Bearing race nomenclature

Topics include, Machine Tools & Tooling, Precision Measuring, Materials and their Properties, Electrical discussions related to machine tools, setups, fixtures and jigs and other general discussion related to amateur machining.

Moderators: GlennW, Harold_V

User avatar
BadDog
Posts: 5131
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Bearing race nomenclature

Post by BadDog »

Stephen:
That was basically my earlier outlook. But then as I pulled the trigger (buying material and budget HT oven), I began to have second thoughts.

Accuracy in terms we usually think about (spindles, etc) isn't even on the radar. BUT, this is a nominal 6T jack (probably built to handle 10T+ for safety factor, it's massive), and the caster uses ball bearings. Adding consideration that both OEM commercially made bearings show massive brinelling, and yet were hard enough that I have only the one that didn't fracture and fail completely, this I've begun to consider this may require more consideration.

There are 2 main points that caused my mental quandary.

First, the remaining intact bearing race shows a noticeably several thou dishing of the bottom. At first I thought maybe that's what caused the breaking, but then realized that the load vector from the balls for axial load are only about 0.140 in from the lip, so maybe not. Maybe the dish was to guarantee the lows was supported outboard to *reduce* the chance of breakage?

Second, with a less the (near) perfect ball race, ball bearing axial load will reduce to almost a "point" load on the race. Of course, everything, including ball bearings and races flex/deform, so it's not actually a point, but still very small area with very high PSI loading. AND I'm producing a race that is substantially less hard than a typical commercial race (hopefully at least avoid fracture), so I start thinking I need to put much more thought/effort into achieving as close to a perfect 0.250 race surface as possible.

That said, no, I hadn't considered a 1/4" end mill. Good suggestion, and it might just work. The OEM race had a slight OD "lip" that wouldn't be possible with the EM, but given the constraining upper race "cup", that might not be a problem. However, I'm not sure the finish would be good enough. But then, maybe the somewhat less hard (than commercial) race would take that in stride? And, is that upper/outer cup race sufficiently "tight" to keep the ball to the center so it avoids a point load? I just don't know enough to guess on that point...
Russ
Master Floor Sweeper
stephenc
Posts: 311
Joined: Sat Jul 05, 2014 6:13 pm
Location: youngstown ohio

Re: Bearing race nomenclature

Post by stephenc »

i observe human nature , ive noticed a personality flaw that most machinists have when it comes to personal projects , hobby or professional . its an inability to accept good results that might be less then perfect .
for example , you make a part for someone else , you do the very best job you can and when your finished you can take pride in a job well done and be quite happy.
but if you made that very same part for yourself you sit back and criticize every aspect of it , always thinking i could have done this that or the other thing a different better way . we also tend to do the very same thing before we even start a project which leads into second guessing and over analyzing every aspect of the project .

i guess what im trying to say is sit back , give it a good think and make sure you aren't trying to achieve unattainable goals and get on with it :)
User avatar
Harold_V
Posts: 20231
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Bearing race nomenclature

Post by Harold_V »

stephenc suggested the use of a ball end mill and rotab. Modern carbide endmills are CNC ground and should faithfully generate the desired profile. The only real problem with ANY ball end mill is that the center is stationary when being used, and often produces a rather unusual profile, or, at the least, a roughened surface or a distinct line. You might circumvent that issue by tilting the head of your mill so the center is not on center. Alternately, you could spin polish the generated groove in your lathe. That would work fine, assuming you generate a true radius, not one with a flat at the centerline.

Quite frankly, I think his suggestion has merit. The intended use of the race isn't critical, requires no particular level of precision, not in my mind. If you get close, it's going to work just fine.

I have not offered any advice in regards to temperature control. That's way beyond my capability. However, if you don't hold the precise temperature, you're still going to achieve results of some manner. I don't think I'd lose much sleep over the issue, although when heat treating to some government specs, target temperature of the furnace must be held to ± 1° F., a rather rigid requirement.

