Ultrasonic in‐line monitoring of polymer extrusion

In‐line ultrasonic monitoring of polymer co‐extrusion and twin‐screw extrusion are presented. Co‐extrusion of high density polyethylene (HDPE) and a thermoplastic elastomer based on polypropylene‐EPDM (ethylene‐propylene‐diene monomer) has been investigated by ultrasonic sensors consisting of piezoelectric transducers and clad buffer rods. One extremity of the rod (probing end) was installed flush with the die surface so as not to disturb the material flow. The other end was air cooled in order to protect the transducer from excessive heating. This approach has been demonstrated to be quite convenient for monitoring and controlling industrial material processes: first, it can work at temperatures up to 1000°C; second, the clad buffer rod probing end can be machined to the same shape as those of commercial temperature and pressure sensors commonly used in the extrusion process. Therefore, no modifications are required for the installation in the original equipment. The information obtained includes the position of the interface between polymers and the stability of the process. The same ultrasonic probe has also been installed on a barrel of a twin‐screw extruder. This study was performed using polyethylene and polystyrene. It has been verified that the ultrasonic sensor can be successfully operated along the extruder screw and that the ultrasound can give access to the material properties while the polymer is being processed. This means that the technique can be exploited to monitor and control in situ the characteristics of the polymer being transformed in operations typically performed on twinscrew extruders, such as compounding, visbreaking or reactive extrusion.     

Dynamic modeling of blown‐film extrusion

Past dynamic studies of blown‐film extrusion have been confined to the stability analysis of the linearized equations. The full set of nonlinear equations comprises a system of partial differential and algebraic equations with boundary conditions that vary from author to author. In this paper, the Numerical‐Method‐of‐Lines, which combines finite‐difference methods with ordinary differential/algebraic equation integrators, is used to solve the full system. Appropriate boundary conditions are selected to give physical results that compare well with experiment. An important boundary condition is the “minimum order reduction” condition on the gradient of the bubble‐tube radius with respect to distance above the extrusion die (the axial position). Transient startups and operational disturbances are examined. Calculations show the influence of oscillations in operating conditions such as heat transfer or inflation pressure on the bubble‐tube radius and film thickness. Steady‐state results obtained by integrating the transient equations for a sufficiently long time are qualitatively in agreement with experiment, in contrast to past simulations of these equations.     

In‐line optical techniques to characterize the polymer extrusion

In the past 15 years our research group has been creating new optical devices to characterize in real time the extrusion process. These detectors are made of a slit‐die fitted at the extruder exit from where the molten polymer flows, with a pair of transparent windows that allows a light beam to pass orthogonally through the molten flow. Following the reduction of the transmitted light intensity one is able to quantify the turbidity, which is a function of the type, concentration and particle size and shape of the second phase present in the flow. By evaluating the scattering pattern of a laser beam (LALLS) it is possible to get information upon the morphology of the molten polymeric system in real time during the extrusion. With the interposition of a pair of crossed polarizers in the optical beam, rheo‐polarimetry, it is possible to evaluate quantitatively the flow birefringence, which is a function of the degree of the polymer matrix orientation. POLYM. ENG. SCI., 54:386–395, 2014. © 2013 Society of Plastics Engineers     

On‐line NIR sensing of CO2 concentration for polymer extrusion foaming processes

An on‐line sensor using near infrared (NIR) spectroscopy is developed for monitoring CO₂ concentration in polymeric extrusion foaming processes. NIR absorption spectra are acquired by a probe installed at the foaming extruder die. The calibration curve relating the absorbance spectrum at 2019 nm to the dissolved gas concentration is derived so as to infer dissolved CO₂ gas concentration on‐line from measured NIR spectra. Experimental results show the developed on‐line NIR sensor can successfully estimate dissolved CO₂ concentration in the molten polymer and illustrate that the developed NIR sensing technique is among the more promising methods for quality control of polymeric extrusion foaming processes.     

Shear-rate-dependence modeling of polymer melt viscosity

Steady-shear-viscosity data sets for commercial-grade acrylonitrile-butadiene-styrene terpolymer, nylon, polycarbonate, poly(methylmethacrylate), and polystyrence are fitted in terms of a generalized Cross/Carreau modeling for the shear-rate dependence. Based upon extensive data sets from the open literature as well as in-house measurements, it is shown that the shear-rate dependence can be more accurately described in terms of the Cross rather than Carreau model. Although the resulting viscosity fits based upon these two models might differ by 20% or more for the same well-characterized data set, the resulting effect upon simulating the injection-molding process is found to be much smaller since such predictions reflect a range of shear stresses (varying linearly from centerline to wall of cavity) over which the two models alternate in relative magnitude. This is demonstrated by detailed representative numerical predictions which are presented for both the filling and post-filling stages.     

