Rubber Mixing Process Research Articles

MINIMIZING TEMPERATURE DEVIATIONS IN RUBBER MIXING PROCESS BY USING ARTIFICIAL NEURAL NETWORKS

ABSTRACT Rubber mixing is a complex manufacturing process that poses challenges for process control due to the high number of control variables, including mixing parameter settings, rheological behavior, compound viscosity, and batch-dependent material variations. Already small deviations from the control variables can influence the compound properties, leading to increased scrap rates. To address these challenges, this paper introduces an artificial intelligence–based approach to enhance process control in rubber mixing by predicting mixing temperatures from input variables. The proposed method uses feedforward neural networks (FFNs) to enable early identification of batch-specific temperature deviations, thereby enabling systematic improvements with each new application. The FFN was trained on a diverse dataset encompassing various rubber recipes and batches. Post-training, the FFN demonstrated remarkable accuracy, achieving a mean absolute percentage error of 1.00% on the training dataset and 1.44% on the validation dataset, thereby showcasing its efficacy in predicting temperature fluctuations within the mixing process. Consequently, the FFN can determine the relevant input variables necessary to achieve specific mixing temperatures, providing a foundation for an automated control system in rubber mixing process. This paper outlines the system architecture of the FFN tailored for rubber mixing and provides a comprehensive overview of the experimental results.

A closed-loop evaluation identification algorithm for rubber mixing process based on physical information and deep learning

A closed-loop evaluation identification algorithm for rubber mixing process based on physical information and deep learning

Formulation and Performance of NBR/CR-Based High-Damping Rubber Composites for Soundproof Using Orthogonal Test.

Multiple functional-material-filled nitrile butadiene rubber/chloroprene rubber (NBR/CR) acoustic composites were extensively studied and prepared. According to the orthogonal test table L25 (56), 25 groups of samples were prepared by using a low-temperature one-time rubber mixing process. With tensile strength, average transmission loss, and damping peak as indexes, the influence degree of different factors and levels on the properties of acoustic composites was quantitatively discussed and analyzed. The matrix weight analysis was employed to optimize the material formula of rubber composites, and the corresponding influence weight was given. Results showed that the acoustic composite with blending ratio of 70/30 for NBR/CR matrix had preferable mechanical and acoustic properties; adding mica powder (MP) and montmorillonite (MMT) in matrix contributed to improve all above three indexes owing to their specific lamellar structures; hollow glass beads (HGB) had a positive influence on improving acoustic property due to its hollow microcavities, however, it had a negative impact on damping property because of the smooth spherical surfaces. Accordingly, the optimal formulation was found to be NBR/CR blending ratio of 70/30, MP of 10 phr (per hundred rubber), HGB of 4 phr, and MMT of 10 phr.

Evaluation and Modification of Rubber Mixing Process

Evaluation and Modification of Rubber Mixing Process

ニーダーの特徴およびゴム練りラインの設計

Rubber mixing process consists of batch weighing, mixing and kneading, sheeting and cooling processes. As several types of mixing machines have been introduced for the core mixing and kneading process in the rubber industry, the selection of mixer type has a significant effect on factory construction and mixing line engineering entirely.

NBR/CR‐Based High‐Damping Rubber Composites Containing Multiscale Structures for Tailoring Sound Insulation

AbstractHigh‐damping acoustic composites with unique sound insulation feature is reported, demonstrating surprisingly stable soundproof properties over a wide bandwidth of frequency (63–6300 Hz) with good mechanical properties. The extraordinary acoustic properties are attributed to the multiscale synergy of lamellar and hollow structures by multifillers adding. The acoustic composites are designed and fabricated through a low‐temperature one‐time rubber mixing process using nitrile butadiene rubber(NBR), chloroprene rubber (CR) as matrix and mica powder (MP), hollow glass beads (HGB), and montmorillonite (MMT) as multifillers. The soundproof mechanism is discussed in detail. Results indicate that the synergism of damping and sound absorption behaviors in the composition has a considerable impact on the acoustic characteristics, which can be controlled by the blending ratio of NBR/CR matrixes and the multiscale structures of MP/HGB/MMT multifillers. The designed multifillers acoustic composites can adapt to broadband engineering noise control including walls in express train constructions, panels of vessel or aircraft cabins, and large transportation pipelines.

Simulation and experimental research on the wear of synchronous quadruple rotor of an internal mixer during the mixing process

As the core component of an internal mixer, the rotor is subjected to friction from rubber and its filler during the mixing process, which causes the surface of the rotor to wear and can also affect the gap between the rotor edges as well as the gap between the rotor and the mixing chamber wall. The more significant the rotor wear, the larger the gap between the rotor ribs and the wall of the mixing chamber, which results in weaker shearing effect of the rotor on the rubber and even leads to uneven dispersion of filler in the rubber, eventually affecting the mixing effect and mixing quality of the glue. In this study, based on the viscoelastic solid theory of the rubber mixing process and the friction mechanism between rubber and metal, combined with the discrete element method, a dynamic simulation model for the wear of the synchronous quadrilateral rotor of an internal mixer is established. The simulations are performed in the EDEM software, and the results are analyzed and compared under different mixing parameters. Further, a control experiment is established to verify the feasibility of the proposed dynamic model for rotor wear.

