Printer Friendly

CQI in the compounding of silicone rubber.

Quality has become the number one priority in American industry, recognized by most manufacturers as a key to overcoming competition and reducing costs. Like other material suppliers, compounders of silicone rubber are realizing that quality assurance practices which were considered milestones just a few years ago are quickly becoming routine procedures.

Much of the initiative for higher quality levels in silicone rubber production comes from OEMs using the fabricated parts. Automotive manufacturers and their suppliers in particular have dedicated tremendous resources toward quality improvement. The performance requirements of these companies generally drive product advancements, and their expectations are the foundation of quality standards.

The escalating pace of quality progress is prompting at least one silicone supplier to develop sophisticated, cross-functional improvement programs. OEMs and fabricators are both reaping the benefits of evolving quality measures, as compounders extend the performance characteristics of the elastomers and increase product consistency. Customer service issues like shipping performance also reflect the systematic changes; average order-to-ship times drop as first-time pass rates climb.

In the past, it has been considered routine to "rework" materials that failed to meet specifications until the necessary physical properties were reached. This assumption often undermines consistency efforts, however, since reworking can alter the processing characteristics. The added variation disrupts production and delivery schedules, and ultimately drives up fabrication costs.

Continuous improvement

At the heart of most structured quality programs is the belief that today's goals are tomorrow's stepping stones. Meeting current objectives becomes the means for reaching successively higher quality levels, prompting materials engineers to recognize that continuous improvement (CI) is a permanent component of their work.

Continuous improvement in the production of silicone rubber is viewed as a systematic, cross-functional commitment that integrates every component of the compounding process to achieve broad-based, progressive quality gains. Much like traditional quality assurance, its objective is product and service excellence. Unlike the conventional view of quality, however, there are no static goals. The CI concept is dedicated to continuous advancement of overall performance quality, based on external (customer-perceived) criteria.

CI addresses both material and operational issues. Some tactics are fundamental, such as development of improved handling techniques to reduce contamination. Others are extremely sophisticated, and employ statistical analysis tools to measure progress.

Consistency is the key

In the past, quality improvement programs in silicone compounding have focused primarily on specifications (also known as conformance quality), and the ability to meet product specifications remains critical as a basis for quality standards.

But the management of performance quality extends beyond the physical characteristics of the material itself. It includes shipping performance, improved response to customer action requests, and even plant safety. CI also emphasizes reduced variation in testing, a facet of material supply rarely considered when suppliers seek improved quality levels.

Although conformance to established tolerances is considered fundamental in the compounding of silicone rubber, it has become apparent that a defect-free product is not absolute assurance that a material will meet all of a fabricator's needs. Even though test data may fall well within the specifications established by the final customer, minor variation from previous lots can cause problems during processing. Consistency, therefore, is critical to efficient fabrication.

Process/product quality

Inherent to effective quality control in compounding silicone rubber is the understanding that process quality is a primary determinant of product quality. Variation in mix time, temperature or blade speed during compounding can cause fabrication problems and variable product performance. Contamination has also become a critical issue, both for functional and cosmetic reasons.

Complicating the compounder's task is the fact that variation in a fabricator's equipment can also affect a material's processing characteristics. Temperature profiles on two theoretically-identical molding machines may vary enough to cause samples from the same lot to cure at slightly different rates, for example. In an extrusion process, equipment wear can contribute to variation and affect a number of product quality parameters. While the compounder may be unable to influence these variables directly, stability in compounding through process control can greatly improve the fabricator's ability to understand this process and reduce scrap.

Statistical quality control

The application of statistical methods to track manufacturing performance has made it possible to identify sources of variation more efficiently and reduce or eliminate them from the process. Unfortunately, most silicone suppliers limit their use of statistical techniques to comparing physical properties and specifications, ensuring that compounds fall within established parameters (conformance quality). But like the computer systems that support them, quality programs have undergone dramatic change in a brief period of time.

Effective quality management just two or three years ago meant computerized tracking of product specifications and test results, with an emphasis on reducing variation in physical properties. Information management systems were introduced to automatically compare test results to tolerances, and to record the properties of each material batch as testing was completed.

Some compounders also established |alarm specs' to provide notification when a tested property strayed close to the product tolerances, but before results actually fell outside of the control limits. Considered major innovations when they were introduced in silicone compounding, such practices are no longer viewed as sufficient to ensure the necessary levels of product consistency. Today's information management systems are used to monitor process parameters to establish control limits.

Raw materials

A comprehensive quality program addresses the entire compounding process. In the past, it was deemed important that incoming materials conform to specifications; now it is understood that they must also be in control. Statistics enhance the compounder's knowledge about the raw materials and how variation affects the products made with them. Further, control charts on the raw materials can also provide insight into the process used to make the primary ingredients themselves, and what role this process plays in determining physical properties of the cured elastomer.

