Video: Metrology Training Lab–Influence of Temperature on Measurement

Hello, I'm Jim Salsbury with Mitutoyo America Corporation and welcome to the Metrology Training Lab. In this episode, we are going to introduce you to the type of problems that we see in dimensional measurement associated with the influence of temperature. You may have noticed that some of today's more advanced measuring equipment has built-in temperature compensation systems. These systems are standard on most coordinate measuring machines, but even this linear height gage has that function. As you can see in this menu, the software gives you the option, if you want, to enter material properties as well as temperature information.

So what does this do and why would you use it? All materials grow and shrink as the temperature changes. You may think these changes are small, but every measurement is impacted by temperature in some manner. You can't avoid temperature influences - they are always there. What we all have to do is understand and potentially reduce the errors from temperature to an acceptable level. So, first of all, all dimensions, all sizes, all lengths, all geometric features are defined at exactly 68°­F, which is also equal to exactly 20°C. This is the official legal reference temperature found in all national and international standards. So, in theory, all your measurements should be done at exactly 20°C. Of course, that is never possible - you never can be that exact, and therefore you always have errors from temperature. In dimensional calibration labs, you may find environmental control at say around 18 to 22°C, and if you are just calibrating small micrometers and calipers, the thermal errors are usually small enough to not cause any concerns. But when you have longer measurements, or higher accuracy equipment, something like this high accuracy linear height gage, then temperature can suddenly be a big deal.

So let's talk about some good practices to reduce the impact of temperature. First of all, you want the temperature to be stable. You want to avoid large doors that frequently open – particularly large roll up doors. You also want good airflow and not air blowing on and off right on top of your measurement. You want to deflect the air away and around your measurement to let it mix together. Also avoid heat sources close to the measurement, like lights and computer systems. People are also large sources of heat, so be careful with large numbers of people, and people coming and going too often. For more critical measurements, try to set up an area that is separate from the rest of the work going on around you. You also want to allow your measurement setup to reach equilibrium in temperature – a process that is often called thermal soak out. If I bring a steel gage block like this over to this linear height gage to be calibrated, I want the linear height gage and the gage block to sit next to each other long enough so that they both reach the same temperature. In our calibration lab here at Mitutoyo America, the practice is to let things sit overnight in the lab.

Now this is an extreme example, but when calibrating gage blocks, we arrange the customer's gage blocks, which are the white ceramic ones here, one by one, next to our master blocks, these steel ones here. The goal is to have the two blocks be at the same temperature. We also use this aluminum plate to help with the heat transfer. If you have a stable environment and things thermally soak out properly, you've taken care of many temperature problems. But what if the average temperature in the room is not the reference temperature - not at 20°C, not 68°F. That's a bit cold, and many labs do run a few degrees warmer than that. Since that situation is common, and not easily fixed, let's look at the math so you have a way to estimate the errors to see if that may be a problem for you. All materials expand as temperature increases. Different materials expand by different amounts. You need to know what is called the coefficient of thermal expansion, or CTE, of different materials to understand the impact. If you look in the Mitutoyo catalog, you will see the published CTE for our gage blocks. Our steel gage blocks expand at 10.8 parts per million per degree C. That means 10.8 microns per meter or 10.8 microinches or millionths per inch for every 1 degree C of change. At 21 degrees C, this 1 inch steel gage block will have grown 10.8 millionths of an inch bigger than at the reference temperature of 20°C. Mitutoyo ceramic gage blocks have a slightly lower CTE, at 9.3, not 10.8. Mitutoyo actually designed the ceramic gage blocks to have a CTE close to steel to reduce thermal problems when they are used. Like a lot of measuring equipment, this linear height gage uses a linear encoder, sometimes called the scale, which is made out of glass. In the operation manual for this instrument, you will find the stated CTE to be 8.1 parts per million per degree C.

Just as an example, let's do a simple calculation to see what happens if we calibrate this linear height gage using a 500 mm – or about 20 inches – steel gage block in a typical office temperature around 22°C. The equation looks like this: The change in length equals the temperature that we're at minus 20°C, that's the reference temperature, times the length that we're measuring (L), times the difference in the CTE between the linear height scale and the gage block material. So we're saying that we're at 22°C in this example, so it's 22 minus 20. Now we were at a 500 mm gage block, so that's half a meter and then the two different CTEs, the gage block, the 10.8, and the scale for the linear height, the 8.1. You can see how the units here, microns per meter per degrees C, that's the same as saying 10.8 parts per million and the 8.1 parts per million per degrees Celsius. Alright, so that's all the numbers that we need and then we can just do some simple math here. The 22 minus 20 gives you 2, pull down the half meter, the difference here is 2.7 microns per meter per degree Celsius then in the end that number comes down to 2.7 microns which, if you prefer inches, is pretty close to about 10 thousandths of an inch.

So is that a problem? Well, if you look up the accuracy tolerance for this linear height, the 2.7 micron temperature error is about twice the accuracy specification for this unit at this length of 20 inches. So that means if you calibrate this unit in this environment, even if the temperature is perfectly stable and the gage block and linear height reach perfect thermal equilibrium, if this instrument, let's say your height was perfect, it would be found to be significantly out of tolerance if the room is just 2 degrees off the reference temperature of 20°C. A good unit will look bad because of temperature. When I said that temperature impacts all measurements, this is what I was talking about.

So what do you do? Well, now you can understand why this unit has built-in temperature compensation. Its purpose is to automatically correct this exact error that we're talking about. For measuring systems with temperature compensation, you should use it. Without temperature compensation built into the unit, you have two other solutions. First, you could calculate the error yourself, you can use the equation I just showed you, and correct all your readings. However, the best solution is to find a much better environment. This gage should really be calibrated in a room that is good to at least 20°C ± 0.5°C or better. Our goal in this episode was to introduce you to temperature effects in dimensional measurement and to give you a few tools to understand them and improve your accuracy. In later episodes, we'll talk more about other temperature errors and ways to reduce those. Thank you, I'm Jim Salsbury, and I'll see you next time from the Metrology Training Lab.