PDA

View Full Version : Compound for tenths



Furnace
08-13-2008, 10:13 PM
First off I have searched for this and didnt see it. My question is: what angle would I set my compound to take off .0001 for every .001 turned on the dial. Lets say this is for a boring operation and every thing was perfect on the lathe. I know I have seen it but couldnt find it when searching. Thanks

toastydeath
08-13-2008, 10:22 PM
About six degrees.

asin(1/10) ~= 5.7392 deg

sin(6 deg) ~= .1045

Edit:

Although just to say, I will be impressed if you can measure tenths in a bore reliably, much less machine to tenths reliably.

Mcgyver
08-13-2008, 11:06 PM
thats what i do, try to approximate 5.75, the line beside the 6 line not at it. It works really well when turning for bearings and such when you want to be within a couple of tenths. I never a saw a tenth as the best i can measure is a tenth. Agree with toasty though, if you can measure bores to a 10th that is something.....still great way to go when you want better than a thou

.RC.
08-13-2008, 11:32 PM
I cannot imagine there would be too many lathes that could turn to something to tenths that is quite round..Probably only the Hardinge HLV the CVA and the Monarch 10EE would do it consistently..

That is why cylindrical grinders were invented

oldtiffie
08-13-2008, 11:49 PM
I agree with TD and Ringer.

If it were me, I'd be more confident about the cross-slide moving in (or out) a "tenth" (or in that zone) if I had a DRO or a good dial indicator with its base on the carriage and its needle/plunger on the tool-post (or cross-slide).

Putting a "thou" on the tilted compound/top slide in the expectation of moving the tool inward/outward a "tenth" is one thing in theory but may well be another expectation (hope?) not met in practice.

This is either "re-run the tool as it is for a second cut" territory or time for a well-used fine file or emery paper.

I can hold 2 tenths in a bore pretty easily and consistently using spring calipers and a good micrometer. It takes some practice, but it works. Ask any older Machinists.

Given the tolerances/allowances allowed in most situations, you should use all the zone. It will be larger than a tenth for either the "shaft" or "hole" independently (if you have to make both) but it becomes much more generous if one part exists and your are making the other (mating) part. In that case you can combine the tolerances/allowances for both - and use it!! It is huge compared to a single "tenth" and it will work perfectly well.

Moving the cross-slide a "tenth" will - in theory if not in practice - reduce a shaft or enlarge a hole by 2 tenths.

KiddZimaHater
08-13-2008, 11:57 PM
80 grit Sandpaper

Forrest Addy
08-14-2008, 12:14 AM
Waste of time. Dialing in a small amount seldom results in that amount of stock removal, your tool may not even cut if there less than 0.0015" or more. You need a certain minimum cut to make the tool work properly and that cut varies from material to material. Better to use a 0.0001" indicator on a biscuit mag base. You'll still need to attach reference surfaces for the indicator to contact.

The way I taught my apprentices back when is to rough down to 0.025 or so over size then take the remainder in two equal cuts adjusting the last cut if necessary to arrive at the finish size. Later as you gain confidence and experience with your machine, you can go right to size. If your tolerance is really picky leave a couple thou stock and use a stone, file, wet or dry, brake cylinder hone etc to make the final size.

oldtiffie
08-14-2008, 12:19 AM
Good sage advice Forrest.

But be careful here. That is "old-fashioned" stuff and is bordering on or really is heresy for some.

But it sure works for me too - always has.

wierdscience
08-14-2008, 12:29 AM
Emery cloth and a file.

Unless the material you are cutting is fairly hard and the toolbit is very sharp and the lathe is very ridgid and balanced then you are in the realm of the OD grinder.

Fasttrack
08-14-2008, 01:37 AM
Waste of time. Dialing in a small amount seldom results in that amount of stock removal, your tool may not even cut if there less than 0.0015" or more. You need a certain minimum cut to make the tool work properly and that cut varies from material to material. Better to use a 0.0001" indicator on a biscuit mag base. You'll still need to attach reference surfaces for the indicator to contact.

The way I taught my apprentices back when is to rough down to 0.025 or so over size then take the remainder in two equal cuts adjusting the last cut if necessary to arrive at the finish size. Later as you gain confidence and experience with your machine, you can go right to size. If your tolerance is really picky leave a couple thou stock and use a stone, file, wet or dry, brake cylinder hone etc to make the final size.

