The local environment can be an important factor in the quality of measurements. No matter what is being measured, the higher the resolution of the measurement (the finer the detail), the more significant the effect of environmental influences can be. The measurement you’re after is one of many variables.
When you’re measuring one thing, you’re actually taking a snapshot of a large number of variables. Most of the time those variables are so small that they have negligible effect on your primary measurement. In fact, they are probably so small they cannot be measured. But that doesn’t mean they aren’t there. And there are times when they most certainly do matter.
Here is a simple example: a piece of steel. If all you want to know is whether that piece is 12 in. long or 12½ in. long, a tape measure is all you need. However, you should remember from science class that steel expands as it’s heated. The simple act of picking up that piece of metal warmed it, but the result of that minor heating was insignificant for that particular measurement. In fact, the linear thermal expansion coefficient of steel is only 0.0000072 per degree Fahrenheit. What does that mean? It means that if I could evenly heat a 12.0000000 in. piece of steel by say 3° F, that steel will actually be 12.00026 in.- long [((0.0000072 x 3) x 12) + 12].
“So what?” you might say. “I can’t even measure to that resolution.” Think about steel used in bridge construction, and the swing in temperatures that bridge you drive over every day is subjected to over the course of a year. You probably know that bridges are made in sections with small gaps between them to allow for expansion and contraction—expansion joints. Think about the size of those gaps and how they change with temperature. If each single section of bridge steel is 100 ft. long and the gap between each section is ½ in., how much does the size of that gap change if the temperature goes from freezing (32°F) to 90° F?
[((0.0000072 x (90–32) degrees) x 100 ft.] =
0.0000072 x 58 x 100ft. =
0.00042 x 100ft. =
0.042 ft. or 0.504in.
This means that a ½ in. gap could close completely with a 58° increase in temperature assuming adjacent sections expanded equally. Too much expansion could buckle the pavement.
Heat and metrology
The bridge example above helps explain the proportional relationship between the gross measurement of a length of steel and the effect of heat on the overall length. As you can see, for a piece 100 ft. long, the effect is minimal. So, when is it important?
Carry this thought further to a common metrology reference standard—the steel gage block. Let’s scale that 100 ft. down to 12 in. (divide by 100). The linear expansion coefficient still applies, but now the change in temperature has a bigger effect on length. And keep in mind that the coefficient works no matter which way the temperature is varying.
The long and the short of it
By now, you understand that a length of steel will be changing as its temperature is changing. In other words, unless the temperature is constant, the length of steel will be different than it was at a higher or lower temperature. As the temperature goes up, it gets longer. As the temperature drops, it gets shorter. You can see why this means that metrology labs must maintain a constant temperature. Unless the temperature is constant it doesn’t matter how precisely you measure that gage block; it’ll be different at a different temperature.
Calibration labs commonly control and monitor temperature and other environmental conditions. A common specification is to maintain a certain temperature plus or minus one degree per a certain time interval, typically one hour, although it may be as short as 15 minutes. You can see from this specification that the specific temperature and the rate of change in temperature are important values. The rate of change is often controlled by limiting the number of people allowed in the metrology lab, because body heat can increase the overall room temperature. Depending on which reference you use, the human body puts out from 60 to 100 watts of heat. Feel the heat from a 100-watt light bulb and then imagine 5 people in a temperature-controlled room to understand the effect.
When you look at a calibration certificate, you’ll notice a section about the environmental conditions under which the item was calibrated. In the steel gage block example, what that means is that the gage block was xx.xxxxx in.-long at those particular conditions. In other words, unless your conditions match those on the certificate, the gage block won’t likely be that specific length.
Controlling temperature
A calibration lab’s temperature is controlled, and there are other practices put into place to deal with temperature effects on measurements. In the case of gage blocks, they are usually “soaked” in the calibration lab for at least 24 hours before being calibrated. They soak by resting against a granite plate that has been at the constant lab temperature for several days or weeks. In addition, the gage blocks are handled with tweezers so there’s no heat transfer from the technician’s hand. The intent of these controls is to minimize some of the variables that can affect the part and the resultant measurements of it.
Temperature in the real world
Of course, temperature isn’t controlled so tightly in typical manufacturing environments, where many measurements must be made. Even air conditioned shop temperatures can vary significantly over a 24-hour period, and some shops turn the air conditioning off overnight or on the weekend. The number of those “100-watt people” in the area varies during the day, and our 12-in. gage block out in the shop is changing commensurately, although not as quickly, because air temperature changes much faster than steel. This means that the thermometer on the wall may say the room is at the temperature on the calibration certificate, but that doesn’t mean that the steel gage block has reached that temperature.
Measurements don’t vary only by temperature
If only we had to think about just the effects of changing temperature... Depending on what parameter is being measured, other varying environmental factors can change measurements. Humidity can be a factor, especially if it’s so humid that moisture condenses on surfaces. Barometric pressure can be a factor. That low-pressure system bringing storms to the area can affect several types of measurements performed during that sunny high-pressure condition last week. Vibration affects many measurements and can come from a number of sources, including forklifts traveling behind the wall or the low frequency rumble of roof air-conditioning units and associated air vents. All these examples may be contributing factors to measurements, especially if they vary during the measurement process.
Too much going on
All of this environmental talk may make it seem that good, repeatable measurements are virtually impossible. That’s far from the truth. Although there are numerous factors that can affect measurements, most are so small as to be lost in the noise. The point is that whatever parameter you’re measuring is but one variable in a large equation that includes all the environmental factors. In most cases, most of those variables are simply too small to measure and, so, too small to influence the measurement. However, the better you can minimize any of those variables, especially temperature changes, the more likely your measurements will be accurate and repeatable.
So what about global warming?
Unless melting glaciers threaten to flood the metrology lab, metrologists should concern themselves more with the local environment they can control. Keep in mind this mangled expression: “The more things change, the less your measurements stay the same.” Until next time, yes, measurement matters.
Add new comment