Source for small precision shafting

Great page James. Keep us posted. I would never think of fans running at 43,500 RPM! A 40mm diameter (20mm radius) fan, spinning at 43.5 rpm has a linear speed at perimeter of 43,500 * 0.02, or 870 meters per second, or approximately Mach 2.5 speed. I wonder what kind of effects would a supersonically rotating fan have on air. It is not obvious to me. Supersonic bodies generate shock waves, and I also wonder what effect would those shock waves have on parts near them. Please let us know how it works in the end and what is the sound level near that fan.

i

Scratch that, rpm is per minute. :)

i

Uh, i, I know you're the algebra guy, but you should check those numbers.

43,500 rpm is 725 revolutions/second. The peripheral travel of a 40mm fan (pi*diameter) is 125.66 mm/revolution. Travel*revolutions/second is 91.1 meters/second. (Roughly 200 mph.)

-- Ed Huntress

Ed, sorry, Ialready realized my mistake, see another post of mine. You are completely right!

i

I should have realized you'd catch that in a hurry, and saved myself the trouble. d8-)

-- Ed Huntress

Sorry, he's much more your speed than mine.

H

So Harold didn't post numbers. That's no reason to insult him. However, because you seem to know it all, (almost), here are some numbers so your knowledge will be complete. How about round to less than .00005, to size within .0001, on parts about 1.375 O.D., 1.500 long, with a 1.125 I.D.? Dimensioned in inches. Parts were nitrided to about 65 Rc. These were done in quantities, cheaply. These parts were first ground by someone who didn't know how to set up the grinder and were out of round (triangular) almost .0015. Since the tolerance was .0002 total all the parts were scrap. The next shop did 'em right. Maybe the person who told you that centerless grinders couldn't make round parts was the same one who ground and scrapped the first run of the above parts. ERS

Not if the material of choice is stainless, it wouldn't be reasonable. Stainless is difficult to grind under flood coolant conditions--and virtually impossible under dry conditions. The moment it starts to heat it expands into the wheel and snowballs. Been there, done that. I'm of the opinion he'd have lots better luck simply turning the end, starting with shaft size that fits the bearings properly. That's how I'd have approached the job at the outset.

We're being spoon fed bits of information such that every time a comment is made, more information is disclosed, changing the concept, making helping almost impossible. I'm at a loss why folks that ask for assistance can't disclose what they're trying to do, and how they'd like to do it. So far, the main effort has been on fitting the end bell to the shaft, totally disregarding the fact that bearings are involved. That's the wrong approach to the job. At those velocities, the fit of the bearings is critical, it can't be ignored, not in my opinion.

Harold

You have only a small lathe, and one

Added video with audio to web page.

James wrote:

I don't understand this.

3.18 mm = 0.125(2) inch

3.08 mm = 0.121(3) inch

If your lathe cannot turn down an

1/8 inch diameter shaft by four thousanths then I respectfully suggest what you need is a new lathe, not all the usenet help that has been supplied!

Jim

How come nobody has suggested passing a one-under 1/8 inch reamer through this end bell? Seems like that's a good deal easier then machining the shaft with a lathe that won't cut to within a thousanth.

Jim

Who was insulting?

I hope I'm misreading you as you misread me. It looks like you're insulting me.

Thank you. This is precisely what I was looking for.

How were the parts measured?

That could be possible. I believe the reference was Cliff. This is why I suggested a discussion between Cliff and Harold. While it is unlikely that such a discussion would take place, there's a certain chance the comment would trigger a further discussion (which it has).

Regards,

Robin

Thanks, Eric. That smart assed punk kid needs to be taught some respect. I've spent more hours running a centerless than that smart mouthed little bastard has been in the shop. It's plain damned amazing how someone can work in the trade for a couple years and get so knowledgeable that they know all about things they've never touched.