I learned that the hard way. With my Smithy, it is flexible enough that the reading on the dial very seldom equalls what is actually taken off. I always do three swipes at the end. When I get within .03 or so, I stop. Measure the part and move the cross slide in .005. Cut and measure again to see if it actually cut .005. If it did, I continue moving in .005 measuring after each pass. If it doesn't, then I can compensate on my second to last pass so my last pass is dead on. Not only is flex a big issue, but I'm not convinced that the screw is ground quite right. The dial is only graduated to .002 so .005 has to be estimated, but I've noticed some discrpenacies with large pieces. If you continually take .010 DOC cuts for a large distance, by the end you are about .005 off where you should be. I thought it was due just to flex, but I checked it with a dial indicator. Sure enough, the table travel measured using some 123 blocks and dial indicator is not the same as whats measured on the dial. Its not far off and most of that .005 is flex... but still :(

Mcgyver
08-14-2008, 08:11 AM
I think you guys are a too negative on this - I find it works well. very sharp hss tool is required. if say 1.8 thou to come off, I can hit it just about dead on by advancing 9 on the compound...obviously its more like guess work trying to advance .0009 on crossfeed.... this is just a simply a way to get greater resolution on on infeed movement. I also use the trial cut strategy, doing a test cut or two as i come up to final size.

I can see small cuts with carbide being a problem but with sharp hss I think its easy to take a small cut, I suppose that will vary with material though but for most stuff i use its ok.

and why would it surprise anyone that a lathe can turn something round to a tenths? you're saying that you turn something, put a tenths indicator on it and moves all over the place? It doesn't have to be a Hardinge to have bearings that will rotate to better than a tenth. Using a decent lathe with a sharp hss tool this technique works, or at least has for me.

PaulT
08-14-2008, 11:21 AM
I find this method useful for boring when I need to get better than 0.001". I'm not saying I get 0.0001" accuracy, but if you work carefully, on bores that aren't too deep (one inch or so) I can get +/-.0002" pretty reliably.

Use a very sharp tool, high RPM, low feed and it works well. Definitely useful for bearing housings, etc.

On my 5914 it has pretty prominent markings at each 0.005" increment, so another way to go is to set the compound at 11.5 degrees, this way each 0.005" tic on the dial takes off 0.0001".

Paul T.

BobWarfield
08-14-2008, 01:10 PM
I think you guys are a too negative on this - I find it works well. very sharp hss tool is required. if say 1.8 thou to come off, I can hit it just about dead on by advancing 9 on the compound...obviously its more like guess work trying to advance .0009 on crossfeed.... this is just a simply a way to get greater resolution on on infeed movement. I also use the trial cut strategy, doing a test cut or two as i come up to final size.

I can see small cuts with carbide being a problem but with sharp hss I think its easy to take a small cut, I suppose that will vary with material though but for most stuff i use its ok.

and why would it surprise anyone that a lathe can turn something round to a tenths? you're saying that you turn something, put a tenths indicator on it and moves all over the place? It doesn't have to be a Hardinge to have bearings that will rotate to better than a tenth. Using a decent lathe with a sharp hss tool this technique works, or at least has for me.

I'm with Mcgyver on this. It isn't that you necessarily want to cut just a tenth as it is you want to be accurate to the tenth. I rough down to 0.010" and then take 2 finish passes based on micrometer readings to the tenth. With the compound set over I can adjust to the tenth what those two passes (each cutting approx 0.005" which is fine for my carbide inserts) need to be. Done quick and dirty I easily get 0.0005" accuracy even on my junky little lathe. With a lot of care I can get maybe 2 tenths. Given that termperature and a lot else really affects these levels of accuracy, a lot of care can be a real pain!

The real challenge is holding this over any length. For that I have to map the taper and "bump in". My little lathe has about 3 thou over 8 inches (need to tune it up some more!) taper with the tailstock perfectly aligned. So I need to bump it 0.000375" every inch of travel. This is most easily done my marking the zones and using the compound set over for fine adjustment:

http://www.cnccookbook.com/img/SteamEngines/VerburgRevOpenCol/P1010423.JPG

A slow feed rate and a steady hand are needed, but the result comes out suprisingly accurate. This "bumping" business is common practice for a seasoned machinist working with a worn machine. It works well.

Cheers,

BW

PS Rumor has it that a Mori Seiki will hold those tenths just fine. There are a number of fine lathes besides the Haridinge or Monarchs that will in fact.

