CNC Your Own Parts

[volvolugnut]

Here goes.
Bearing manufacturers make common size bearings. There match inside diameter, outside diameter, and width of other manufacturers so they can interchange. However, they may not have the same loading and life ratings because some manufacturers make better bearings from better materials. So if you are replacing a rolling bearing in something, you can find replacement by the size unless you are critical to get original equipment ratings.
In a previous work life, I specified spherical roller and cylindrical roller bearings for use in pumps. The largest of these bearings had OD about 12 inches, ID about 7 inches, and width about 3.5 inches. They cost several hundred dollars each. I preferred to use SKF brand, but sometimes their delivery was several months and we could not wait.
These bearings had lots of company supplied design information including tolerances for machining the mating parts. As I recall, the tolerance to machine the pump case for the largest bearing was something like 0.0003 inches. All machined bores were specified to 10 thousands of an inch. The machine shop we used, made the parts with large CNC machines and did not typically need grinding to hold these tolerances. But they would sometimes reject parts made out of tolerance.
Actually, if the machine shop made parts out of tolerance, I would get a call asking if I could accept the parts. Most of the time I would have to reject the parts because the results would harm the life of our pumps.
volvolugnut

[MrAl]

Here goes.
Bearing manufacturers make common size bearings. There match inside diameter, outside diameter, and width of other manufacturers so they can interchange. However, they may not have the same loading and life ratings because some manufacturers make better bearings from better materials. So if you are replacing a rolling bearing in something, you can find replacement by the size unless you are critical to get original equipment ratings.
In a previous work life, I specified spherical roller and cylindrical roller bearings for use in pumps. The largest of these bearings had OD about 12 inches, ID about 7 inches, and width about 3.5 inches. They cost several hundred dollars each. I preferred to use SKF brand, but sometimes their delivery was several months and we could not wait.
These bearings had lots of company supplied design information including tolerances for machining the mating parts. As I recall, the tolerance to machine the pump case for the largest bearing was something like 0.0003 inches. All machined bores were specified to 10 thousands of an inch. The machine shop we used, made the parts with large CNC machines and did not typically need grinding to hold these tolerances. But they would sometimes reject parts made out of tolerance.
Actually, if the machine shop made parts out of tolerance, I would get a call asking if I could accept the parts. Most of the time I would have to reject the parts because the results would harm the life of our pumps.
volvolugnut

Hi,

Some more interesting stuff.

OD about 12 inches, isn't that kind of small for a bearing?

Just one thing...
You said the tolerance was 0.0003 inches, and that is 0.3 thousands, but the bores are specified to 10 thousands which is 0.010, so how is that so? Just wondering.

0.0003 inches is pretty darn good. One of my brother in laws worked in a shop where they made machines that were used to make other machines that would make parts like this. I think he said the tolerance was 0.1 thousands or maybe it was 0.01 thousands, I just remember being very impressed when I heard that so it was probably 0.01 thousands. I'll have to ask him again.

[volvolugnut]

I misstated the tolerance. They were specified to tenth of a thousandth of inch - 0.0001.
There is an old story about Toyota after they started getting very serious about quality. The engineers asked what to do after they could no longer measure the difference between good parts and bad parts. The answer was to get better measurement tools.
volvolugnut

[MrAl]

I misstated the tolerance. They were specified to tenth of a thousandth of inch - 0.0001.
There is an old story about Toyota after they started getting very serious about quality. The engineers asked what to do after they could no longer measure the difference between good parts and bad parts. The answer was to get better measurement tools.
volvolugnut

Hi,

That's interesting that you should mention Toyota.
From what I understand, Toyota was founded by the Toyoda family (funny huh).
They changed the name to Toyota because it only takes 8 strokes, considered a lucky number in their culture, while Toyoda takes 10.
I thought it was strange when i first heard the name Toyoda in charge of Toyota.
I first thought it might have something to do with the way they pronounce certain English letters, like 'R' for 'L', as in "Brack Friday" instead of "Black Friday".

Ok then the tolerance my brother in law quoted must have been 0.00001 inch because I knew about the 0.0001 tolerance for a long time and i was very very surprised and amazed to here that smaller tolerance. A lot of micrometers go down to 0.0001 and have 0.0001 increments.

[volvolugnut]

If you have or see old (before 1980?) micrometers, they likely have 0.001 increments. That is what my old, used micrometers have. Good enough for most automotive work.
volvolugnut

[BlackBart]

12" x 3.5", yow! What kind of grease do things like that use?


I just read a magazine article on the first Ferrari V12 and I thought of this thread on tolerances. 1947.

Like many engines over the years, it was a cast aluminum block with steel cylinder sleeves. They heated the block up and pushed the cooled sleeves in. When it all came to temp, the block has grabbed on to the sleeves not to let them go. I'd think you'd have to do some experimenting with what those dimensions are and whether it would crack the block or produce a loose cylinder. Track testing at high temp, etc.

[volvolugnut]

These big bearings were lubricated with 50 PSI synthetic oil. The oil flow also cooled the bearings. I can't remember how oil was introduced to the crank bearings. I don't think the straight, splined crank shaft was cross drilled for oil holes. We had enough problems with breaking pump shafts without stress riser holes.
The straight crank shaft had eccentric assemblies which had the connecting rod big end bearings (the large bearing sized above) and slightly smaller bearings on each side for the crank shaft bearings. The eccentric was splined to the shaft to drive the pump connecting rods and pistons.
volvolugnut

[volvolugnut]

Expansion and strength of cast iron and aluminum is fairly simple to calculate if you know the expected temperatures. If you are designing race engines you should know what temperatures to expect.
volvolugnut

[MrAl]

Expansion and strength of cast iron and aluminum is fairly simple to calculate if you know the expected temperatures. If you are designing race engines you should know what temperatures to expect.
volvolugnut

Hi,

If you remember back in early science class, there was a device with a ball on one end of a handle and a ring on the end of another handle.
You could not get the ball through the ring because the ring was slightly smaller than the ball diameter.
If you heated the ring it would expand just enough so you could get the ball through.
If it cooled, you could not get the ball out of the ring so the two handle things were stuck together.
Heating the ring a second time, you could get the ball through again and thus separate the two again.
They are both made of the same metal I think it was brass.

And with that, here's a question.
If you start with them where the ball is through the ring already, and normally you would heat up the ring to get the ball through again, what would happen if you heated them BOTH up at the same time. That is, heat the ball and the ring at the same time.
Could you then still get the ball out or would they expand the same amount and so you couldn't get the ball through?

The key points I think are that the ball volume expands out in all directions, while the ring expands mostly along the circumference. Does that make a difference though.

[abscate]

Simpler version. If you heat a square piece of brass with a circular hole in it, does the hole diameter grow or shrink?

Show your work.

Expansively of brass is 54E-6/C