When temperature is fouling up the measurement process.









Raye, a regular reader of my monthly rants here, sent me a problem I am confident many readers experience. His company makes bushings out of brass or bronze that can range to 100 inches or more in diameter. He notes that while his lab is held at 68 F, it is not uncommon for work to be at 90 F surface temperature when it comes in for final inspection.

Air conditioning the shop is an obvious method of getting things under some kind of control, but that can be a very hard sell to management. Raye notes that they may leave the work for up to one day to cool down, but this can cause a delivery delay. If the part is a reject and the machine has been producing parts while the first off gets measured, more than one reject may be involved. If it’s a one off, a machine may sit idle waiting on a verdict before moving on to other work.

To make matters more interesting, the composition of the materials involved means differences in their rates of expansion or contraction and adds another dimension to problems caused by temperature variations from the standard of 68 F or 20 C.

When temperature is fouling up the measurement process, there are a couple of ways to look at the problem. The first is to control the temperature better, the second is to compensate and adjust for what it is. Before you can do either, you have to know what you’ve got to deal with and how it came to be. Monitoring the shop temperature over the working day will provide useful information. The only thing worse than a high or low temperature is one that is moving around all the time.

Controlling the temperature can mean controlling the environment in which the part is made, but that can be ineffective if the machining operation is the one that is causing the most grief. Flooding the work with coolant during machining may make the difference. That, combined with less aggressive machining cuts, could bring things down to a workable range.

If controlling the overall environment is not in the cards then go after the component temperature on its own-that is what you really need controlled anyway. A couple of fans circulating air around the work may do the job for you. For extreme temperatures or rapid cooling, a more aggressive approach using a liquid heat sink may work best. Simply put, it’s a tank or tray the work is immersed in while machine coolant or other liquid is circulated around it-a cold tub, so to speak. This process can be accelerated by hooking up a refrigeration system to keep the liquid temperature low and stable.

Not to be forgotten in all of this is the machinist who has to measure the work during machining. Monitoring the process at his/her end can provide a lot of useful information. Record the temperature before and after machining along with the machinist’s readings of size. If you do this with a variety of parts and materials that you are likely to encounter you can create charts that can save a lot of guessing. These charts detail how the machinist can measure the work wrong so it will be right when it gets to the correct temperature.

Many skilled machinists already use this concept with great success. They know the work is hotter than it should be, they know it will drop in size when it cools down and machine it “oversize” knowing it will drop in temperature into the size limits. A series of charts based on measurement data can aid this process, particularly for machinists who don’t have adequate experience to wing it.

There are temperature-compensating devices that work with electronic measuring instruments to show the “size” of the work at the right temperature even though it may be many degrees away from that ideal. Another approach I’ve seen involved rail car axles being machined for bearing journals. The company attached a steel disc to each one prior to machining so it became the same temperature as the axle after machining. The disc had been calibrated to the size being machined and became a master used to provide a final setting for the dial snap gage used to measure the work.

Whether you control it or correct for it, there are a number of ways to prevent temperature from messing up your measurements.



Hill Cox president of Frank J. Cox Sales Ltd. (Brampton, Ontario, Canada). He may be reached at [email protected].