What's new
What's new

In-machine grinding and probing

TAIWA NUMBA WAAN

Aluminum
Joined
Mar 9, 2021
I've recently gotten some strange requests from customers. They have parts being made in fairly large qts, 30,000 to 60,000 annually, that have difficult features to machine. I can't post the drawings here but, I'll try to describe the parts:

50-54 Rc steel,
30mm to 67mm diameter
Length about 1.5x diameter
Has a big through hole off center by a few mm, so the wall thickness varies.
The thicker walled side has more axial through holes from 3mm to 7mm (high length/dia ratio)
There are a lot of slitting saw crescent cuts on both the inside and outside.
Worst of all, the tolerances on the OD, off-center ID, are 0.015mm! This requires grinding since it's full hard.

The customer has/had them made in China with 60 to 100 percent reject rate - passed Chinese QC and the customer paid for them all, but scrapped when inspected independently after delivery to the western company..... Yes, I know this is as dumb as it sounds, keep in mind I'm not the one doing this.

So the customer now wants to bring the manufacturing in-house in a mill turn setup with a rough op, heat treat, then finish grind with in process probing as preliminary QC.

I am skeptical of putting a production grinding process in a machine that also does regular turning and milling, but I'm kind of obligated to help this particular customer, so if anyone has any advice please let me know!
 
customer, so if anyone has any advice please let me know!

Marposs may have the most experience in the world at this. In-process high-production grinding measurement is their specialty.

But I could be out of date, companies these days change who and what they are at the drop of a hat. So, at least used to be true, still might be. Anyway, it's a startpoint.
 
So you would want .0015 mm (50 millionths) size control and adjustment.
That is expensive territory.
Rather than grind you may want to look at CBN cutting tools.
I worked at place that had to do this in this hardness. (CV joints)
Previously they had "matched" color sizes for assembly. This was a huge bottleneck and the conversion to one size painful.
In another place it was decided to only make one size hole/piston back in the early 80s. We got the hole boring well down under this.
Then the piston making department could not hold the tolerances for only one size and that became another journey.

You do not need super accuracy on the gauging but you do need super repeatability. (one can "tweak" zero but it has to repeat)
Marposs indeed great, Reinshaw may have help.
Bob
 
Sounds like a job for a HEMBRUG multitask. Probably going to be a 7 Figure check.
 
Bob... one too many zero's past the decimal point.
Process control and adjust needs to be 1/10 of your print. SOP in my world. CP, CPK. PPK and all that weird stuff.
Inspection will demand this since they only want to check a small sample and then trust the rest of the batch.
Making one part and 60,000 have different demands and rules.
Imagine that 6 axles out of 280,000 fail. Now Uncle Sam mandates a recall of all those cars.
Not only do all have to be replaced but you have to provide loaner cars to all those customers during the fix it process.
And.. the big automaker will put a crew of people in your plant 24/7 to help "fix" you. Big fun for a part that goes out the door at $60 each complete assembly.
 
Last edited:
Process control and adjust needs to be 1/10 of your print. SOP in my world. CP, CPK. PPK and all that weird stuff.
Inspection will demand this since they only want to check a small sample and then trust the rest of the batch.
Making one part and 60,000 have different demands and rules.
Kind of out of my wheelhouse but I remember seeing a add at one point or another for the hemburg i mentioned above, hard turning & grinding all in one, repeatability was like .0002mm with in process inspection.
 
You old farts and your 1:10 stuff ! I bet you don't even believe in inspection on the same machine it was made on !

Troglodyte ! :D
I know it's not a legitimate metrology practice, but probing the OD and ID to check diameter is more of a go/nogo in this case, just because some previous batches have been defective. They would still have to do a full inspection later but the purpose of probing in the same machine is to catch any process faults before they scrap the entire run.

0.015mm is 6 tenths, so tight but not impossible.
 
Some multitask machines advertise grinding capability, but it's generally supplemental due to the small size of the wheel. It boils down to wheel wear. Small wheels wear quickly.

Given the quantities, this is a job for dedicated abrasive machines, i.e. OD grinding and honing. If location tolerance is tight, semi-finish machining prior to abrasive finishing might be necessary.
 
Some had mentioned CBN, that's fine and dandy, but hard to turn when the center bore is offset. I know Hermle and Makino do grinding on their machines, they even have special grinding cycles included in the software for dressing wheels and such. Have a look at the Makino online webinars, they have an entire topic with on machine grinding and checking that is pretty useful.
 








 
Back
Top