SQC is viewed as necessary on all raw materials, allowing the compounder to verify consistency and determine whether there is a correlation between physical properties of a compounded product and the raw materials that were used to produce it. For example, surface area is one of the characteristics measured on the silica filler used to compound silicone elastomers. Surface area has been found to have a significant effect on the final properties of the fully compounded material.

To quantify the relationship, different statistical analyses can be employed. Initially, a linear correlation coefficient (LCC) may be calculated to establish whether a relationship exists between the surface area and physical properties. LCC is a statistical tool that will indicate whether there is strong signal demonstrating a causal link (figure 1).

To help technicians fully understand the consequences of surface area variation, multiple regression analysis (MRA) can be used to examine the relationship between the surface area and specific physical properties of the cured elastomer. By working backward through the statistical records of previous material lots, an equation can be developed that will help predict the compound's properties from the surface area of the incoming filler.

Design of experiments (DOE) have also helped to separate the statistical signal from background noise in the process. These tightly-controlled test runs are a statistical approach to experimentation, under which a matrix of experimental parameters is employed as a technique for maximizing the efficiency of the test sequence. With DOE, several parameters can be evaluated at once, allowing the compounder to extract the most information from the fewest number of trials. The use of these statistical tools is based on the premise that process quality must be designed in to enhance material performance.

The mix cycle

Quality engineers at Dow Coming STI also concentrate on reducing variation during mixing. One of the company's most basic chart procedures is the SPC check on rheometer tests performed on samples of large-volume products. Maximum torque and cure time are plotted even before the mixer is emptied. As long as the results verify that the compound is statistically in control, the batch is dumped and weighed. Any significant inconsistency is detected immediately, and an operator-driven investigation begins.

Another fundamental element is mass-balance standard deviation, calculated by comparing the theoretical weight of the finished mix with its actual weight. The theoretical weight is a projection based on lot size and ingredients, indicating what the mass of the compounded material should be. By plotting the difference between theoretical and actual weights, operators can get an early indication whether the product is in control.

Operating procedures

Another building block of CI is a focus on the procedures used by equipment operators. Documented standard operation procedures (SOPs) help assure lot-to-lot consistency, and are also a key element of safe manufacturing practices. The most advanced methods are extremely thorough, interactive procedures which leave little room for interpretation.

For each product, the exact manufacturing procedure is clearly outlined, including the specific order of ingredient addition, mix time and temperature. Detailed SOPs help eliminate guesswork and variation among individual equipment operators that can cause inconsistency.

At STI, for example, each material lot starts with a shop packet, a permanent set of computer-generated documents that becomes part of the manufacturing records. The packet contains product formulation, along with a |pick slip,' which records data on the raw materials. The SOPs serve as a set of guidelines to mixing every lot of a given material the same way on every shift.

A critical feature of the SOPs is the fact that they are equipment-specific. Because there are unavoidable differences between mixers, for instance, slight revisions in operating procedure may be necessary from one machine to the next, even on equipment for the same manufacturer. Equipment-specific SOPs help ensure that no matter how long it's been since the compound was last produced, adhering to the documented steps will produce a material which processes and performs consistently every time.

At the end of every mix, process data are generated to record critical data from the compounding process. Mix time, power consumption, blade speed and temperature are part of the information that completes the packet on each material lot. These process data are captured by computer at a specified frequency, and are displayed in graphics and text.

The shop packet becomes more than a recipe for making the material. It is also a record of what did in fact occur as that individual silicone compound was produced. In the past, some compounders have used a basic batch card as their only documentation, simply indicating the necessary ingredients and quantities.

Prevention, not correction

This computerized |trail' of process characteristics emphasizes the role that manufacturing methods play in determining the consistency of the compound. Processing information can be extremely useful in conducting investigations when a material exhibits some variation. But proponents of CI maintain that statistics are most valuable not in determining what caused a material variation, but preventing that variation from occurring at all.

To that end, the objective of quality engineers is to begin control charting these variables on a real-time basis. That is, to have immediate access to the current statistics on a batch in progress and be able to correlate physical properties to the process data. Methods of just a few years ago, which accepted a certain percentage of rejection and rework as a fact of life in the compounding of silicone, begin to seem archaic when compared to emerging quality standards.

Test equipment and procedures

Without consistency in the laboratory to augment that of the manufacturing process, quality improvements programs will reach a plateau and advance no further. In fact, process engineers may be misled by inconsistent testing into making decisions that actually introduce greater variation.

Lab consistency hinges on both the equipment used to determine physical properties and the operators who control it. New automated digital durometers are replacing analog equipment in advanced laboratories, improving the accuracy of measurements by carrying them to an additional decimal place. Operator-introduced variation is also reduced, since there is no need to interpret a reading on a dial.

Plastometers can be enhanced with digital readouts and automatic printers, which are pre-set to take a measurement at a specified moment during the test, according to industry standards. Data are recorded with precise timing, and a hard copy of results is generated without operator intervention.

Even the process of cutting test samples has been scrutinized. Lab technicians have found that as a die begins to lose its sharpness over hundreds of cycles, tensile properties of the cured samples tend to drift. A new cutter design with interchangeable blades produces more consistent samples and greater test reliability (figure 2).

