite but also has the highest smoke production rate; that transformer oil has the highest heat release rate while gear oil has the lowest; and that the fire hazard parameter is the highest for transformer oil and the smoke parameter is the highest for lubricating grease. The potential of this type of work to design safer wind turbines under performance‐based approaches is clearly clarified.     

Experimental validation of CFD simulations of a lab‐scale fluidized‐bed reactor with and without side‐gas injection

Fluidized‐bed reactors are widely used in the biofuel industry for combustion, pyrolysis, and gasification processes. In this work, a lab‐scale fluidized‐bed reactor without and with side‐gas injection and filled with 500–600 μm glass beads is simulated using the computational fluid dynamics (CFD) code Fluent 6.3, and the results are compared to experimental data obtained using pressure measurements and 3D X‐ray computed tomography. An initial grid‐dependence CFD study is carried out using 2D simulations, and it is shown that a 4‐mm grid resolution is sufficient to capture the time‐ and spatial‐averaged local gas holdup in the lab‐scale reactor. Full 3D simulations are then compared with the experimental data on 2D vertical slices through the fluidized bed. Both the experiments and CFD simulations without side‐gas injection show that in the cross section of the fluidized bed there are two large off‐center symmetric regions in which the gas holdup is larger than in the center of the fluidized bed. The 3D simulations using the Syamlal‐O'Brien and Gidaspow drag models predict well the local gas holdup variation throughout the entire fluidized bed when compared to the experimental data. In comparison, simulations with the Wen‐Yu drag model generally over predict the local gas holdup. The agreement between experiments and simulations with side‐gas injection is generally good, where the side‐gas injection simulates the immediate volatilization of biomass. However, the effect of the side‐gas injection extends further into the fluidized bed in the experiments as compared to the simulations. Overall the simulations under predict the gas dispersion rate above the side‐gas injector. © 2009 American Institute of Chemical Engineers AIChE J, 2010     

Thermal‐Enzymatic Hydrolysis of Wheat Straw in a Single High Pressure Fixed Bed

An integrated flow‐through fixed‐bed reactor concept for lignocellulose separation at 3 L scale is presented. Wheat straw is degraded using sequential liquid hot water treatment and enzymatic hydrolysis in a single‐pot reactor setup. System responses on process parameters are modeled and optimized using response surface methodology. Accurate empiric models and significant factor influences could be identified for both treatment steps. With optimal factor settings, 23.9 and 49.2 wt % hot‐water‐soluble lignin and C5 oligomers could be recovered, respectively. 70.0 wt % cellulose solubilization could be achieved in the fixed‐bed enzymatic hydrolysis after 72 hours.     

A C1 microkinetic model for methane conversion to syngas on Rh/Al2O3

A microkinetic model capable of describing multiple processes related to the conversion of natural gas to syngas and hydrogen on Rh is derived. The parameters of microkinetic models are subject to (intrinsic) uncertainty arising from estimation. It is shown that intrinsic uncertainty could markedly affect even qualitative model predictions (e.g., the rate‐determining step). In order to render kinetic models predictive, we propose a hierarchical, data‐driven methodology, where microkinetic model analysis is combined with a comprehensive, kinetically relevant set of nearly isothermal experimental data. The new, thermodynamically consistent model is capable of predicting several processes, including methane steam and dry reforming, catalytic partial oxidation, H₂ and CO rich combustion, water‐gas shift and its reverse at different temperatures, space velocities, compositions and reactant dilutions, using the measured Rh dispersion as an input. Comparison with other microkinetic models is undertaken. Finally, an uncertainty analysis assesses the effect of intrinsic uncertainty and catalyst heterogeneity on the overall model predictions. © 2009 American Institute of Chemical Engineers AIChE J, 2009     

Time series analysis based on running Mann‐Whitney Z Statistics

A time series analysis method based on the calculation of Mann–Whitney U statistics is described. This method samples data rankings over running time windows, converts those samples to Mann–Whitney U statistics, and then normalizes the U statistics to Z statistics using Monte‐Carlo generated null parameters. Based on the Z statistics’ magnitudes this algorithm can identify time windows containing significant incidences of low or high data rankings, where the window length is determined by the sample size. By repeating this process with sampling windows of varying duration ranking regimes of arbitrary onset and duration can be objectively identified in a time series. The simplicity of the procedure's output – a time series’ most significant non‐overlapping ranking sequences – makes it possible to graphically identify common temporal breakpoints and patterns of variability in the analyses of multiple time series. This approach is demonstrated using United States annual temperature data during 1896–2008.     

Effects of the injection‐molding temperatures and pyrolysis cycles on the butadiene phase of high‐impact polystyrene

Recycling is a thermal process in which polymers are melted to produce new products. It is possible that these thermal processes could modify their mechanical and thermal properties. Polymer degradation can be characterized with thermogravimetric analysis and differential scanning calorimetry. Recycled materials tested with these methods have shown variations in some thermal properties, such as the glass‐transition temperature and thermal degradation onset, but the sensitivity of these methods is not sufficient to investigate the changes in the characteristics of polymers when materials are exposed to moderate temperature conditions or several thermal cycles. To study these structural changes, a much more sensitive technique, such as pyrolysis–gas chromatography/mass spectrometry (Py–GC/MS), is needed. Small variations in the structure can be determined by Py–GC/MS. Each pyrolysis product can be identified by its retention time and mass spectrum with the use of reference literature. In this work, we have studied structural changes in high‐impact polystyrene as a function of the injection‐molding temperature and pyrolysis cycles. The results do not show significant changes in samples processed at different temperatures with Py–GC/MS, but the values of the pyrolysis products differ as a function of the pyrolysis cycles. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci, 2007