View Full Version : Shrink fit....ARGGGHHHH

03-30-2005, 09:00 AM
Hey guys! Oh boy...I've been at it again. I've made two hubs for a customer. Both are for 1" bores and a 1" shaft (duh). I machined the hubs to .0015, or so I thought. I just measured them. They where a bit warm while machining. You could just barely feel it.
This morning the shop is cooler. Maybe 55 or 60F. The hubs now measure .003 undersize. Could this make that big of a difference or do I read measuring tools different depending on the time of day? http://bbs.homeshopmachinist.net//biggrin.gif
I have a lot of time in these and with a shrink fit you only get the one chance so I want to be dead sure here. Thanks!

03-30-2005, 09:12 AM
Yep,you need to wait till your parts are room temp before measuring for size.That's one big argument for flood coolant,it's not so much for cooling the cutter as it is the part being machined.

03-30-2005, 09:32 AM
What material? Aluminum rate of expansion is (about) 13 millionths of an inch per inch per degree F. Steel rate of expansion is (about) 6 millionths of an inch per inch per degree F.

So, if you have an aluminum part, 1" dia, 60 degrees difference, you have
0.000013" x 60 degrees x 1", or about 0.0008". Assuming the 60 degrees is correct, it doesn't seem like enough to cause the difference you see.

03-30-2005, 09:38 AM
Sorry...I forgot to add, all parts are mild steel. Crap...now I'm late for work http://bbs.homeshopmachinist.net//biggrin.gif

03-30-2005, 09:42 AM

The "standard temperature" for measurement is 20*C (68*F). Any temp above or below that, workpiece & micrometer both, will give an inaccurate reading.

Try this some time......calibrate your tenth-reading 1" micrometer with a standard that you've held in your hand for one minute. Let the standard cool to room temp & check reading. The hotter your hand is & the cooler your room temp is, the greater the variation in measurements.

BTW, here's a link to five class-of-fit charts. I printed these & frequently use them: http://www.engineersedge.com/general_tolerances.htm

Barry Milton

03-30-2005, 11:08 AM
Now I know why nothing I make ever fits!! All these years I thought it was my fault!!

Forrest Addy
03-30-2005, 01:15 PM
Yeah, temperature is the most frequently overlooked variable in the machine shop. If the work and the measuring tools are steel and the same temperature it doesn't metter how hold or cold they are within reasonable limits. As soon as there's a difference you better get out your thermometer, your scratch pad, your calculator and your table of coefficients of thermal expansion.

We used to have a 42" Bullard Cutmaster that ran hot maybe 30 degrees above ambient. Anything you put on it ran undesized after it cooled. I used to make a of of aluminum bronze submaring fittings on it because other than running hot the machine was in very good condition. Having learned the lessons my mentor taught me I compensated for temperature errors and made parts that once cooled to room temperature were right in tolerence.

Crusty old farts use to abuse me "You don't need to do all that figgerin' - waste of time. Just cut it to size." "Ever get a good part off this machine?" "Nope, never did." "I got some there on the pallet. When they cool they'll be right on the button."

Naturally that got me a reputation for being a smart a$$. Time hasn't changed me a bit.

03-30-2005, 02:52 PM
Why don't one of you guys you make this the subject of a short article for HSM or MW?

03-31-2005, 09:02 AM
Thanks guys! I don't know what I did wrong. Geez, I measured these 10 times and was happy that it looked right. Measured them again last night and they really are too small. Never had time to fix the problem but should be ok tonight.
Ha....took the night off to go see some miniature live steam engines. WOW! I thought it'd be ho hum. Was I wrong! The machine work is incredible.