RST-T Temp Sensor

[[To...]]

We have a customer that routinely has measurement plans that take upwards to 5 hours to run. The cmm is in a temperature controlled environment but don't always have the luxury of letting the part get down to the ideal temperature. They're trying to find out if they can use the RST-T sensor to take a temperature reading at fixed intervals, i.e. every 20 minutes or so, and have the compensation adjust for the updated temperature. They've never used the sensor that came with their machine about 15 years ago.

[[Da...]]

It's a tiny little bit possible, but generally not.

Temperature compensation is a little tricky, because what happens under the hood is so incredibly basic, a lot less complicated than what some people think.

You can calculate a compensation value for a given length yourself: "Length" x "Temperature difference to 20° C" x "Temperature coefficient". The temperature coefficient depends on the material, for steel it's around 12 x 10^-6 mm (in mm world), the temperature difference is what your temperature sensors pick up, but the length, well, the CMM measures a lot of different lengths on a single part. And since length is part of the function, the compensation value gets larger the longer the lengths are. On the other hand, the compensation value becomes zero at an origin.

Where that origin is, doesn't matter at all, as long as the other factors don't change. But your customer wants to change the temperature difference factor constantly, and so he needs a defined origin. Now, this origin is NOT the same as your base alignment origin, since the deciding factor is how your part is clamped. Ask yourself, if the temperature changes, where will the part "go"? If the part is tightly clamped down, it can't move at all. It could bend, but that's not what you want, and that can't be easily calculated.

But if you make a fixture, where your part can move around a fixed point, like a freeway bridge on rollers, that might give you a chance to control the origin and therefore regularly changing temperature parameters.

Usually, Calypso invokes the temperature compensation only once, after calculating the base alignment. There is a command that is supposed to control the origin when working with a part made of different materials, when you absolutely need more than one temperature compensation, but you need to design the fixture to control the fixed origin.

Under normal circumstances: Don't do it!! Ever! It will most definitely mess up all your results, unless you know what you're doing. That is also the reason, why even with temperature compensation, usually your part temperature must not change too much from beginning of the measurement to the end. Stability is key!

[[To...]]

Thanks Daniel. The reason this question came up was something they read in the specs of another brand machine that they have. Is it deceptive marketing? I don't know. Maybe there are some users with experience on brand M cmms that have some feedback. With the exception of the phrase "and for the workpiece itself", I'd say it's on par with most cmms. But after digesting your post, I see the potential risks with variable temperature compensation.

Brand M.jpg

[[Da...]]

That seems somewhat deceptive, I agree. Always a good idea to ask a tech guy questions about this, I would love to hear an explanation for this.

[[Ri...]]

I have two new CMM's, neither have Planner or Stylus System Creator.
How do I add an RST-T sensor?
One CMM has the sensor, but the software will not allow me to export the PCF file.
I tried to import a PCF file from an Offline seat but the new CMM will not allow this, it was something along the Sensor ID was assigned to another system.

[[Je...]]

Please sign in to view this quote.

Variable temperature compensation is not supported in CALYPSO. The temp comp value is implemented at the time of Base Alignment and temperature readings after this point are merely for reporting and can be used for Temp Delta reporting.

My advisement to customers that are using temp comp is to build in a sequence of procedures that minimizes the risk of improper application;
1 - Controlled Heat-soak time (this is obvious and likely already implemented where possible)
2 - Set Temp measurements with delta alarms if T > X°_Delta - (PCM required to work in a robust manner)
3 - Construct measurement plan to measure LONGEST measurements first and LARGEST bores with tight tolerances so that the physical data is gathered at time nearest temperature reading. (This takes quite a bit of planning and forethought. Likely will increase inspection time due to inefficient travel paths)
4 - Break inspection plan up into several mini-plans and run them consecutively using AutoRun or a $Master inspection compiled plan ($Master is an advanced topic)

[[To...]]

Please sign in to view this quote.

This sounds like the simplest option to me.

[[Jo...]]

I have a similar situation, however my program is only 45 minutes. I measure a Datum at the beginning of run, and if remeasured at the end of run, it has changed .0006". The part movement is volumetric. It would be interesting to measure a couple of features at the beginning of run, and the end to quantify the Delta. You then have a better idea of error.

[[To...]]

Please sign in to view this quote.

I used to cover this technique in my Advanced training classes but it was more related to the stability of your set up, i.e. did something move?. I like the way you're using it.