Polynomial ARMA model identification for a continuous styrene polymerization reactor using on‐line measurements of polymer properties

The multiinput–multioutput identification for a continuous styrene polymerization reactor using a polynomial ARMA model is carried out by both simulation and experiment. The pseudorandom multilevel input signals are applied for model identification in which input variables are the jacket inlet temperature and the feed flow rate, whereas the output variables are the monomer conversion and the weight‐average molecular weight. The use of a polynomial ARMA model for identification of the multivariable polymerization reaction system is validated by simulation study. For the experimental corroboration, correlations are developed to convert the on‐line measurements of density and viscosity of the reaction mixture to the monomer conversion and the weight‐average molecular weight. The on‐line values of the conversion and weight‐average molecular weight turn out to be in good agreement with the off‐line measurements. Despite the complex and nonlinear features of the polymerization reaction system, the polynomial ARMA model is found to satisfactorily describe the dynamic behavior of the polymerization reactor. Therefore, one may apply the polynomial ARMA model to the optimization and control of polymerization reactor systems. © 2000 John Wiley & Sons, Inc. J Appl Polym Sci 76: 1889–1901, 2000     

A novel on‐line optical method for algae measurement

In recent years, algal blooms have occurred worldwide, and algae‐rich water often has adverse effects on water production. The technique of algae measurement is a critical issue for adjusting water treatment processes according to the numbers of algae cells. The algae particles in the water are generally 2–200 µm in size with only a few smaller than 2 µm. The traditional algae measuring method is by visual observation with an optical microscope. However, traditional visual observation often needs 48 h fixing time, which makes the measurement results lag behind the needs of water production. To solve the problem, this study employed on‐line optical devices to improve the efficiency and accuracy of algae measurement. A photometric dispersion analyzer (PDA) and particle counting analyzer (PCA) were jointly utilized to monitor on‐line the algae concentration in natural water. Algae cells can be classified by different sizes. It was found that there was good correlation between R²_PDA and total algae counts in water. The PCA could quantitatively characterize the algae counts and species distribution of dominant algae species in real water. PDA and PCA could be used jointly to define on‐line the characteristics of real water containing mixed algae. Copyright © 2010 Society of Chemical Industry     

Error‐triggered on‐line model identification for model‐based feedback control

In industry, it may be difficult in many applications to obtain a first‐principles model of the process, in which case a linear empirical model constructed using process data may be used in the design of a feedback controller. However, linear empirical models may not capture the nonlinear dynamics over a wide region of state‐space and may also perform poorly when significant plant variations and disturbances occur. In the present work, an error‐triggered on‐line model identification approach is introduced for closed‐loop systems under model‐based feedback control strategies. The linear models are re‐identified on‐line when significant prediction errors occur. A moving horizon error detector is used to quantify the model accuracy and to trigger the model re‐identification on‐line when necessary. The proposed approach is demonstrated through two chemical process examples using a model‐based feedback control strategy termed Lyapunov‐based economic model predictive control (LEMPC). The chemical process examples illustrate that the proposed error‐triggered on‐line model identification strategy can be used to obtain more accurate state predictions to improve process economics while maintaining closed‐loop stability of the process under LEMPC. © 2016 American Institute of Chemical Engineers AIChE J, 63: 949–966, 2017     

Validation of three on‐line flow simulations for injection molding

Polymer process control is limited by a lack of observability of the distributed and transient polymer states. Three simulations of varying complexity are validated for on‐line simulation of an injection molding process with a two drop hot runner system to predict the state of the polymer melt in real time and thereby improve product quality in situ. The simplest simulation is a Newtonian model, which predicts flow rates given the inlet and outlet pressures. An intermediate non‐Newtonian and nonisothermal simulation utilizes a modified Ellis model that expresses the viscosity as a function of the shear stress in which the modeling of the heat transfer utilizes a Bessel series expansion to include effects of heat conduction, heat convection, and internal shear heating. A numerical simulation was also developed that utilizes a hybrid finite difference and finite element scheme to simultaneously solve the mass, momentum, and heat equations. Numerical verification indicates that the flow rate predictions of the described simulations compare well with the results from a commercial mold filling simulation. However, empirical validation utilizing a design of experiments indicates that the described analyses are qualitatively useful, but do not possess sufficient accuracy for quantitative process and quality control. Specifically, off‐line validation using optimal transducer calibration with well characterized materials provided a coefficient of regression, R², of ～0.8. However, blind validation with previously untested materials and no transducer re‐calibration provided a regression coefficient of ～0.4. While the direction of the main effects was usually correct, the magnitudes of the effects were frequently outside the confidence interval of the observed behavior. Several sources of variance are discussed, including sensor calibration, constitutive modeling of the polymer melt, and numerical analysis. POLYM. ENG. SCI. 46:274–288, 2006. © 2006 Society of Plastics Engineers     

