Prozessüberwachung und -optimierung in der Elastomerverarbeitung auf der Basis statistischer Versuchsmethodik und mathematischer Prozessmodelle

There is a growing need for precision rubber parts and complex rubber components for the automotive sector. The quality requirements rising thereby lead to the increasing meaning of the quality assurance and the quality management in the rubber industry. In the rubber processing, there is an extra step or processing phase the material compounding step, which does not exist in the processing of plastics. This additional discontinual processing step may account for the higher risk for deviation between rubber batches or within a batch. This in turn may lead to problems in the subsequent stages and result in deviations in the final product properties. This thesis propounds guidelines and methods for efficient process analysis and optimization with particular focus on the rubber mixing process in an internal mixer. The research methods used for this thesis were process analyses and modeling which allowed the exact determination of correlation between the mixing process and the rest of the process chain. The experiments carried out focused on selected steps of the production of the rubber parts on mixing process in the rubber internal mixer, transfer molding process an on the injection molding process. In the rubber industry, there are many different testing approaches used and their validity is not always sufficient. At the same time they are inappropriate. Consequently the rubber characterization methods were analyzed. The focal point of the investigations was to find out the method most correlated to the rubber behavior of the rubber compound during processing and to the properties of the final product. As an approach to this problem statistical methods of the process analysis, like correlation analysis or regression were used. The validity of the rubber compound testing methods was determinated by correlational coefficient and the coefficient of certainty. The results showed that there is a direct correlation between the viscose and elastic compound properties and the qualities of the final product. In additional to that the rheological properties of the rubber mixture gives valuable information about the processing behavior in an injection molding machine and mold. The investigations were carried out using various measuring methods such as Rubber Process Analyzer (RPA), High Pressure Rheometer, Rheovulkameter, Mooney Viscometer. The RPA and the Mooney-Viscometer measurements proved to be the most accurately (reliable) of the method used. This thesis also covers the development and application of mathematical models for online-prediction of the compound properties during the mixing process in an internal mixer using Artificial Neural Networks (ANN) and linear regression analysis. These methods were used to predict the different rubber compound properties based on different types of rubber and on different fillers (carbon blacks and silica). Using the regression models in most cases it was possible to describe the interrelationship with a very high level of certainty. The advantages of using models based on ANN were evident when dealing with non linear interrelationships, e.g. (for silica-compound) In this case it was possible to achieve coefficients of certainty of about 90%. It was possible to prove the practical relevance of this modeling method for a tire compound produced in a 320 liters production internal rubber mixer. In this case it was possible to calculate the compound property – Mooney-viscosity with a high degree of certainty using ANN. Based on the results, an online prediction system was developed, presented and explored in this thesis. This system can be used in practice to reduce the number of compound tests carried out for quality control in the rubber industry.

Grantor dc:publisher

Publikationsserver der RWTH Aachen University

Year dc:date

2001

Author dc:creator

• Ryzko, Peter Jaroslaw

Contributors dc:contributor

• Haberstroh, Edmund

• info:eu-repo/classification/ddc/620
• Ingenieurwissenschaften
• Kautschukverarbeitung
• Prozessanalyse
• Prozessmodell
• Mischen
• Innenmischer
• Prozessoptimierung
• Optimale Versuchsplanung

Statement dc:rights

• info:eu-repo/semantics/openAccess

Language dc:language

ger

Identifier

https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123676%22

OAI identifier oai:identifier

oai:publications.rwth-aachen.de:62082