Ambient Temperature Affect on Manufacturing

You'll need to determine the expansion coefficient of the material(s) you are using.

Look in the "Machinery's Handbook" for expansion coefficient values and the formula required to compute them.

Thermal Expansion Rate for Steel 12 x 10^-6/Degree Celsius.

Therefore 20 Degrees * 12 x 10^-6 * 50 mm = .012 mm

This is not the entire story but shows that temperature can be a factor. (Calculations were for straight line expansion as a simplification, not for a cylinder.)

I'm no mechanical engineer and I didn't study how to calculate expansion due to temperature.

Just off the top of my head, Shop B, which is fabricating the bore, is warmer (hotter) than Shop A, which is fabricating the shaft. Assuming that temperature expansion also affects the calipers Shop B is using, the size of the bore would be actually larger than specified. The shaft will easily mate with the bore. I don't know how the looser fit will affect its performance.

The industry standard is as you stated: 68deg. F/20deg C. Anyone making parts for you is responsible for making them to print at that temperature. Incorrect ambient temperature is not an excuse for not meeting print tolerences.

As another person has already commented you can calculate the thermal expansion and predict the results. However, take these things into consideration. When measuring steel parts, you are probably measuring with steel or SST instruments which are affected equally with the steel parts being measured. Also the heat generated in the machining process is a factor in addition to the ambient temperature.

Aluminum has a much greater coefficient so you must be much more careful. Any more than 5deg. F trying to hold +/-.001 and you could have problems.

I can give you a couple of examples. I recently visited Hardinge Mfg. in Elmira Ny. In their spindle repair and assembly rooms the parts are kept for at least 24hrs to normalize temperature before they are worked on.

In my own shop I recently machined some polypropylene parts that were 9 inches long. The tolerences was only +/-.005 but I found that the part moved over .001 for every degree F.

The variation i have calculated in excel (simple really). It is not much significant : provided the thermal expansion of job(s) and micrometer is almost same, and the job(s) and micrometer is at same temperature. The pl note that the room temperature is not usually the job temperature. It may be hot due to machining heat, or cold being kept on a cold table.

If measured values are 49.98 and 50.02 , at room& temps 10 and 30, micrometer temp 10& 30 - coeff exp job(11) and micrometer(8), the fit actual will be 0.0325 instead of measured 0.0295, but if mic temp in both cases are 30deg c the actual fit will be 0.405 (error of 0.011).

In practice i have seen this happening very often esp since my components are > 300mm dia and in some cases reach > 1m dia. In these cases I try to compensate the measuring instrument by maintaining same job and mic temperature. It works within acceptable error limits ,(both are steel and have similar alpha). pl write if you have difficulty in creating the excel file.

Problem is not ambient temp. Material get hot when you machine them. Parts are usually check when setup on the machine. So part would be hot when measure. If your part is really critical, you'll need to inspect them in temp controlled room with the parts have enough time to settle to same temp. It also means inspector can't hold them in his/her hand for too long.

For 6061 coeff is 13uin/in-F. With 36F different its 0.000468in/in. So your 50mm shift will have +0.0234mm. Stainless are around 9uin/in-F.

I don't think temp different will be that much when the shop machine the parts. You'll see the different after shipping depends on outside temp.

Don't inspect critical parts until they get to room temp.

Pineapple

I think I did not make myself clear - the problem is the job vs micrometer temperature and their coeff of expansions (if different).

At my place min temp is about 10 deg C. At this temp, even if jobs are kept for 1 day, at noon the job temp may be about 15 deg (not lost the night cool), ambient is about 25. In addition when one measures several pieces , micrometer get heated due to inspector body temp (37 deg C) - this variation I have seen and compensated. I normally keep the mic on the job/ measuring table in physical contact so that the heat balance is by conduction (more efficient and less time to heat balance). Also the job should not be kept where there is wide temp fluctuation - near fan, door, in open etc.