H
Wise people talk because they have something to say. Fools talk because they have to say something.
User avatar
BadDog
Posts: 5131
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Bearing race nomenclature

Post by BadDog »

Thanks all.

Stephen:
I definitely identify with your description. On some things I confidently make my choices, do my best, and move on. But on others, as with this race, there are points of thought that lead me to what I have heard called "analysis paralysis". It's more common for me than I would like to admit.

In this case, it was daunting enough to take on making a race from material I'm unfamiliar with including precision heat treat (as opposed to critical temp, then quench and temper) for the first time. But I was fine with that. The details that stymied me more than anything is observing the severe brinelling attributed to high point load damage and a multitude of fractures, seemingly speaking to excessive fragile hardness, leading to complete failure in a commercial race with likely a far more detailed manufacturing process than I could hope to achieve.

As for the EM, I'm glad you (Harold) called that back out. Yet again, I need to focus better and read for comprehension. What I took away was use a 1/4 EM side milling to form the 1/4" ball race surface, which would produce the 1/4" round profile precisely, but eliminate the (seemingly) desirable "lip" just past the ball midline to keep the ball from wanting to wander around (with point load rather than seated in the concave race surface). Now I see that the advice was for a ball mill, which both allows for the lip, AND presents a much easier tool presentation (long projection) without being crowded by the table/fixture, though at the "zero speed center" issue described to deal with. I rather like that idea.

However, in amidst family visits earlier today, I already started on the round HSS blank holder and bit. I think at this point I'm going to pursue that for a bit, at least far enough to finish it and test on a piece of cheap stock. If it looks promising, I'll go forward (likely next weekend I hope) with that and report the results (either way). If it proves problematic, I'll regroup with the ball mill (angled) and rotary table, and try again. Essentially, taking Stephen's advice to heart and get on with it...
Russ
Master Floor Sweeper
User avatar
NP317
Posts: 4557
Joined: Tue Jun 24, 2014 2:57 pm
Location: Northern Oregon, USA

Re: Bearing race nomenclature

Post by NP317 »

I like stephenc's suggestion of a rotary table and ball mill.
RussN
johnfreese
Posts: 219
Joined: Thu Jan 21, 2016 2:10 am

Re: Bearing race nomenclature

Post by johnfreese »

I can't remember the name of the company that made unground thrust bearings. The races had V grooves resulting in 4 point contact.
User avatar
BadDog
Posts: 5131
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Bearing race nomenclature

Post by BadDog »

Well, not much shop time with grandkids over, but I hoped to have some good results today. Sadly, I have utterly failed. Wasted 1/2" of expensive alloy, and 3/8" of 1/4" round HSS bit trying to machine this race. It actually started well, and I almost had one complete, when the edge of the round bit failed and it work hardened. Fixing the bit wouldn't touch it, and I had to use carbide to carve off 1/8" of material and try again. Two more bit refits later and another "clear out the work hardened material and I'm ready to throw in the towel, at least for tonight.

For anyone that cares, I used a piece of 3/4" square (from back end of brazed carbide bit I never use) and drilled a 1/4" hole canted 5* in both axis to provide 5* of clearance for both axial and cross feed. The top of the bit was carefuily dished out with a new sharp 1/4" carbide bur under magnification and with marking to clearly see as I got to a sharp edge. I think at least a large part of the problem is that this results in a rather steep positive rake. I had hoped to avoid work hardening and have easy in-feed due to that rake, which worked only for a short time on the first try. After that, it failed almost instantly. Note that the round is only used for finish form cutting the bearing surface, with the bulk of the material removed with a sharp (narrow land) modest small (don't recall) point radius CNMG insert, which cuts it just fine and easily with or without work hardening.

At this moment I think I need to just buzz off that 1/2" of foobar end of my my material and start over. I'm rethinking the whole 1/4" round for cutting this stuff. If I had the proper tools and skill to create a better edge geometry, I think it would work, but I can't come up with a good way to get what I want.

