Taking it to the final size

8ntsane

Active User
Registered
Joined
Dec 1, 2010
Messages
1,172
Ok Guys, I think we have all been here before.
I was over at a friends place, and he hasn,t been machining very long. He was not a happy camper that day, and swore his lathe was junk. I thought I would start this topic because we have all been in this bad spot before.

The target of the part is say, 2.750
At 2.780
creeping up on the final size
2.770
2.760
Then your allmost there, should one start sneaking up on the final size .002 at a time, or take the final cut?
I think we all have taken that final cut, and found out 2.750 ended up a 2.748, and damn near had a fit as the part did take some time to make.

Ive found over the yrs, many things can screw you over just when you think your mins from being done, now your starting again.

Here are just a few things Ive seen over the yrs.
1 Made the measurement while the part was still hot.
2 Used a caliper instead of a mic

3 trying to hit the final measurement without leaving any room for expansion of the work piece, and not leaving a few thou extra for the final fitting

4 Trying to hit your final size to a few tenths, and trying to do a fuzz cut or three to get there.

Now these are just a few ways to add another piece to the scrap bin.

Tell us how you get that final size, fit and finish you want.

Hand lap
Hone
Abraisives/ sand paper
tool post grinding

We all have our method to the madness, so lets hear how you do them. This thread could help many newbies from the frustration of trying to get the parts to size, and have it go bad at the end of the job.
 
my small chinese lathe will flex so .005 may be .003 or .004
I try to make the final pass with tool steel not carbide
and if fitting to a bearing or needing exact fit i have a fine stone tool post grinder I made from a hf saw blade sharpener that takes very fine cuts.
I have tried the files and sandpaper and I prefer to put on the grinder
steve
 
There is a tad more involved here. Tha main one, for me anyway, heat. If you take your part to final cut and it is hot, when it cools it won't be finish size anymore. It will probably be undersize. Also when I get close, say .010, I will take more than one pass at each setting. You would be surprized at how much the second pass takes off. Final cuts for me are .002. Yes, I like all of you have missed the mark before.

Please remember this is the way I do it, that doesn't make it the right way, it does make it my right way.

"Billy G" :whistle:
 
Well, since this is lathe work we're focusing on, I'll go....

I usually run flood coolant, especially if the material requires high spindle speeds/feeds....I know it will get warm. That minimizes the shrinkage effect.

I take the last .100 or so in even DoC. This gives me the time and information I need to take the final cut, as well as time for the part to cool back to ambient if needed. If I have 0.100, I take 0.050 and see if it really did, and make an adjustment if needed in the DRO, dial, or indicator (whatever I am using). If it takes 0.048, or 0.051, I mentally note that and create an offset in the tool nose. The next cut will be 0.025, based on that offset. Hopefully it will be much closer to "as dialed". If not, I repeat the process. Rarely do I need to make more than a few tenths correction at that point. If I do, the insert is dull, not on center, etc. That's another problem. If the correction is on the order of a couple of tenths, I make the correction and do a pass I've learned to call the "mic check" pass. This assumes of course that the insert will be repeatable. If I have 0.015 (or 0.010, depending on what kind of material it is) and that agrees with the dial, DRO, or indicator, I dial it off and cut. I just finished with some parts that have a bearing fit specified to be 0.4996 +/- 0.0002. Worked every time. No files, paper, or fuzz cuts.
 
There we go, thats a good start to this thread.
Tips from our more experianced Good Ole Boys, allways good!

Keep em coming
 
I'm with Tony on this, right down to the inserts. Start taking cuts to get the finish you need and that normally means some consistency. I just find out what it is consistently cutting. Tony calls it an offset but I think of it as a multiple. IE if I take .015 this cut then .051 should have me there in 4 passes. Check/measure each time and adjust as needed (hmm Tony;s offset?)but stay in the same ballpark. It usually works well but sometimes it seems the decreasing surface speed, gravitational forces, or something makes the final cut not as pretty as the previous cuts and at that moment I just start hoping I missed on the large side. It worked a lot better on the Navy's heavy lathes than my current lathe but it's the method I know.

The cuts I really hate are the ones where the final size is so close to the stock size it's pretty much skim, measure, skim and measure, much the same as above but with HSS and less feed and much lighter cuts.

I use shear type tools to get me out of a bind when I'm like say .001 off. With that tool you can literally make swarf that floats down like snow flakes.