Deep Semi-Supervised Just-in-Time Learning Based Soft Sensor for Mooney Viscosity Estimation in Industrial Rubber Mixing Process.

Soft sensor technology has become an effective tool to enable real-time estimations of key quality variables in industrial rubber-mixing processes, which facilitates efficient monitoring and a control of rubber manufacturing. However, it remains a challenging issue to develop high-performance soft sensors due to improper feature selection/extraction and insufficiency of labeled data. Thus, a deep semi-supervised just-in-time learning-based Gaussian process regression (DSSJITGPR) is developed for Mooney viscosity estimation. It integrates just-in-time learning, semi-supervised learning, and deep learning into a unified modeling framework. In the offline stage, the latent feature information behind the historical process data is extracted through a stacked autoencoder. Then, an evolutionary pseudo-labeling estimation approach is applied to extend the labeled modeling database, where high-confidence pseudo-labeled data are obtained by solving an explicit pseudo-labeling optimization problem. In the online stage, when the query sample arrives, a semi-supervised JITGPR model is built from the enlarged modeling database to achieve Mooney viscosity estimation. Compared with traditional Mooney-viscosity soft sensor methods, DSSJITGPR shows significant advantages in extracting latent features and handling label scarcity, thus delivering superior prediction performance. The effectiveness and superiority of DSSJITGPR has been verified through the Mooney viscosity prediction results from an industrial rubber-mixing process.

Characterization of the incorporation of a granular medium into elastomer matrices during mixing process

AbstractMixing of filler particles with elastomers is one of the most important steps in tire manufacturing process. Precipitated Silica (Si), as a filler, shows enhanced properties compared with carbon black when incorporated with an elastomer such as low rolling resistance, high wet grip performance, and high wear resistance. According to the standard rubber mixing process, the fragmentation of the filler particles in an elastomer during mixing must yield to breakage of particles to a size that enables homogenous mixing of the material. Hence, this experimental study is concerned with understanding the behavior of silica during its incorporation in the elastomer. Incorporation is done through shear forces present in the Internal Mixer. Different experimental methodologies are used to measure the particle size distribution. For free silica particles that have been fragmented but not incorporated, we use Laser granulometry. Silica embedded in elastomer is analyzed using X‐ray tomography. The results obtained suggest that free silica and silica embedded in elastomer follow erosion mechanism.

Ensemble Just-In-Time Learning-Based Soft Sensor for Mooney Viscosity Prediction in an Industrial Rubber Mixing Process

The lack of online sensors for Mooney viscosity measurement has posed significant challenges for enabling efficient monitoring, control, and optimization of industrial rubber mixing process. To obtain real-time and accurate estimations of Mooney viscosity, a novel soft sensor method, referred to as multimodal perturbation- (MP-) based ensemble just-in-time learning Gaussian process regression (MP-EJITGPR), is proposed by exploiting ensemble JIT learning. This method employs perturbations on similarity measure and input variables for generating the diversity of JIT learners. Furthermore, a set of accurate and diverse JIT learners are built through an evolutionary multiobjective optimization by balancing the accuracy and diversity objectives explicitly. Moreover, all base JIT learners are combined adaptively using a finite mixture mechanism. The proposed method is applied to an industrial rubber mixing process for Mooney viscosity prediction, and the experimental results demonstrate its effectiveness and superiority over traditional soft sensor methods.

Semi-Supervised Hybrid Local Kernel Regression for Soft Sensor Modelling of Rubber-Mixing Process

Soft sensor techniques have been widely adopted in chemical industry to estimate important indices that cannot be online measured by hardware sensors. Unfortunately, due to the instinct time-variation, the small-sample condition and the uncertainty caused by the drifting of raw materials, it is exceedingly difficult to model the fed-batch processes, for instance, rubber internal mixing processing. Meanwhile, traditional global learning algorithms suffer from the outdated samples while online learning algorithms lack practicality since too many labelled samples of current batch are required to build the soft sensor. In this paper, semi-supervised hybrid local kernel regression (SHLKR) is presented to leverage both historical and online samples to semi-supervised model the soft sensor using proposed time-windows series. Moreover, the recursive formulas are deduced to improve its adaptability and feasibility. Additionally, the rubber Mooney soft sensor of internal mixing processing is implemented using real onsite data to validate proposed method. Compared with classical algorithms, the performance of SHLKR is evaluated and the contribution of unlabelled samples is discussed.

Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes.

Although several data-driven soft sensors are available, online reliable prediction of the Mooney viscosity in industrial rubber mixing processes is still a challenging task. A robust semi-supervised soft sensor, called ensemble deep correntropy kernel regression (EDCKR), is proposed. It integrates the ensemble strategy, deep brief network (DBN), and correntropy kernel regression (CKR) into a unified soft sensing framework. The multilevel DBN-based unsupervised learning stage extracts useful information from all secondary variables. Sequentially, a supervised CKR model is built to explore the relationship between the extracted features and the Mooney viscosity values. Without cumbersome preprocessing steps, the negative effects of outliers are reduced using the CKR-based robust nonlinear estimator. With the help of ensemble strategy, more reliable prediction results are further obtained. An industrial case validates the practicality and reliability of EDCKR.

Energy Consumption Analysis to Detect Process Failure in Rubber Compound Mixing Process

There are many ways to detect the abnormality of rubber mixing process regardless of the availability of indicator system of the abnormality. One way is to detect the problem after the result of rubber production tested by laboratory teams. Abnormality will be late to be anticipated because the result of the test will appear after 8 to 10 batch afterward. The abnormality in the mixing process will be detected if the results of the testing compound are out of the specification. The energy consumption will be measured for every batch by using the internal current sensor in the AC motor main drive. By reading the figure of measurement, energy consumption character will be identified and the specification standard of energy consumption can be defined. The process will follow the specification standard of energy consumption and abnormality condition will be corrected immediately.

TWO-DIMENSIONAL NONISOTHERMAL NUMERICAL ANALYSIS OF SPEED RATIO EFFECTS ON DISPERSION AND DISTRIBUTION IN HIGH-VISCOSITY PARTIALLY FILLED RUBBER MIXING

ABSTRACT Two-dimensional, transient, and nonisothermal computational fluid dynamics simulations are conducted for high-viscosity rubber mixing in a two-wing rotor-equipped partially filled chamber of fill factor 75%. Calculations presented assess the effect of three differential speeds or speed ratios of the two rotors for the rubber mixing process: 1.0 (also called even speed), 1.125, and 1.5. A Eulerian multiphase model, the volume of fluid technique, is employed to simulate two different phases, rubber and air, by calculating the free surface between the two phases, in addition to the main governing equations such as the continuity, momentum, and energy equations. To characterize the non-Newtonian, highly viscous rubber under nonisothermal conditions, the shear rate–dependent Carreau-Yasuda model along with an Arrhenius function are employed. A set of massless particles is introduced into the chamber to calculate several parameters related to dispersive and distributive mixing characteristics. Specifically, the mixing index and maximum shear stress are analyzed for the dispersive nature, whereas cluster distribution index and length of stretch are calculated for investigating the distributive nature of the mixing process. Also, the temporal viscous heat generation rate, a good indication of the temperature rise throughout the domain, which is critical in the process and equipment design, is analyzed here. Results showed that the 1.125 speed ratio was the most efficient in terms of distributive mixing and heat generation.

NON-ISOTHERMAL EFFECTS IN PARTIALLY FILLED RUBBER MIXING SIMULATIONS OF MANUFACTURING PROCESSES

ABSTRACT A finite volume technique is used to analyze the isothermal and non-isothermal flow behavior for the rubber mixing process in a two-dimensional, partially filled (75%) internal mixer, which consists of two counterrotating rotors rotating at 20 rpm. In order to capture the interface between air and rubber, an Eulerian multiphase model called volume of fluid (VOF) has been employed here. The transient flow behavior was accomplished by a sliding mesh technique, and the highly viscous, non-Newtonian properties of the rubber have been characterized using the Bird–Carreau model. Most of the previous computational fluid dynamic (CFD)-based investigations of rubber mixing assumed isothermal flow, and consequently negligible viscous heat generation, temperature rise, and viscosity drop associated with heat generation. Hence, a non-isothermal simulation is carried out, and results are compared with those of an equivalent isothermal simulation. In addition, dispersive and distributive mixing characteristics are assessed using statistics calculated from particle tracks generated by a set of massless and neutral particles that have been injected in the simulation. For this purpose, quantities such as the cumulative distribution of maximum shear stress, length of stretch, and cluster distribution index are calculated and compared between isothermal and non-isothermal conditions. Results showed a significant difference between the isothermal and non-isothermal simulations, thus making the isothermal assumption critical. Also, the non-isothermal simulation predicted better mixing during the entire mixing cycle.