I failed to disclose any numbers because that serves no purpose. A centerless grinder is capable of working to millionths, but not in the hands of a fool, as you discovered. I ran countless production runs of relatively small quantities (a few hundred parts) that were held to .00050" with no effort. Roundness was never a problem, and size was only because the machines take several hours to warm up, so you're constantly backing off the wheel until it settles down. A job was often finished before the machine was up to full operating temperature.

By the way, Cincinnati talks about the tri-lobed condition in their operator's manuals. It's real easy to avoid.

Your job would have been a serious challenge due to wall thickness, but light passes, 5/16" above center, with plenty of coolant, with all the grinding coming at the front of the wheel, allowing the balance of the wheel to spark the part, and they would come out beautifully.

Harold

But doesn't this sound like a Swiss screw-machine job to you? That's what the things were made for.

The velocities are high but no motor I've ever seen has an armature so perfectly balanced that the shaft itself has to be machined to leadframe-die tolerances. If balance is that critical, you have to balance the assembled armature.

Eh?

-- Ed Huntress

I apologize that this post has turned into name calling and insulting. I apologize for not providing enough information to get the help I needed.

The lathe I'm using is the Taig micro lathe. The armature is stationary and the rotor turns. The rotor is pressed steel and is very accurate. I install permanent magnets in the rotor using a jig I made and they are in balance. Witness the video on the web page where it's turning almost 40K rpm with no vibration.

One company said they can supply the rod I need for $19.95 a meter so I ordered 3 meters.

Thanks everyone for the great ideas and sources.

James

James wrote:

That *should* be able to turn the shaft to teh tolerances you need.

Also consider passing a one-under reamer through the end bell to size it to the shaft, rather than the sizing the shaft to fit the existing hole in the end bell.

This will render the hole nearly 0.124 which will get you close to a good press fit. You can purchase an *adjustable* reamer to size the hole closer than that.

Jim

Harold,

I don't appologise for writting what I wrote. You and Eric have misinterpreted me, thinking I was insulting you and suggesting that you do not know what you're talking about.

I have never run a centerless grinder. I have (briefly) run a cylindrical grinder. I cannot comment, from experience, as to the performance of such a machine. As such (and because I'm in the field) I was interested in the type of accuracy obtainable using a cylindrical grinder.

I believe it does serve a purpose. You yourself stated that the fit between a shaft and a bearing running at greater than 50,000RPM becomes very important. Knowing what a centerless grinder is capable of is important - esspecially when it is obvious that the original poster doesn't have a very deep background in metalworking.

These are numbers, and that's important. Esspecially when you state "millionths". I am interested, however, in further quantification. Are we talking 0-10 millionths? 10-100?

Because I work in the automotive industry, I have a certain interest in this information (although my job does not deal with engine/transmission components at all). Most in this group are not particularly interested in the last 4 and 5 decimal places. There is a gentleman on AMC who deals with these numbers as he deals with engine components, and I find his work very interesting.

Lastly, if you have a problem with me, it would be a courtesy to address me directly (publically or privately). To slander a member of a public forum in that forum is rude and dishonourable.

Regards,

Robin

"Ed Huntress" wrote in news:4ZRSd.7817$ snipped-for-privacy@fe11.lga:

After looking at the pics, I'd have to say I could make the shaft and the motor mount complete in one op. on a Swiss in my sleep. Give me good c'less ground stock and I'll hold .0002" total on all diameters, .0001" is possible. If you're going to be making these in any knid of volume, another option would be to use a box tool, followed up by a roller burnisher (Cogsdill type). Even the most clapped out lathe can hold tight tolerance and finish that way. The burnishing tool has .0001" adjustment. You could pick up an old Brown and Sharpe 00 hand screw ( I sold mine for $500.00) or a Hardinge DSM hand screw and knock them out to +/- .0002 and a 16 Ra all day long. I agree with the bit about machining the shaft extension after assembly. I've seen it done in high volume production in Japan. You could also two op. the motor mount and hold similar or better tolerance in the bore.

Dan