Steve
 
My first lathe was a 12" Craftsman Atlas. That blasted thing was too flexible and would lie to me all the time!! Drove me nuts. 2nd. lathe was a Jet Taiwan 1024. INFINITELY BETTER. Did not bend under a cut and give false readings.
 
Steve, I think I picked up the term "offest" from the CNC days where the controls would allow you to add or subtract a little to compensate for either pushoff, tool wear, erroneous zero setting, etc.. If you expected the part to measure 1.0000 and it didn't, you could enter an "offset" factor on the control to adjust it's theoretical position with the actual position. It's basically the same as cutter comp on the mill. Most carbide end mills actually measure a couple under nominal, and cut small as well. So....you "lie" to the control and tell it the end mill is smaller than it is. For circular interpolation, you can dial in a close size with this method without actually having to change the program. That way, next time you run the program, with different tools, you don't get bit by a wrong tool setup.
 
Last edited:
All good infro guys.
Hitting the target size can be a challenge for many less experianced members here. For the experianced members, getting to final size comes much easyer than a fellow thats startin out.

As allready mentioned, heat plays a big factor in measurement, and final size. What you use for a cutting tool will affect the quality and surface finish of a final cut. I see some like to grind to the finished diamentions, and yet others still use inserts. Interesting that many different methods are used.

I have found that heat has screwed me over enough times, that I will shut the machine down and wait till the part cools when Im getting close.
Depending on what Im working with material wise, I will set the compound over to 45 degree,s and feed into the cut with the compound rather than the crosslide for the last few thou.

I know once I do start feeding in with the compound, rather than the crosslide, Im without my DRO, and have to take a measurement after each pass. When Im getting down to the final passes, I will usually switch over to HSS cutting tool. The HSS cutter will also has enough nose radi to produce a smooth finish.

Allso, if switching over the tooling, you should do this well ahead of time, and not on the last .002 of the cut. You,ll need to make a few cuts, measure and check your surface finish prior to getting to final size. Once getting close, I also like to make a spring pass too, and check measurement, and see if Im getting possibly a better surface finish to boot. I perfer using a Mic to verify my cuts, and spring passes. Calipers are fine for reading the roughing in, but its to easy to apply to much pressure when using a caliper, and can easyly fudge your reading. Yes, us experianced guys can use either, but for the newb, he may not have the feel yet.

Tony,s method using flood coolant is a good idea, and I do it sometimes myself. It can shorten the cooldown period, and waiting time sometimes required to accuratley measure your work piece.
The flood coolant can also help with surface finish as well. But not all home shop machines have this.

As George has mentioned about his machines not being the most ridgid, and I have used machines that would not be reliable to get with in .002 for final size. I see no problem with using sand paper of various grits to get you to the final size either. I have used some lathes that I would not trust to take me all the way, and the paper of the right grit will get you there, and clean up the surface finish at the same time.

The tool post grinder isnt something I like to use, but at times it is a life saver. Ive had some times when I can,t get the smooth finish I want, and need. The material is hard, and what ever I try isnt working out to plan, out comes the tool post grinder. It seems when ive only got a few thou left to go, and I know nothing else will do, the TP grinder has saved the day, and the job more than a few times for me.

So, as one can see from the replys in this post, there is many things that effect your final size, fit and finish. And many things that can bite your ass in the process. Hopefully, the replys from all that have posted will help the newbie from making his scrap bin larger.

For the newb, getting a decent fit is what most are striving for, and getting there is part of the learning process.

Keep the infro coming guys
 
On some lathes, in certain situations, when you feed up to a shoulder, or through a bore, simply shut down and reverse the feed (or hand feed slowly back off the cut if it's short), and you won't have disturbed the actual position of the tool. Often you'll see a bit of material coming off the part during this process. This is pushoff being cut, either from the part deflecting, or the tool being literally "pushed off" the part. It's the same as a spring pass, except the back of the tool is usually sharper than the front that just cut off all the material, especially if you have taken several passes. When you get close to finish size, doing this will allow you to take moves only in one direction, allowing the carriage and cross-slide assembly to remain in the "stressed" position they get in during a cut. You'll get more predictable cuts this way. If you are really sneaking up on it, do this with 0.005 left. Dial in 0.003 and "drag" the tool back (with the spindle running) and then stop for a measurement. Dial what is left and poof, you're there. Feed on.......feed off. Presto! (and no, it's not like Wax on!....Wax off!)
 
Back
Top