Northwest Analytical’s Quality Analyst allows Willamette's Wilsonville Lab to quickly and easily analyze product data.

A container that jams a packing machine and shuts down an entire line has a major impact on a company's production and profits. For Willamette Industries, applied statistical methods are powerful tools in meeting this challenge. Statistical quality control (SQC) is used throughout the company to assure customers that specifications are being met, to make operations more efficient, and even to reduce waste. The key to achieving these benefits with SQC is reducing variation, both in the manufacturing process and in the finished product. 

Willamette's West Coast Development Lab in Wilsonville, OR, is a key player in the company's quality program. The lab performs testing and analysis for two paper mills, eight corrugating plants, one finishing plant, and two bag plants for the company's Western Region. In addition to testing finished paper and products at the lab, the staff also does in-plant quality audits, serves as a consulting group for process capability analysis and improvements, performs field evaluations at Willamette customer sites, and does annual supplier certifications. 

Because the lab is not connected with any one Willamette plant, it generally is viewed by both customers and the plants as a neutral authority. If a batch of containers doesn't meet customer specifications, the lab can use its test results to help assess where the problem lies and then mediate between the paper mill, the corrugating plant, and the finishing facility to help resolve it. Similarly, if a customer has a problem with a Willamette shipment (such as packing equipment jams), the lab can help the customer ascertain whether the problem is with Willamette's containers, the environmental conditions, or with the packing equipment itself. 

The Wilsonville Lab's success also is due to its rigorous adherence to testing standards and effective analysis. In addition to gauges and testing machines, the lab relies on NWA Quality Analyst (QA), an off-line statistical software package made by Northwest Analytical, Inc. The specialized statistical software enables Willamette to quickly and easily analyze data, as well as produce high-quality charts that enable both company employees and customers to see and interpret test results. Many of the plants also use this same software for statistical process control. 

To produce reliable and convincing analysis, the lab must control variation in its own methods, ensuring gauges are correct and maintaining standard environmental conditions. The goal is to spot any problems early and correct them before they create a shutdown situation. To achieve this, the lab collects temperature and humidity data three times each day and analyzes it with NWA Quality Analyst, which automatically produces x-bar and range control charts. The control charts graphically show the extent to which key variables (such as humidity) are in or out of statistical control. Each chart includes upper and lower "control limits," calculated to distinguish between in-control points (inside the limits) and out-of-control points. With the specialized SQC software, the lab technicians do not have to calculate the upper and lower control limits themselves. The software does this automatically based on the data entered. Once assured that the data is statistically in control and meeting the specification limits, the lab technicians look for trends. If the measurements start trending toward either of the limits, the technicians can clearly see something is starting to go wrong and can make changes before the situation becomes serious.

The Wilsonville Lab's use of statistical methods plays an important role in reassuring Willamette's customers that the containers they purchase meet specifications. The lab tests the paper from which the corrugated is made, the corrugated board, and the finished containers. Paper testing includes moisture content, caliper, STFI (short span compression), mullen, slide angle (slipperiness), and the scott bond test (which measures the strength needed to pull fibers apart). Corrugated and container tests include manufacturer's joint skew (a critical variable in causing packing machine jams), slot depth, glue line width, pin adhesion, edge crush, and register (to make sure print lines up). After collecting the data, the lab uses NWA Quality Analyst to perform a variety of statistical analyses to make sure that the product (paper, corrugated, or finished container) is being made to internal and/or customer specifications. Much of the analysis is done with x-bar and range charts to make sure the product is within the upper and lower statistical control limits (which may be different from specifications). Once assured the measurements being taken are in control (and are therefore statistically predictable), the technicians can perform a process capability analysis to see whether the product meets specs. The technicians also may perform Pareto analysis, which charts defect rates and shows types of defects by frequency of occurrence or cost. When the results of these tests reveal the product is not meeting specifications or there is a trend towards unacceptable variability, the lab teams up with plant supervisors and operators to solve the problem. Any product that does not meet specifications is scrapped. In-lab testing also plays a vital role in determining whether new product or customer specifications can be met. Before mass production begins, the lab acts as Willamette's own supplier certification tester, making sure that quality is maintained before it ever becomes an issue for the customer. 

While Willamette uses in-lab analysis to detect whether finished product meets specifications, the lab also uses statistical methods (called statistical process control or SPC) to detect process variation that eventually leads to non-conforming product. Several times a year, Wilsonville Lab technicians visit and audit manufacturing facilities. Measuring key process variables, they analyze the data on-site, work with the operators to adjust the process, and then regraph to see whether variation is reduced and/or output is improved. These audits help to correct process problems, make improvements, or show that no changes are needed. Audits also provide plant management with information needed to make larger process changes. For example, a recent in-plant analysis by the Wilsonville Lab used NWA Quality Analyst to show how a reduction in variability of adhesive viscosity, temperatures, and gel points could justify the purchase of a new starch kitchen with automated starch batching. 

Plant managers support the lab's work because it helps them improve customer relationships. Their analysis and reporting allow the plants to show their customers they are capable of meeting specifications. Also, it helps the plants to be more cost-effective and efficient by improving processes and reducing waste. 

Occasionally the lab also goes to customer facilities to do analysis. If a customer is having a problem with a corrugated product, the lab can help determine whether the problem is with the product, environmental conditions, or the client's process. Then the Willamette team works with the customer to make improvements.

Maintaining good relationships with customers is the key to success, and Willamette's West Coast Development Lab plays an important role in that relationship. The lab's stringent testing and analysis, presented in NWA QA charts, provides convincing proof that customer specifications are being met. The lab also provides important—and independent—feedback to Willamette's manufacturing facilities, helping them to improve processes, meet specifications and reduce waste. In an industry that is being driven by customer change, Willamette's philosophy is to get ahead of that change, continuously improving the company's ability to meet tight specifications before customers ask for them.

back to top