Temp Range For CMM Environment

[parrottk]

Can someone tell me what is the best working temp for a CMM?
We have a PFX and a Global.
I think it should be 68-72.

Thanks

[Matthew D. Hoedeman]

67-69, no more than 2 degrees variation is best.

[craiger_ny]

Depends on tolerance. All of our machines are on the shop floor and temps range a great deal. Our biggest problems do not come from big temperature ranges as much as from fast changes. Gradual temperature changes we do alright with, fast, abrupt changes do have detrimental effects.

What kind of tolerances are you guys holding?

[T Miller]

Golbal 7 7 7
right or wrong I try to keep my room at 70 deg year round.

[Jan d.]

What's the spec for your machines? Typically it is 66 to 70.

ISO standard number 1, yes the very first one that the world could agree on, says that all measurements shall be taken at 20C. I agree with Matt though, the closer the better. 67 to 69 would be real good for close tolerance work. 72 is too high. Gradients in the machine space and abrupt changes (like air vents right above the machine) can also have very negative effects on machine performance.


Jan

[Matthew D. Hoedeman]

Now, all that being said, here is some more to 'chew' on:

Your machine, all by itself, can be perfectly in calibration at 110 degrees AS LONG AS it was calibrated at that temp AND the artifact used for said calibration was at it's correct calibrated temp (67 to 69) and the temp of the room/CMM stays the same. HOWEVER, anything you measure that has thermal expansion capabilities (everything, in other words) will change it's true shape at that temp. and your real-world readings will be skewed.

If your company will not heat your lab above 65 in the winter or cool it below 75 in the summer, that is too much temp change. A good part in the winter could be too big in the summer, even if they are EXACTLY the same, when checked at the same temp. They need to figure out which is 'cheaper' for them, heating or cooling and then set the temp at which the lab will stay. If cooling is cheaper, 66 to 68, if it is balanced, 68 to 70, if heating is cheaper, 70 to 72, but keep the range down. Anyway you slice it, the temp can have a BIG effect on what you are measuring as well as what you measure with.

[Bill **********]

As Matt said your machine can run at almost any temp. Here is the rule that I go by. Lab temp at the time of calibration +/- 2 degrees. The owner still comes around once in the spring and once in the fall and adjusts the temps in here. He thinks that 70 +/- 2 degrees means he can change the temp in here to be 68 in the winter and 72 in the warmer months. I track temp in here on a strip chart that records temp and relative humidity. When I see the strip climbing in the warmer months I know he's been in and I go adjust it back to 70 from 72. When I see the temps fall in the cooler months, again I know he's been in and again I go set the thermostat back to 70 from 68. I have a tag on the thermostat that says DO NOT adjust temp, but of course he gets to ignore it.

Bill

[cmmguy]

Your machine, all by itself, can be perfectly in calibration at 110 degrees AS LONG AS it was calibrated at that temp AND the artifact used for said calibration was at it's correct calibrated temp (67 to 69) and the temp of the room/CMM stays the same.

Temp for calibration is around 20ºC. When the temp deviates from that, because artifacts and CMMs are not usually made of the same materials, you can only calibrate the CMM when the CMM and the artifact are corrected for temperature at calibration back to 20ºC. That is why temp comp is important on CMMs that operate outside of the normal "window".

Short Term Variation is the killer when it comes to temperature.

[AJHVW93]

Thanks to all, For the different points brought up here on said topic ! I have had many discussions about this with management. Hopefully they will be more inclined to agree with me ,if they see the replys to the post. I know its every day knowledge to you and I!!

[John Kingston]

When Hilton Roberts gets in he'll be all over this like flies on a dead festering racoon at the side of the road.

Temperature swings will mess with the mechanical squareness if your machine.
You have this massive granite block that takes a long time to adjust or soak to the temp change. At the same time, the thin materials of the bridge are moving in all directions due to the varying thicknesses. Even when you get the temp stabilized, you may or may not (probably not), be where you were at calibration.
Hence the reason Hilton pushes for measuring a "standard" part or widget. Tracking this will help you to see those changes, and you will catch a problem before it bites you in the proverbial butt.

I can only hope I explained this as well as Hilton.

[parrottk]

Thanks All.
The range for our CMM Room over the past 3 months has been 68.3 F-69.5 F.
Theywant it warmer in there on nights 75 F
Now with the info I recieved from this forum I can dig my heels in and keep it as it is.

[Jan d.]

Thanks All.
The range for our CMM Room over the past 3 months has been 68.3 F-69.5 F.
Theywant it warmer in there on nights 75 F
Now with the info I recieved from this forum I can dig my heels in and keep it as it is.

That's great! Absolutely, dig in deep and do NOT change that! 68.3 to 69.5 is as good as it gets. Keep it there year round and you do not have to worry much about any adverse effect. 75 is rediculous. Don't fall for that crap.


Jan.

[jaymunday]

wggeerbeb

[NUGUY]

A sidenote: In many cases, the boys on the floor will be working and inspecting parts by their machine tool in thermal conditions quite different than QA lab conditions. We have a few parts that require holding the temperature the same in the shop as I have in QA, within 2 degrees F. We accomplish this by making sure the coolant is the same temperature as inspection temperature by inserting a thermometer in the tank, and adding cold/warm water as needed. Works terrific with aluminum parts.

[biohazard]

wggeerbeb