Quantifying for Quality

This is the first “Measurement Matters,” a column that will show how much measurement matters in achieving and maintaining quality, especially in manufacturing. I will cover topics such as types of measurements and measurement devices, how measurements are made, what can be done with the measurement data, what can influence measurements and the interpretation of measurement data. Some of the topics may be common knowledge to many of you, but I hope there will be something in every issue that will be useful to you, or at least that makes you think. Let’s start with a basic definition. Measuring is the act of quantifying something. That quantification is a measurement. Take this very simple example: You can tell that something is hot or cold—that's a qualitative assessment of temperature. Such an assessment isn't a measurement; determining the actual temperature of the item is measurement. Another example is the diameters of golf balls and basketballs. We can tell by looking at them that basketballs are bigger. That's a crude measurement with nothing more than our eyes. But until we use some kind of measuring tool, we don't know how much bigger the basketball is—we need to quantify it. Bigger and smaller, hotter and colder, and other such examples are qualitative measurements. Saying,“It is warmer today than yesterday” is a qualitative statement. “It's 4 degrees Fahrenheit warmer at noon today than it was at noon yesterday” is a quantitative statement.

The science of measurement is metrology. Many times you'll hear people use measurement and metrology interchangeably, but a metrology system provides measurements. Using a measuring device to collect measurement data is part of the science of metrology. A scientist of metrology is a metrologist.

Let’s go back to the simple hot/cold example.

Considering something to be hot or cold is a relative determination of the magnitude of the hotness or coldness of the item. When two items are at opposite extremes of temperature, it's obvious that one is hot and the other is cold. There can even be visual cues such as the orange of a hot ember, or visible ice crystals at the other extreme. However, when the difference in temperature gets very small, one item may be hotter or colder than the other, but you may not be able to discern the difference. They may feel equally cold or hot to the touch. This brings up the issue of resolution—your finger may not be sensitive enough to indicate that the temperatures aren't equal. It may take a sensitive thermometer to show that there's a slight difference in temperature.

This raises another measurement matter—how accurately do you need to know the difference in temperature? It depends on the application or requirement. As long as two items feel like they're at the same temperature, that may be good enough. At the other extreme, it may be necessary to know temperature differences to fractions of a degree, because of thermal expansion affecting the fit of two mating parts. Measurements are important in context. Although it's possible to measure to small fractions of a degree, it may not be necessary. On the other hand, lack of measurement resolution may force decisions that can have severe implications.

As you might expect, there's the important measurement matter of selecting the proper tool for the measurement. There are a number of considerations in the selection process including the range and resolution of the measurement that is required, where and when the measurement is to be made, how important the measurement is (which can influence how much money to spend on it, which can, in turn, affect which technologies to employ), and the degree of difficulty involved in using the tool properly. Determining what to do with the measurement data is another subject with multiple considerations.

As you can see, there are many factors to consider about measurement. In future columns we will discuss many of these factors individually. One example is measurement variability. When measurements are repeated and the results vary, did the part change between measurements? Did the measuring device drift? Did the measuring environment change? Did the operator of the measuring device change? And, most important, did the measurement vary enough to matter?

There are also calibration issues. Is absolute measurement always necessary? How often must I calibrate? What’s the difference between calibration and certification?

There are lots of measurement matters that you may think about every day, or may have never considered. Companies that provide a service or sell measuring machines think about these things, so you don’t have to. But understanding measurements can help you better understand your measurement data and make better business decisions with those data.

Until next time, remember, measurement matters.

source www.qualitydigest.com