Does anybody care to educate me on centreless grinding?

[mc]

I'm toying with a new product, but my problem is I need ground shaft stock (0.498 - 0.499" to be precise), as polymer bushings need to run on it (hence the ground surface to meet the recommended surface finish requirement).

Now I'm aware such things are typically finished with centreless grinders, but that's where my knowledge on the subject ends, so I'd be grateful for anybody who could at least increase my knowledge a bit.

How much material would you typically leave to be ground?

Links to more in depth information about them would be welcome. I have done a couple google searches, but other than the basics of how they operate, I've not really found that much information.

And how mad would I be to consider building one?
A quick search suggests they're typically into double kw figures for power, which is well beyond what I currently have available, so I'm thinking something maybe 3-4 foot square foot print, with a 2-3 hp motor. I'm not looking for a high throughput, and if I could combine it with an automatic parts loader, I could dump a pile of parts in it, and ignore it while I do something else.

[seadog]

Can't help, other than to point you at this company - https://www.westmidlandgrinding.co.uk/services/centreless-grinding/

[Lew_Merrick_PE]

Hi mc  --  You are (I hope) aware that 3 points can define a circle.  Think of centerless grinding as being 3 tangent points where 2 of those points are high friction drive wheels and the 3rd point is the grinding wheel.  Problems crop up because high friction drive wheels are usually flexible and grinding wheels tend to wear away.  This makes holding a diameter closer than (say) .0005 inch somewhat problematical.  Back in the dark ages when I did a fair amount of centerless grinding most parts had less than .010 diametral inch of clean up.  [1988 was the last time I ran a centerless grinder.]

[sbwhart]

Done a bit long time ago:-

As said in Lews' post you have to watch out for lobeing due to the three point contact. You can get small fixtures that fit onto a service grinder that will change it to a centreless. the main benefit of centreless is that you can do long bars by simply feeding it through the machine we've used bars 3m long that were ground this way, also you can simply pass multiple part through one after the other. You'd leave about three thou on for finishing. If your part is relatively short say 4/5" I'd just have it ground between centres, you could do this with a tool post grinder or you lathe.

Stew

[mc]

Thanks for the input guys.

The finished parts would vary between around 1 to 3", and some kind of between centre grinding setup would work for the parts with a centre hole, but not all parts would have a central hole, hence why I was looking at centreless grinding.

I spent a bit of time today looking at items that are already available, and it would appear nobody else is grinding their parts, as all the parts I've seen so far are just turned finish, so a roller box is now looking like a more cost effective option, at least for now.

[PekkaNF]

Maybe you are familiar with this?

Because grindingwhees can be hands full, mabybe belt would a way to go...something like this:

Real machines tend to be powerhungry behemonts, here is the tiny one:

Pekka

[mc]

Thanks for those vids Pekka.
That surface grinding one had popped up on my feed, but I had not watched it until now. It certainly gives me one potential option if I go down this line.

The belt one, I'm going to guess is aimed more at getting a brushed/clean finish rather than good tolerance.

And yes, industrial centreless grinders all seem to be big, but then they're all aimed at big throughput with minimal adjustment. If I was to go for grinding, I'd hope to have something reasonably compact, but I'm well aware the smaller the wheel, the more adjustment it'll need to keep tolerances.

[PekkaNF]

Happy to nudge a little, right direction, I hope, from the OP I suspected that you pretty much know what need to accieve:

I'm toying with a new product, but my problem is I need ground shaft stock (0.498 - 0.499" to be precise), as polymer bushings need to run on it (hence the ground surface to meet the recommended surface finish requirement).
...

If you need relatively short and are willing to do manual repetition, then Robin method is superior.

But belt grinding can be dialed in to be pretty accurate when corret contact wheel and belt is used. And obiviously has greater troughput capacity on constat size shaft. My limited knowledge on polymer bushing is that they are more forgiving on actual diameter than surface roughness, therefore I put this belt option here.

Being DIY forum that beltgrinder option seems to show pretty good DIY potential compared to proper centerless grinder.

[WeldingRod]

Are you worried about surface finish or tolerance?  Roller or diamond burnishing can take a lathe finish to what you need in one step.  AND the finish will be the better-for-sliding biased type: peaks selectively knocked down.
Single point diamond burnishing tools aren't super expensive, but are also not super-impressive (at least when I tested them)  Completely flexible on diameter, though!  I was shooting for biased and around 2 microinch, so I was asking a LOT.  They need clean coolant/lubricant.  We used a whole house filter with a micron element.
Roller tools are not very diameter flexible, but super fast, smooth finishing, and controlled.
I still have the diamond tool somewhere, so I might be able to do a demo.

Sent from my SAMSUNG-SM-G891A using Tapatalk

[mc]

A bit of both.
I'd like to maintain a nice constant tolerance, and I realise it's not that tight a tolerance, but I think I'd struggle to maintain it on my lathe using normal turning. The shorter parts would probably be OK, but trying to hold less than a thou (I'd ideally like to aim for maintaining within 2-3 tenths) on a bit bar at 6 times longer than diameter is likely to be a challenge.

The finish is where things get a bit interesting. IGUS provide lots of information and wear rate charts, for their various compounds. The generic option specifies 0.6Ra (N5-6 or rougher end of ground finishes) finish as being optimum (too smooth and stiction/material transfer becomes a major problem), yet their low friction compound doesn't show any greatly increased wear even up to twice that roughness (in the N6-7 area which is the overlap zone between typical ground and turned finishes). As this will be a oscillating application, they also mention the high spots will wear of the shaft, so a rougher shaft can actually give longer service as it'll wear in. Off course I need to tread the line between wearing in to give good performance, and wearing out so things start rattling!


At the moment, I've found a supplier of reconditioned roller boxes, and although it probably won't work in my CNC lathe (it would stop my turret from fully rotating which could be worked around, however I think it'll hit the enclosure when trying to part off), I'll give it a run on my manual lathe to see how things will work out.
If that works, I think this idea will have enough potential profit to justify a new CNC lathe build designed around gang tooling.