The Effect of Ingredients Mixing Sequence in Rubber Compounding upon Vulcanization Kinetics of Natural Rubber: An Autocatalytic Model Study

This study examined the effect of ingredients mixing sequence to the vulcanization kinetics of natural rubber. The effects of mixing temperature, vulcanization temperature, and the carbon black type upon the kinetics were also studied by using rheography and an autocatalysis reaction model approach. The results showed that this model is good in providing information on vulcanization reaction kinetics of natural rubber. High vulcanization temperature resulted in high reaction rate constant. The more black carbon mixed at the beginning of the rubber mixing process, the higher reaction rate constant would be. The mixing of carbon black and rubber chemicals mixed into the rubber subsequently resulted in the higher reaction rate constant compared with that of simultaneously.

Just-in-time semi-supervised soft sensor for quality prediction in industrial rubber mixers

Increasing data-driven soft sensors have been adopted to online predict the quality indices in polymerization processes to improve the availability of measurements and efficiency. However, in industrial rubber mixing processes, most existing soft sensors for online prediction of the Mooney viscosity only utilized the limited labeled data. By exploring the unlabeled data, a novel soft sensor, namely just-in-time semi-supervised extreme learning machine (JSELM), is proposed to online predict the Mooney viscosity with multiple recipes. It integrates the just-in-time learning, extreme learning machine (ELM), and the graph Laplacian regularization into a unified online modeling framework. When a test sample is inquired online, the useful information in both of similar labeled and unlabeled data is absorbed into its prediction model. Unlike traditional just-in-time learning models only utilizing labeled data (e.g., just-in-time ELM and just-in-time support vector regression), the prediction performance of JSELM can be enhanced by taking advantage of the information in lots of unlabeled data. Moreover, an efficient model selection strategy is formulated for online construction of the JSELM prediction model. Compared with traditional soft sensor methods, the superiority of JSELM is validated via the Mooney viscosity prediction in an industrial rubber mixer.

Industrial Mooney viscosity prediction using fast semi-supervised empirical model

In industrial rubber mixing processes, the quality index (i.e., Mooney viscosity) cannot be online measured directly. Traditional data-driven empirical models for online prediction of the Mooney viscosity have not utilized the information hidden in lots of unlabeled data (e.g., process input variables during each mixing batch). A simple semi-supervised nonlinear soft sensor method for the Mooney viscosity prediction is developed. It integrates extreme learning machine (ELM) and the graph Laplacian regularization into a unified modeling framework. The useful information in unlabeled data can be explored and introduced into the prediction model. Furthermore, a bagging-based ensemble strategy is combined into semi-supervised ELM (SELM) to obtain more accurate predictions. The Mooney viscosity prediction in an industrial internal mixer exhibits its promising prediction performance of the proposed method by incorporating the information in unlabeled data efficiently.

Comparison of Three Rotor Designs for Rubber Mixing Using Computational Fluid Dynamics

ABSTRACT In recent years, there has been an increasing demand for efficient mixers with high-quality mixing capabilities in the rubber product industry, with the focus of producing fuel-efficient tires. Depending on the functional characteristics of the tire and thus the compounding ingredients, different types of mixers can be used for the rubber mixing process. Hence, the choice of an appropriate mixer is critical in achieving the proper distribution and dispersion of fillers in rubber and a consistent product quality, as well as the attainment of high productivity. With the availability of high-performance computing resources and high-fidelity computational fluid dynamics tools over the last two decades, understanding the flow phenomena associated with complex rotor geometries such as the two- and four-wing rotors has become feasible. The objective of this article is to compare and investigate the flow and mixing dynamics of rubber compounds in partially filled mixing chambers stirred with three types of rotors: the two-wing, four-wing A, and four-wing B rotors. As part of this effort, all the 3D simulations are carried out with a 75% fill factor and a rotor speed of 20 rpm using a computational fluid dynamics (CFD) code. Mass flow patterns, velocity vectors, particle trajectories, and other mixing statistics, such as cluster distribution index and length of stretch, are presented here. All the results showed consistently that the four-wing A rotor was superior in terms of dispersive and distributive mixing characteristics compared with the other rotors. The results also helped to understand the mixing process and material movement, thereby generating information that could potentially improve productivity and efficiency in the tire manufacturing process.

Fast property prediction in an industrial rubber mixing process with local ELM model

ABSTRACTOnline property prediction in industrial rubber mixing processes is not an easy task. An efficient data‐driven prediction model is developed in this work. The regularized extreme learning machine (RELM) is utilized as the fundamental soft sensor model. To better capture distinguished characteristics in multiple recipes and operating modes, a just‐in‐time RELM modeling method is developed. The number of hidden neurons and the value of regularization parameter of the just‐in‐time RELM model can be efficiently selected using a fast leave‐one‐out strategy. Consequently, without the time‐consuming laboratory analysis process, the Mooney viscosity can be online predicted once a mixing batch has been discharged. The industrial Mooney viscosity prediction results show its better prediction performance in comparison with traditional approaches. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2017, 134, 45391.