Development of on‐line pyroprocessing for liquid thorium fueled reactors

The electrochemical behavior of thorium and uranium in molten LiF was examined separately and simultaneously at 1173 K using cyclic voltammetry. Inert molybdenum wire served as a working electrode. A platinum wire served as a quasireference electrode. The quality of voltammograms was highly dependent on the degree of dryness of the salt. The experimental results confirmed previous findings for similar salt matrices that thorium is reduced in a single, 4‐electron [Th(IV)/Th] step, whereas uranium is reduced in a two‐step process, including a 1‐electron exchange [U(IV)/U(III)] followed by a 3‐electron exchange [U(III)/U]. Diffusion coefficients for thorium and uranium were calculated and plotted as a function of concentration. Calibration curves of peak current density vs. analyte concentration were developed. Separability of uranium from the thorium‐rich matrix was confirmed feasible. Satisfactory concentration monitoring of uranium was demonstrated, whereas thorium concentration monitoring in the ternary salt was found problematic. © 2015 American Institute of Chemical Engineers AIChE J, 62: 1236–1243, 2016     

In‐line viscosity fuzzy control

A fuzzy control strategy on viscosity of a polymeric melt during the extruding processes is proposed to ensure producing the quality‐consistent end products. Because the viscosity model in the control system varies with respect to different conditions of the system command (e.g., screw speed), we focus on using a fuzzy controller accomplished with a fuzzy modeling technique to solve the model‐variant viscosity control system. Also, system stability analysis is presented. According to the simulation and the implementation results, better performances and a suitable using method are concluded. © 2000 John Wiley & Sons, Inc. J Appl Polym Sci 79: 1249–1255, 2001     

In‐line viscosity control in an extrusion process with a fuzzy gain scheduled PID controller

In the extrusion process, rapidly tracking the set point of quality factor and eliminating its variation to reduce the off‐specification product is important. In this study, the fuzzy gain‐scheduled proportional‐integral‐derivative (PID) controller is used to control the melt viscosity during extrusion processing. A second‐order model related to the viscosity and the extruder screw speed is developed empirically to approximate the extrusion system. It is concluded that, in comparison to the well‐known Zeigler‐Nichols PID tuning control scheme, the performances of the proposed control strategy is preferable both in simulation and implementation. © 1999 John Wiley & Sons, Inc. J Appl Polym Sci 74: 541–555, 1999     

Mechanical characterization of parts produced by ceramic on‐demand extrusion process

Ceramic On‐Demand Extrusion (CODE) is an additive manufacturing process recently developed to produce dense three‐dimensional ceramic components. In this paper, the properties of parts produced using this freeform extrusion fabrication process are described. High solids loading (~60 vol%) alumina paste was prepared to fabricate parts and standard test methods were employed to examine their properties including the density, strength, Young's modulus, Weibull modulus, toughness, and hardness. Microstructural evaluation was also performed to measure the grain size and critical flaw size. The results indicate that the properties of parts surpass most other ceramic additive manufacturing processes and match conventional fabrication techniques.     

A hybrid neural model (HNM) for the on‐line monitoring of lipase production by Candida rugosa

A mechanistic model was proposed by Gordillo for the representation of lipase production by Candida rugosa, with the bioreactor in batch and fed‐batch operation. However, the model was not able to represent the lipolytic activity. The objective of the present study is to propose an efficient hybrid neural‐phenomenological model (HNM) for this process. The experimental data used corresponded to fed‐batch operation with constant substrate feed rate at 2.8 × 10^?7; 5.6 × 10^?7 and 9.7 × 10^?7 kg s^?1. Artificial neural networks (ANNs) were trained to represent the aqueous and intracellular lipase activity and were further associated with a reduced version of the mechanistic model of the proposed HNM. When compared to the experimental data, the HNM exhibited higher accuracy. The HNM can be employed in process monitoring using only on‐line measurements of CO₂ and substrate feed rate to infer enzyme activities and also substrate and biomass concentrations. Copyright © 2007 Society of Chemical Industry     

