Predicting mechanical properties of elastomers with neural networks

A polystyrene (PS)/rubber blend compatibilized with PS-g-rubber copolymer, prepared via emulsion polymerization using redox initiator system, is used to investigate the utilization of central composite design (CCD) and artificial neural network (ANN) approaches in correlating polymerization conditions to mechanical properties (tensile strength and abrasion loss) of unfilled compound vulcanizates. The conditions were manipulated by changing four factors: reaction temperature and time, percentage of deproteinized rubber in the mixture containing natural rubber, and amount of chain transfer agent. The results show that the relationships between the conditions and the mechanical properties for compatibilized PS/rubber blend are too complex to be explained by polynomials, but are well described by the ANN models, developed for each response. In addition, simulation results for the tensile strength response as a function of those factors using the obtained ANN are in agreement with literature, whereas those results for the abrasion loss do not quite agree with literature due to the interference of the large measurement error. This suggests that only experimental data with high precision should be used to train an ANN to achieve a model with not only best performance but also high reliability.

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Prediction of mechanical properties of compatibilized styrene/natural‐rubber blend by using reaction conditions: Central composite design vs. artificial neural networks

Sresungsuwan

Hansupalak

2012

“…In brief, ANN has basic elements, which are three layers (so‐called input, hidden, and output layers), weights, bias, and transfer functions 37–39. It should be noted that there can be more than one hidden layer, but usually a network containing one hidden layer and numerous neurons is enough to perform a task.…”

Section: Methodsmentioning

confidence: 99%

“…In brief, ANN has basic elements, which are three layers (so-called input, hidden, and output layers), weights, bias, and transfer functions. [37][38][39] It should be noted that there can be more than one hidden layer, but usually a network containing one hidden layer and numerous neurons is enough to perform a task. Each neuron, or node, in the input layer corresponding to each independent variable sends a weighted vector of the variable to all neurons in the hidden layer.…”

Section: Modeling the Relationship Between The Styrene Conversion Andmentioning

confidence: 99%

Prediction of styrene conversion of polystyrene/natural rubber graft copolymerization using reaction conditions: Central composite design versus artificial neural networks

Aroonsingkarat

Hansupalak

2012

Gross copolymer or the total product of graft copolymerization of polystyrene (PS) and rubber, prepared via emulsion polymerization using a redox initiator, is used to investigate the utilization of central composite design and artificial neural network (ANN) approaches in correlating the graft copolymerization conditions to the monomer conversion. The conditions were manipulated by changing four factors: reaction temperature and time, percentage of deproteinized natural rubber (DPNR) in the rubber mixture also containing NR, and amount of chain transfer agent. For DPNR preparation, the incorporation of ultrasound energy into a deproteinizing method (i.e., urea treatment) was preexamined. A shorter reaction time, a lower total nitrogen content, and no agglomeration of rubber particles suggest the success of the incorporation. Results exhibit that the relationship between those factors and the response can be better described by the ANN model, which is further proved to be an excellent tool for the prediction of the conversion at other reaction conditions. In addition, the thermal behavior of gross copolymer is similar to its parents, the rubber and neat PS, but more to the former owing to the larger amount of rubber component. © 2012 Wiley Periodicals, Inc. J. Appl. Polym. Sci., 2013

“…Coran14 postulated a simplified reaction scheme to explain the kinetics of delayed action sulfur vulcanization and this simplified scheme adequately accounted for the kinetics of crosslink formation. Although chemical mechanisms differ, vulcanization with other crosslinking systems may be explained in a similar way with Coran vulcanization scheme 15. Coran method was used for the interpretation of the experimental results.…”

Section: Introductionmentioning

confidence: 99%

Mechanical, vulcametric, and thermal properties of the different 5‐ethylidene 2‐norbornene content of ethylene‐propylene‐diene‐monomer vulcanized with different types and compositions of peroxides

Çavdar

Özdemir

Usanmaz

2009

Ethylene-propylene-diene-monomer (EPDM) rubber is an important commercial polymer. The vulcanization process significantly changes its thermal, mechanical, and vulcametric properties. This study was carried out to find optimum formulation of EPDM composite for a better application in automotive industry. Sixteen EPDM polymer samples having different 5-ethylidene 2-norbornene (ENB) and ethylene contents were vulcanized with different types and compositions of peroxide and coagents. The mechanical and vulcametric properties of these samples were measured and compared. The type of peroxide, coagent, and EPDM grade affected the mechanical, thermal, and vulcametric properties of the EPDM rubber to some extend. Use of aromatic peroxide and coagent increased the thermal stability slightly. Mechanical properties were changed very slightly with the change of peroxide type for the same content of peroxide and coagent. Scorch time and cure time decreased with initial increase of the peroxide content. EPDM compound vulcanized with BBPIB peroxide and TAC/S coagent has higher cure time than EPDM compound vulcanized with DMBPHa peroxide and TMPTMA coagent. Coran method was used for the modeling of experimental data. Velocity constant for the formation of peroxide radical and polymer radical were found for each case.