Thermal expansion and accuracy in huge machines?

[Splint]

Hi,
considering the size of some machines like this one http://www.janicki.com/services/CNCmachining.htm how do the machines maintain such fine accuracy when the thermal expansion can make substantial dimensional changes to the size of such big machinery. Do they live in temperature controlled rooms? The new machine at Janicki looks like the structure that supports the gantry may be made from concrete which is a material which reacts slowly to temperature change, but that is the first machine I have ever seen which would be made of concrete.
How do they deal with it?

Cheers
Splint

[ESjaavik]

Polymer concrete is widely used in machines today. And the producers of any high quality machine puts a large amount of experience and calculations into controlling the temperature in the machine. This is done by distributing masses, cutouts, webs etc. at the right places. Also by using local temperature control and insulation materials. But controlling the room temperatures is also done. The machine constructors cannot take into account temperature differences in the machine's environment.

This I believe is one way to keep the dimensional changes in a selfmade machine within reasonable limits. Keep it in a room with a stable temperature. It's all the more important when you mix building elements of steel, aluminum and other that have great differences in thermal expansion. Ball guides on aluminum profiles may stay within the relaxed demands of hobby use in room temperatures. But in your garage, on a cold winter day, you could make banana shaped parts.

[Geof]

Never mind thermal effects how about tidal? I worked with a guy who had run a big shaft lathe in a shop built on a tidal river delta; ground level about 30 inches above mean high water mark. For critical work they had to take the final cuts at the same state of the tide as the machine was trued up at.

[DAB_Design]

I've ran some pretty large machines. For the most part, on normal tolerances (+/-.005 and up), it really wasn't an issue. There was only 1 instance that I can remember thermal expansion being a problem. We had to drill some holes in some wingskins several hundered inches long, and they had a tight true position tolerance and couldn't seem to get them right. After looking into possible problems with programming, and machine malfunctions, it was decided the problem was due to the temperature change caused by the vacuum fixture being used.

The vacuum made the part cooler than it was supposed to be. So each time the holes were to be drilled, the temp of the part was taken, and it was entered into a formula within the program that adjusted the position of the holes.

[Mcgyver]

for the most accurate work, regardless of machine size, they are temp controlled environments, no direct sunlight etc. coolant cooling systems.

one of the biggest problems with the highly accruate work is the measuring instruments. The comparitive long, small mass in contact with human hands is prone to thermal expansion

[Splint]

Thanks for your input guys. I know what you mean Mcgyver about thermal expansion in measuring instruments, that's why some instruments have plastic hand grips to act as an insulation barrier between the opperators hands and the metal of the instruments.

Thanks
Splint

[jphagen]

Interesting topic. Wonder where NASA thinks it is.. Seems the Other guys
with there low teck. plumming might save the day..
Think about it all this fine spec. then it goes out the door to where?
Just a non-climate controle area.. There goes the specs??
What I have found is they spec. a lot, the send it into neverland
where they have no control..
Once made a plug.. Could not be .001 over or under..
Looked at what they did with them.. One plumber with a sledge
the other with a holder and the drove them in..
Heat excanger plug..
So it goes..

[uscra112]

You've raised one of the major issues in production machining, especially aluminum, in industries from automotive to aerospace. The best way is to control the entire environment, including coolant (if any). Machine tools as long ago as the 1950's had coolant jackets in spindles to extract bearing-generated heat that distorted the machine frames and disrupted bearing preloads. Today machining line builders for auto plants use controlled-temperature coolant to maintain part temperature during machining, among other strategies, including making machine frames as symmetrical as possible. They also in some cases have active compensation with thermal sensors at key points that feed data to a model that adjusts compensation values. This is still very tough to model, so normally it's only use to adjust spindle growth.

Then the car guys wash the parts at 150 degrees F. and expect 'em to be the same size!

Pneumo Precision in Keene NH made turning machines where the entire machine frame was constantly flooded with oil at a controlled temperature. The whole machine then lived in an insulated box. These things could single-point turn a telescope mirror and have it be within 1/2 wavelength of light! (Yes, you can machine glass, with very light cuts, extreme rigidity, and diamond tooling.)

The modern way to inspect these parts is using Coordinate Measuring Machines - no contact by human hands, controlled-temperature environments, measuring feedback from scales made of a glass-ceramic composite with zero expansion coefficient, massive real-time correction of static and dynamic structural alignment errors, etc. Typical "uncertainty" of measurement under the ISO 10360 test protocols (3D volume) is 2 microns (, .0001") per meter. I work for one of these companies, and we even measure the PART temperature and apply a correction model to it so the results are given for the part at a standard temperature, (usually 20 degrees C.) Ain't computers wonderful?

IF you can find it, get a copy of Wayne R. Moore's coffee-table book, "Foundations of Mechanical Accuracy". Required reading. Yes, Moore as in Moore Special Tool. No, I will never loan out my copy.

It never ceases to amaze me that Boeing still machines 100 foot wing aluminum wing spars in a non-thermally-controlled environment! But they do, and they work!

I used to joke about large machine tools being influenced by phases of the moon, but it turns out to be true! In my case a huge 5-axis K&T machining center installation next to San Francisco Bay. The final foundation design was a monstrosity, but totally necessary unless they were willing to move the machine to another plant.

[DieGuy]

IF you can find it, get a copy of Wayne R. Moore's coffee-table book, "Foundations of Mechanical Accuracy". Required reading. Yes, Moore as in Moore Special Tool. No, I will never loan out my copy.

Great book, wonderous reading. I worked in a dimensional metrology lab back in the 80's and it was required reading. We were just getting into CMM technology then and used Zeiss CMMs, (they had 27 of them). We alsd had some Moore and SIP measuring machines. I got to retrofit a Moore bridge measuring machine with HP laser displacement and geometry correction equipment. We were way ahead of our time on lots of real time correction. The software was all custom and written in HPL. It was great fun.

I would tell you where it was but then I would have to kill you.

Edit: just check amazon and a used copy in fair condition goes for $165.00

[gripper]

So my old steel 6" ruler is no good now we've gone digital....

[uscra112]

No, it can still be used to cut straight lines with an Exacto knife.

I am however the possessor of a wondrous collection of NON-electronic measuring tools, fishtail indicators, watches, even my old high-school slide-rule . . . Somehow all this electronic and computer stuff takes the fun out of it when I'm working at home.

[Geof]

.... even my old high-school slide-rule ....

I still have the one I bought for $36 in 1967. What is that equivalent to in todays dollars? I found a slide rule collector's website through Google and some of the prices quoted suggest mine could be my retirement fund in a few more years.

Perhaps we should not worry about trying to makes things the correct dimensions; just make them, measure them and then define a temperature at which they should be assembled so everything fits.

BTW scroll up to post #3 I got there first with phase of the moon.