In‐line optical detection in the transient state of extrusion polymer blending and reactive processing

Using an opticaldetector we followed the transient state of blends and composites, including a reactive blending during extrusion. The detection system is composed of a slit‐die with transparent windows fixed at the extruder exit, an optical arrangement with a W incandescent light microbulb with fixed luminescence, and a CdS photocell. As the tracer passes though the light path, it absorbs and backscatters part of the light, reducing the total transmitted light intensity. This is followed by changes in the voltage induced by the photocell to an electric circuit. We calibrated the response of the photocell at room temperature using a set of various films with a second phase dispersed, and obtained a logarithmic relationship. The tracers were particulate (phthalocyanine, TiO₂) and polymeric (PS, PA6) phases that absorb and scatter light, producing a residence time distribution (RTD) curvelike trace. Measurements were taken from a twin‐screw extruder Werner‐Pfleiderer ZSK 30 equipped with K‐Tron gravimetric feeders operating at various screw configurations and speeds, and feeding rates. The transient state of PP/PA6 blends can be easily detected optically and recorded using one of the components (either PP or PA6) added as a pulse in a steady‐state flow of the other component. With the simultaneous addition of a compatibilizer (polypropylene grafted with acrylic acid (PP‐g‐AA)) with the PA6, the intensity of the detector signal is substantially increased as a result of the PA6/PP‐g‐AA reaction. Quantitative off‐line infrared spectroscopy of the total amide group corroborated the in‐line measurements. These observations suggest that an in‐line optical detector may be a fast and simple way to study the flow behavior of blends and composites, including reactive processing. POLYM. ENG. SCI. 45:11–19, 2005. © 2004 Society of Plastics Engineers.     

Measuring number‐concentrations of nanoparticles and viruses in liquids on‐line

BACKGROUND: The surge of studies on artificial and natural nanoparticles has revealed a new world for engineering and life sciences, but in most instances, the small scale has made their number‐concentration determination in liquids a challenging problem. Former success has mostly been limited to special particles measured by indirect techniques. A new approach is required for this determination to facilitate the production and application of nanoparticles in different systems. RESULT: Here, an approach is described using a nanoparticle tracking system based on Brownian motion, which can be used to determine the number‐concentration of nanoparticles, including viruses, in liquids on‐line. Extensive analysis showed the influence of suspension concentration and particle size on the accuracy of measurements. Natural nanoparticles of Adenovirus and several types of artificial nanoparticles, including precision nanobeads, uniform inorganic silica microspheres, monodisperse gold metal colloids and aggregated Aerosil nanoparticles, were measured and compared by counting the monitored particle number obtained using light scattered from individual particles, from which the particle number‐concentration, the product yield and the aggregation could be evaluated. CONCLUSION: This approach was compared with the mathematical calculation method and the emission spectrophotometry technique used for practical applications. The results showed this new approach had improved accuracy for determination of the particle number‐concentration. Copyright © 2010 Society of Chemical Industry     

Optical characterization of concentrated dispersions: applications to laboratory analyses and on‐line process monitoring and control

Light scattering methods are often used to study the stability of suspensions or emulsions and to estimate the dispersed phase properties such as particle size and volume fraction. However, such optical methods often require a previous dilution of the dispersion because of a limited measurement range, and are then unable to give information about the real physical state of dense heterogeneous media. A new technology based on multiple light scattering analysis and called Turbiscan has been recently developed by a French company, Formulaction, to fill this gap and to characterize both diluted and concentrated dispersions. In the first part, we review the physical concepts of multiple light scattering by dispersions. In relation to the optical analyser Turbiscan, we present physical and statistical models for the radiative transfer in dense suspensions. In the second part, we investigate the influence of particle volume fraction and particle size (polystyrene latex bead suspensions) on the backscattered and transmitted light fluxes measured by Turbiscan. The experimental data are compared with results from the physical models. In the last section, we use the optical analyser Turbiscan Lab to detect and characterize various concentrated dispersions destabilization (coalescence, flocculation, creaming and sedimentation), and then the Turbiscan On Line to monitor and characterize an emulsification process under ultrasonic agitation. Copyright © 2004 Society of Chemical Industry