Scishopguy
08-14-2008, 01:26 PM
Something that was overlooked so far is the use of a machinist's scraper to sneak up on a tenth. Most of the old timers I had the pleasure to work with used the scraper to get that last little bit when fitting a shaft to a bronze bushing. You mic it and scrape it until you get the fit you want. As a side benefit, you can make it (by how fast you advance the blade) so you cut microscopic oil grooves. This allows for less than 100% contact of the shaft, giving more places for oil to be retained.
However, this is just one type of turning and the examples you folks gave are a bit different. If you are careful and mic between each cut you can get good accuracy even on worn out lathes, you just have to "hold your mouth right," as they used to say. ;)

Fasttrack
08-14-2008, 03:09 PM
I think you guys are a too negative on this - I find it works well. very sharp hss tool is required. if say 1.8 thou to come off, I can hit it just about dead on by advancing 9 on the compound...obviously its more like guess work trying to advance .0009 on crossfeed.... this is just a simply a way to get greater resolution on on infeed movement. I also use the trial cut strategy, doing a test cut or two as i come up to final size.

I can see small cuts with carbide being a problem but with sharp hss I think its easy to take a small cut, I suppose that will vary with material though but for most stuff i use its ok.

and why would it surprise anyone that a lathe can turn something round to a tenths? you're saying that you turn something, put a tenths indicator on it and moves all over the place? It doesn't have to be a Hardinge to have bearings that will rotate to better than a tenth. Using a decent lathe with a sharp hss tool this technique works, or at least has for me.


Thats kind of what I was thinking... I've got a dial indicator with .0005 resolution but you can still get a feel for how round something is by watching needle movement between the graduations. On my Pacemaker, I dialed in pinion gear shaft from the other Pacemaker to within +/- .0002, meaning that the TIR was less than one graduation on the bearing surface. After cutting the section that the gear rides on, I checked it again just for kicks. The needle showed no perceptible movement. Maybe a flicker one side or the other of the "0" mark, but no real movement. I was skeptical and spent about 15 minutes trying to figure out what was going on. I thought that maybe the indicator wasn't contacting the surface but all was as it should be. Of course, surface finish comes into play at those sizes. It was a nice shiny surface where as most of the mild steel stuff I turn has a dull slightly rough finish on it.

Anyhow, my point is... this is on a 65 year old machine with pretty worn ways. I think any good quality lathe should be able to machine something round to 1 or 2 tenths. I mean, I don't expect that on my import 3-in-1 machine, but a "real" lathe should be able to.

Scishopguy - That is interesting. I can't imagine needing to work to those kind of tolerances. On my go-kart transmission project, I've got a 5/8" shaft riding in bronze bushings. I figured I was doing awsome at - .002. My oil pump will just have to provide a high enough flow to act as a "hydrodynamic" bearing like AK mentioned. :D

toastydeath
08-14-2008, 03:46 PM
You can't measure roundness on the machine you cut it on for obvious reasons.

At work, we have very tight tolerances (as we make air bearings). I can tell you from experience that a person may think they are working to tenths, even in roundness or over moderate spans, but I can assure you that they are not.

A person cannot machine to a tolerance they cannot accurately measure; one can only guess.

Fasttrack
08-14-2008, 04:16 PM
Oh... duh

:o

Must be the paint fumes getting to me...

lane
08-14-2008, 07:30 PM
Thats is what sand paper is for. Take some different size pieces of CRS 1/4 , ,3/8 1/2 . about 8 inches long Saw a slot down the center about 2 inches deep insert sand paper and roll it around bar . stick in hole and sand to size .

LES A W HARRIS
08-15-2008, 03:26 AM
About six degrees.

asin(1/10) ~= 5.7392 deg

sin(6 deg) ~= .1045

Edit:

Although just to say, I will be impressed if you can measure tenths in a bore reliably, much less machine to tenths reliably.

Tenth on dia would be fifty millionths on radius?:confused:

http://i37.photobucket.com/albums/e97/CURVIC9/2008%20GENERAL/LES15082008_011219.jpg

approx 3 on compound.

Another point, if machining to a shoulder one's length is constantly changing.

Cheers,

ka6gzj
08-15-2008, 04:17 PM
We learned in school to set the compound to 84deg. But my text book says
84deg. 16'.

hope this helps. jim

toastydeath
08-15-2008, 10:48 PM
Whoops on the math, I'm so used to having direct diameter readout that I forget compound slides are radius.

My apologies for the screw up.