My second pass, which I may just test out on the ruined stub, will be to take one of those "I never use them anyway" brazed carbide bits and grind a 1/4" radius into it. Good excuse to use my diamond wheel carbide grinder I haven't had much use for in ages. With magnification, and maybe haul out the optical comparator, I think I can likely produce a more successful tool that way, though it will take some time and care. Testing it on the junk stub from today I can also verify seating of the ball using light bypass to see if it's judged close enough.

Failing that (a distinct possibility), I'll shelve that project until I can find a 1/4" round insert. I just don't think my skills are up to doing this with HSS when fighting work hardening stainless.

This is how it sets after the last try, 2 failures showing clearly on the bit.


Image of holder and bit

And yes, I overkilled both edges. I should have given up far sooner, but it was late, and I was getting pretty fed up with the progress on that front. Knowing I had to restart the whole thing, I'm embarrassed to say I allowed myself a small bit of sad satisfaction at the end...
Russ
Master Floor Sweeper
User avatar
GlennW
Posts: 7284
Joined: Sun Apr 15, 2007 9:23 am
Location: Florida

Re: Bearing race nomenclature

Post by GlennW »

Flood it with cutting oil.
Glenn

Operating machines is perfectly safe......until you forget how dangerous it really is!
User avatar
BadDog
Posts: 5131
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Bearing race nomenclature

Post by BadDog »

I don't have a setup to flood, but did apply a needle stream of light cutting oil with a lab washdown bottle at about 60 sfpm (110 rpm) with light feed. It actually made nice slice chips that curled nicely with a great finish... until it didn't. The edge just didn't hold up quite long enough to get it finished the first time, and that was the closest I managed to get.
Russ
Master Floor Sweeper
User avatar
BadDog
Posts: 5131
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Bearing race nomenclature

Post by BadDog »

I'm a little less frustrated with it now, and considering trying again with the HSS round.

Is it very likely that flood oil will make the difference? If so, I have a small portable pump/tank that I could setup with a rigged catch/return that might work well enough for the 2 pieces. My lathe has provision for flood, but it needs about 5 gallons (I think?), and I normally don't like the idea of using it just for the mess, so really not thrilled with going Full Monty.

What about that grind profile. I'm pretty sure the biggest problem is an edge that is too acute and easily damaged. With SS, the very moment that happens, it's all over without carbide. With the bit tipped 5*, it's got support as much as possible, but the "top rake" is pretty extreme using the ball bur in a die grinder. I tried on later attempts to keep it more shallow, but it took several attempts to get there without it jumping and having to start over, then, for all that work, it failed within seconds in the cut with work hardening having been cleared by carbide insert (same as used for rough shape that almost worked the first time).

Anyway, without some sort of dedicated tool grinder, anyone got ideas on how I can produce a more robust edge that works well? The only ball bur I have is 1/4" itself. If I had a 1/2" ball, I think it might work much better producing a much less acute edge with the bur just held centered (360* edge). Maybe I should just find/order one? Probably pretty pricey, but I'm kinda neck deep in excess investment on this thing anyway.
Russ
Master Floor Sweeper
stephenc
Posts: 311
Joined: Sat Jul 05, 2014 6:13 pm
Location: youngstown ohio

Re: Bearing race nomenclature

Post by stephenc »

My only helpful suggestion is going to cause you even more work ;)

Instead of grinding the tool by hand put your lathe to work . Make a tool holder for the die grinder or what ever you happen to be using to power the burr and mount that sucker on the tool post .
It'll at least give you a more precise way of grinding the tool .
User avatar
BadDog
Posts: 5131
Joined: Wed May 17, 2006 8:21 pm
Location: Phoenix, AZ

Re: Bearing race nomenclature

Post by BadDog »

Good point, thanks.
Russ
Master Floor Sweeper
Post Reply