To ensure optimum control in lab ovens and fluid baths, their heat sources, seals and blowers receive high maintenance priority. Digital recorders constantly monitor temperatures during heat aging and fluid immersion testing. These improvements most directly impact the automotive industry, where OEM standards require extended testing, and consistency throughout the duration of the procedure is a key to accuracy and repeatability.

At times, there may be a need for consistency even beyond accepted standards. When a sample rest period is part of the test regimen, for example, standard methods may require that a cured test slab be conditioned in a controlled environment for 16-96 hours. STI quality engineers feel strongly that using a sample rest of the same duration in every trial will produce more meaningful, repeatable results, and a consistent time period is used for all materials tested under such methodology.

Similarly, sample weights are kept consistent from test to test. Even though accepted methods may suggest a sample size ranging from 11-13 grams for specific procedures, an exact sample weight will be used every time to standardize the parameters.

Detailed interactive SOPs are also a key to consistency in the laboratory, much as they are in compounding. Even calibration of test equipment has been improved through their use. In the past, an outside technician would examine equipment periodically and adjust it to standards, but this practice had a tendency to introduce variation.

Calibration checks now begin by running internal standards on the equipment and control charting the results. Documentation includes NIST (National Institute of Standards and Testing) traceability, tolerances and any adjustments or maintenance performed. A second standard is then run and charted, and the equipment is returned to service when the final check is within the established control limits.


The objectives of consistency within the lab could not be realized without implementation of SPC to the testing process, in much the same way that statistics are used in compounding. On each shift, lab technicians test standard lots of material and control chart the results. Subsequent analysis determines what differences exist over time between operators (on the same piece of equipment), and whether significant variation is present from one piece of equipment to another. The goal is to have all lab personnel arrive at precise and accurate results on the standard samples. In addition, the quality program at STI includes a round-robin procedure under which the different geographic regions test identical samples in order to further examine the reliability of the results from each individual site. Technicians are certified through written examination and observed performance.

Coefficient of variation

Statistical measurement is becoming widely recognized as critical to quality improvement, and its utility is becoming increasingly apparent. Statistics are employed to monitor physical properties, process data, lab variation and even customer service. By collecting detailed information on all facets of the compounding process, the custom mixing of silicone rubber is reaching unprecedented levels of quality and consistency. In order to assess the progress of its CI programs, STI is using a measurement tool called coefficient of variation (COV), developed as an overall barometer of the entire body of complex statistical information. COV is a calculation derived by taking the standard deviation and dividing by the average, then multiplying by 100 to generate a percentage. That is sigma/X-bar x 100.

In itself, COV is a unitless number, and does not provide any specific guidance on sources of variation. What is does provide is an overview sense of the CI program's effectiveness. The technique is applied to specific properties that are tracked as part of the statistical program, as well as to the plant's overall performance (figure 3).

The aim in using COV is to determine whether the quality improvement tools being employed are having the desired effects over time. If the CI program is successful the long-term trend should indicate that the coefficient is steadily dropping. COV also serves as a way to monitor progress in improving lab consistency. The coefficient can be determined either from very focused data (the variation in rheometer testing, for instance) or on the testing function as a whole. Part of the value in using the technique is its ability to describe variability without regard to specifications. It is a pure indicator of process and product variation.

Results of CI

Fabricators are seeing the benefits of CI programs, and most are developing their own improvement techniques. Some integrate the compounder's SPC data into their quality control efforts. A few have reduced their physical property testing to an audit level, rather than testing every lot of material.

In some cases, end users who have seen their materials consistently in control for long periods are no longer requiring that full-scale testing be conducted on every batch. The practice is a testament to the success the quality program, one that brings tremendous cost savings to the fabricator. Since some tests require an extended time to complete (oil immersion and heat aging procedures), order-to-ship times can be drastically reduced, as well. The view of the CI program at STI is one of a constantly maturing definition of performance quality, based on specific customer input. Direct participation from fabricators and OEMs is essential in determining this definition. Customer feedback is seen not simply as a means of assessing product and process quality, but as a strategic tool for management of the company's future.
COPYRIGHT 1993 Lippincott & Peto, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:continuous quality improvement
Author:Eloph, Chris
Publication:Rubber World
Date:May 1, 1993
Previous Article:Synthesis and morphology of TPEs.
Next Article:Jobs & People.

Related Articles
Custom mixing of silicone rubber.
Curing rate and flowing properties of silicone rubber at injection molding.
High performance precipitated silica equivalent to fumed silica in silicone reinforcement.
Liquid silicone rubber gasketing materials.
Moving from quality assurance to continuous quality improvement.
How we straightened out the meal service using CQI: a small example of a major approach to problem-solving.
'CQI' software goes beyond SPC.
A processing guide to silicone rubber extrusions.
A processing guide to silicone rubber extrusions.
A processing guide to silicone rubber extrusions.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters