Putting Control Charts to Work
The Process Control column in Ceramic Industry’s April 2003 issue (p. 38) provided a very cursory look at the function of the control chart. While the control chart is certainly not the only tool available for statistical process control (SPC), it is the most important and basic.
If your company is not routinely using control charts, it is unlikely that significant or consistent control of its manufacturing processes exists. It is also unlikely that your company is aware of this lack of control. Trying to manufacture products without control charts is like trying to navigate in a strange city without a street map—you really can’t know where you are.
The payback that can be generated by the modest investment needed to put control charts to work can be enormous.
Basic RequirementsTo make effective use of control charts, four requirements must be met. Your company must have:
• Vision, commitment and leadership at all management levels, but especially at the top.
• Personnel trained in ceramic processing and technology.
• Personnel trained in basic statistical process control procedures.
• A good computer-based SPC software package and accessible computer hardware to go with it.
The first requirement is by far the most important. If your company lacks these qualities, you might as well quit reading this column now.
Are you still with us? If so, please go back and read the first requirement at least one more time to be certain it is met.
The second and third requirements go hand-in-hand. How many of your supervisors, foremen, technicians, industrial engineers, inspectors and managers are trained in ceramic technology? They don’t have to be ceramic engineers. Don’t waste your time training personnel in using control charts or other SPC tools unless and until they have a good basic understanding of practical ceramic technology. Without such training, your staff will not be in a position to act upon the information provided by control charts or other SPC tools.
Personnel also need training in SPC, especially as it applies to ceramic or glass manufacturing. Where can both types of training be obtained? One good source is Ceramic Correspondence Institute (CCI) (see http://www.ceramics.org/education/cci.asp). CCI offers three courses that answer the basic needs of requirements two and three: Fundamentals of Ceramic Technology, Ceramic Processing, and Statistical Process Control for Ceramics and Glass. For the glass industry, Glass Technology would replace Ceramic Processing. Other sources of training in SPC also exist, but ceramic technology training can be hard to find at other organizations.
There are a lot of good (and some not so good) SPC software packages available. If you would like help in finding and evaluating them, drop me an e-mail. It doesn’t take a fancy computer to run SPC software, but a large hard drive and a good inkjet color printer are desirable. A really good monitor is essential—the larger the better.
Using Control ChartsAt what manufacturing steps in your process should you employ control charts? The April column showed an example of batching data, and that is certainly one area where control charts are very useful. Another is charting critical raw material properties. You might also want to chart casting slip properties, body plasticity, particle size distribution information, pressed density, formed dimensions, moisture content before and after drying, size after drying, drying shrinkage, warpage, dry density, green strength, firing conditions, firing cones, fired size and dimensions, firing shrinkage, water absorption, defect levels, and other properties, depending on the nature of your product. For example, if your product is an electrical ceramic, there are a variety of electrical properties you would want to chart. Some property measurements might be redundant, such as porosity and density—then again, both of these measurements might be useful for your particular plant. There is no magic list. It is better to begin by tracking too many processes and properties, and then eliminate those that are telling you the same thing. Correlation analysis can help you do this, but you need the data first.
I don’t think I can overstress the importance of properly used control charts. In fact, I feel so strongly about this that I challenge you to e-mail me a spreadsheet with process data, and I will prepare a control chart for you. We might need to discuss the nature of the data you will send, but I will do almost anything to show you how powerful this tool can be.