Material Temperature Sensor Placement for Measuring Linear Displacement
Material Temperature Sensor Placement herein refers to Measuring Linear Displacement with an Interferometer Laser Calibration System (henceforth referred to as “Calibration System") however the principle applies to any system under the same circumstances.

Not long ago, the rule was to place the Material Temperature Sensors on the Axis being measured for Linear Displacement then adjusting for the Positioning Errors from the results. This involves moving the Material Temperature Sensors from Axis to Axis as they are measured and adjusted. Unintentionally, this rule negatively impacted uniformity between the Axes. That's because the Axes are predominately at different altitudes and therefore at different temperature zones. The end result relative to the Workzone is, an inch in one Axis is not the same inch as it is in another. If the temperature zone is different between all the Axes, then they all position to a different inch.

An example of this is:
The Material Temperature Sensors are placed along side of the X-Axis Linear Scale. The Sensors feedback a temperature of 68° F. X-Axis is displaced 100" and the Positioning Error is adjusted for the 100". Since the material temperature is the standard of 68° F. the Calibration System does not compensate for material temperature and the Axis actually positions at 100" when the Calibration System displays 100".

Next the Material Temperature Sensors are placed along side of the Y-Axis Linear Scale. In this case the Y-Axis Scale is 15 feet higher than the X-Axis Scale and the temperature at that altitude is 4° F. warmer. Now the Material Temperature Sensors feedback a temperature of 72° F. This causes the Calibration System to compensate for the 4° over 68°. With the Material Coefficient of Expansion set at 6.5 PPM, Y-Axis has to position at 100.0026" for the Calibration System to see that as 100". Any Position Error is adjusted so that the 100.0026" position represents 100.0000".

Next a 100.0000" square Part is machined in using X and Y-Axes. For the sake of explanation, the altitude or temperature zone that the Part was machined in, is the same as the X-Axis Scale's. The Part would be 0.0026" longer in the Y-Axis direction than it is in the X-Axis direction.

A few decades ago, it was realized that an improvement in quality is to place the Material Temperature Sensors in the Workzone where the machine performs it's tasks, then correct the Linear Positioning Errors for each Axis without moving the Sensors. The end result of that is, the Axes all position relative to their ultimate common objective rather than to themselves individually steered around by various temperature zones. Their individual inch is closer to the same inch.

An example of this would be:
Imagine we have to measure a 100" square part with a Coordinate Measuring Machine. We have a very tight tolerance on the part and we want to check the CMM to make sure it will measure correctly at 100". But, we don’t have a Calibration System to check it with. We do however have a 100" long Gage Block that has the same Coefficient of Expansion as the Part, 6.5 PPM. So we place the Gage Block on the CMM where we plan on measuring the Part, precisely align it for the X-Axis direction and measure it with X-Axis. If the CMM doesn’t measure the Gage Block as 100" then we correct the error in the CMM so that it does.

Next we rotate the Gage Block 90°, align it to Y-Axis and repeat the same process we used with X-Axis to correct Y-Axis. Now if we measure the 100" square Part at the same coordinates that we had the Gage Block at it will measure 100" square if it actually is 100" square. It has to because it has the same Coefficient of Expansion as the Gage Block, the CMM’s Linear Positioning Errors where corrected at the same coordinates, and both Axes Linear Positioning Errors where corrected for and tied to the same temperature zone.

What would have happened if we would have moved the Gage Block up by the Y-Axis Scale on the CMM and adjusted Y-Axis from that there? In the same scenario as the machine in the first example where it is 4° warmer up there, the Gage Block would warm up and expand 0.0026". If Y-Axis was adjusted to that 100", then when the Gage Block is placed back down where the Part gets measured, the Gage Block will cool back down again and contract 0.0026". Y-Axis will measure the Precision Gage Block as 99.0074" because it has been adjusted to position to a dimension 0.0026" longer than the 100" the Gage Block is at now that it is in a cooler temperature zone. Y-Axis will also measure it's side of the 100" square Part as 99.0074" if it where 100" wide too.

Wait a minute, what about Abbe' Offset Error? Putting the Gage Block up there is way off from the Workzone and from the Calibration System's correct beam path isn't it? Instead of doing that lets just leave it down there in the Workzone aligned to Y-Axis where it originally was. Since it doesn't have Material Temperature Sensors, heat it up 4° so it will expand to 100.0026" long then set Y-Axis to 100".

That represents the same scenerio as calibrating Y-Axis to 100" with the Calibration System's Material Temperature Sensors feeding back a 4° warmer temperature than that area of the Workzone. The difference is, when the Gage Block cools off, Y-Axis will still measure it 0.0026" too short because it doesn't have Material Temperature Sensors out there in a remote area telling it how long it's supposed to be in the Workzone.

The difference between heating up the Gage Block and heating up the Temperature Sensors is, the Calibration System only has to adjust the values that are displayed, recorded, axes are positioned to, and calibrated to. The effect is the same. The Gage Block's dimensions don't change simply by rotating it 90°s. However, The Calibration System's laser length standard can easily be changed simply by Material Temperature Sensor Placement. That in turn can change the calibrated length of two Axes 90°s apart.

Likely, the average shop doesn't have machinery as large as the examples mentioned here so the results wouldn't appear to stand out as much. Proportionally the results would be the same under the same relative circumstances, only on a smaller scale . Regardless, not moving the Material Temperature Sensors is a worthwhile step however small the end result. It requires no extra effort not to move them. It’s a free gain in the Quality of the Product and the Machine's Performance.

I am working on GEISS 5 axis CNC machine. I want to measure the temperature of the CNC by placing sensors on it, Could you please help me on the best places where the temperature of the machine can be measured by placing the sensors. Could you help me on it.
Thank you
Ahmed 07/01/2011
I would put them on the machine structure like the columns (not sheetmetal) at a height that would be in the middle of the workzone for the parts to be machined.
Admin 07/01/2011
Are we calibrating for machine growth or the feedback system? You need to compensate for the change in temperature of the glass scale, steel tape or the ballscrew, depending what is used for feedback. Just because the workzone is 75 deg yes the steel part has grown but the glass scale with its air purge and 68 deg temp will move 1 in and laser will correct 0 for scale error and something for the laser wave length, so when at 68 air temp both will = 1.

What am I missing?
Mike Knemeyer 01/04/2012
Mike, The point of not moving the material temperature sensors for each axes is so that one inch for example will be the same for all the axes.
Admin 01/04/2012
Introduction to Interferometer Lasers