[mactec54]

hy mactec, technology for crafting the ball is not unique for the entire size range; frequent sizes are smaller, so there is justified a punch-forming machine

also, what i have described is not a full process; what about cold/hot forging the ball + hard turning + grinding & sorting ( 2-3 baths ) + lapping ( 1-2 baths )

maybe you should try a ball bearing ... kindly

I know what Ball Bearings are I used to build custom protype Bearings for SKF , they are never hard turned as you say

[deadlykitten]

this is a fact, from a plant : technology for crafting balls involved forging from dia~25, and hot forging from dia~60 to over100

forging operations are followed by turning in 1 or 2 steps : removing burs + kind of prefenish, before grinding operations

okey, now what ?

[mactec54]

this is a fact, from a plant : technology for crafting balls involved forging from dia~25, and hot forging from dia~60 to over100

forging operations are followed by turning in 1 or 2 steps : removing burs + kind of prefenish, before grinding operations

okey, now what ?

Please show us this primitive plant doing this turning operation, it does not matter what size ball is need they are either hot forged or cold headed then Ground from that point on

https://www.precisionballs.com/Giant_Balls.php

[deadlykitten]

hy jarith

We’re having trouble holding the sphere for the 2nd operation

what kind of trouble ? what is your fixture for the 2nd operation ?

i want to turn it to a tolerance of around 2 thousands. But I’d be happy with 5 thou all the way around or even more.

2-5um ? hmm ...

However that isn’t my main aim at the moment, holding the work for the 2nd operation is my main concern

no, no, you should not see them different ... i mean, you have to "link" the "desired tolerance", with the "stability" of the fixture; if you wish to achieve a rough tolerance, machine setup will be fast, but, for a more precise tolerance, you need to invest more time in preparation

so, what quantity do you need, and what tolerance is allowed ? or is just an experiment, see what happens, etc ?

ok, now, i will share a few things :

1) in the 1st operation, you may encounter a not-so-nice finish towards X0, because the machine has to increase the rpm too high, beyond it's limits, so, in the middle, there will be a spherical cap, with rough surface roughness; does this bother you ? do you wish to minimize this effect ? if yes, then pls check footnote *1 below

2) X axis backlash may be within 2-5um, so you may not need to worry about it now; if axis backlash is > 5um ( or whatever value you consider ), then you need to discuss with your machine dealer, about the posibility of managing axis backlash by parameters / code; if it is not possible to manage the backlash, then you will have to map it "longhand", thus by g-code, and hoping that it will work

3) for the 2nd operation, if you use soft jaws, and still wanna achieve 2-5um tolerance, pls look at attached image :
... left : yellow represents the machined surface in 1st operation
... right : yellow represents the machined surface in 2nd operation

after clamping the sphere ( 2nd operation ), and before cutting, check the fixture tir : is it <2-5um( or whatever value you consider ) ?
... if yes, then you can continue, and check tir periodically ( like once at 4 hours, etc )
... if no, then you can not continue, since fixture stability is not enough; i can share infos about clamping with low tir, like few um, but this requires more time seting-up the machine; even if you use soft jaws, with low hidra pressure, your chuck has to be in good condition, and the machine size has to be small, because, hidraulics, even at low pressure, may generate a force big enough to mess up a tir, way beyond 2-5um; again, pls say if you need a solution to improve tir for 2nd operation; check a brief info in footnote *2

4) 2nd operation will have to manage the "meeting point", thus, somehow, you will need to tangent the surface from 2nd operation, with the surface from 1st operation; check the details in right image : this requires a hard material, and a sharp tool, so to minimize the difference; if material is soft, and/or tool is not sharp, the difference will be more visible; i don't know, in the end, polish it a bit ?

kindly

---

Please show us this primitive plant doing this turning operation

hy mactec, if that thing is working for them, then is ok

in that image that you shared, it seems that there is the "deflashing" operation, that removes the "big bur"; maybe that it can handle more than 1 sphere at a time, maybe it can perform faster than turning, maybe those cast iron plates will last for a looong time, i don't know ...

in that link, it seems to be an automated solution, that does " grinding / lapping / polishing " ... but look what is there " We usually machine the blanks ... on spherical turning machines, but cast ingots or formed and welded hemispheres can be the raw material. " - it seems that that custom-cnc can do rough operations, but is preffered to do the roughing before, on another machine

so, i guess that is not so "primitive" to have a machine only for rough turning ?

please, do you have a video with this machine ? kindly

---

*1

you need to map your rpm range across the sphere, see what happens, in order to minimize the " dirty " cap surface; in other words, you have to keep the cutting speed constant for as long as possible :
... try a lower V ( M/min ). so to delay the moment when the machine reaches it's spindle limit, and, as a result, the cap surface will decrease; there is a balance, because :
...... if V is decreased, overall surface quality may get lower, because finishing requires speed, but, the cap surface may be barely noticeable
...... if V is increased, overall surface quality may improve, but the cap surface may increase and look nasty; this balance has to be found by running trials, and it depends on material and tool
... instead of a turning knife, you may use a round-end-mill, so to stabilize even more the cutting speed : as a result, you no longer deal with the "cap" problem, since M axis can keep it's rpm constant regardless of X, but, for this to work, you may need to keep the spindle runing in S mode, not in C mode, thus you may need a machine interlock; if you rotate the spindle with C axis, motion will be discontinued, since you can not achieve constant rpm in C mode, so, you need to run continuosly, in S mode

*2

somekind of hollow semispheres, one being machined in the spindle, and another being inside the tailstock; this setup removes the "noise" caused by the hidraulic pressure inside the chuck ... however, it requires a smooth tailstock; if tailstock pressure is too big, than "noise" re-appears; in such a case, you may tighten the tailstock mechanically, using a screw + nut system

this "noise" represents the position shift range, caused by the hidraulics; if you remove/lower the hidraulics, you will remove/lower this "noise"; does this makes sense ?

[skrubol]

hy jarith
2-5um ? hmm ...

Thou normally refers to thousandths of an inch.

[deadlykitten]

hi skrubol, i don't know, in post 19 he sais that dia should be 30mm ...

[skrubol]

hi skrubol, i don't know, in post 19 he sais that dia should be 30mm ...

In some areas is it common to use 'thou' to describe thousandths of a mm? I've never heard it used that way.