PFFC - December 2008 - (Page 12) PROCESS MANAGEMENT Special Causes: Enemy or Opportunity? ll converting processes are subject to variation, and product at the end of the line is shipped within a narrow window of performance specifications. This kind of fluctuation for a process running in statistical control is known as common cause variation, of which 99.73% falls within calculated upper and lower control limits. When trouble strikes, it is the sudden onset of special cause variation. It is communicated in many ways from consumer company to converter. An example would be: “My filling line is down because the print is scuffing off.” The immediate response is to make sure the line is back up and running. After that, there are choices, such as “start a witch hunt,” defined by Wikipedia as often involving moral panic, mass hysteria, and mob lynching! A better approach, of course, and the one recommended, is to look for a root cause. How is this done? Of all the tools for analyzing data, the control chart is the most useful and needs to be in place. No other tool captures the voice of your process better. Research by statistician Dr. Walter Shewhart indicates that by establishing upper and lower limits at three times the standard deviation of the process (plus and minus, respectively), 99.73% of the common cause variation will fall within these limits. Therefore, a process is said to be in statistical control when the process measurements vary randomly within these control limits; that is, the variation present in the process is consistent and predictable over time. The upper and lower control limits are not the same as tolerance or sales specification limits. Control limits are a function of the way your process actually performs over time. Sales specification or tolerance limits are a function of what your process is designed to do and must be within the capability of the process; otherwise, constant “tweaking” is required, leading to waste of time and materials. In my experience, special cause variation seems to be inevitable, even in the best-run manufacturing operations. That being the case, the enemy has to be turned into opportunity. In the search for the root cause of a serious problem, many opportunities for improvement and reduction of normal variation are uncovered. In a specific case in which I was involved, an ink scuffing problem revealed issues with primer skipping, primer application weight variation, use of an incorrect ink formula, variation of foil wetting, and inconsistent press settings. Added to this, the QC ������ test specified by the consumer company was not ��������� the best predictor of filling line performance. A perfect storm of these factors all occurring at once caused a catastrophic failure in the field. The process could withstand many of these variations and deliver good product but not all of them at the same time. Not only was the variation outside of normal variation, it was also out of sales specification. ������������� The lesson learned here is that root cause analysis may not reveal a single reason for failure but will al������������������� ways uncover opportunities for process improvement in many different areas. These will include incoming raw materials specifications, machine settings, operControl charts are used to determine whether ating procedures, and test methods. your process is operating in statistical conSo, if the occasional special cause trol. Until it is, any improvement efforts is inevitable, the question becomes: are, at best, mere process tampering. How capable and prepared is your Basically, a control chart is a run organization for root cause analysis? chart that includes statistically generHow quickly can it respond? ated upper and lower control limits. This scenario is analogous to GREEN The purpose of a control chart is to safety training. You hope never to PRINTING detect any unwanted changes in your use it, but you had better have your process. These changes will be signaled staff prepared and trained to respond by abnormal points on the graph. in an emergency. A By David Argent Contributing Editor Run Chart ����� ���� ��������������������� COMING NEXT MONTH Process improvement expert David Argent has 30+ years of experience in process analysis with particular emphasis on ink and coating design and performance. Contact him at 636-391-8180; djvargent@sbcglobal.net. 12 | DECEMBER 2008 WWW.PFFC-ONLINE.COM http://WWW.PFFC-ONLINE.COM
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.