What's new
What's new

Accuracy of centre drill holes

Superbowl

Cast Iron
Joined
Feb 12, 2020
Yes again. If misaligned, the deeper the 60° hole, the bigger the gap at the opposite end with the same angle...
Therefore, to do the most accurate work possible, you want to drill the shallowest holes possible that will keep the work in place on the lathe. Depth, then depends on the size of the work and how agressive you plan to cut.
 

eKretz

Diamond; Mod Squad
Joined
Mar 27, 2005
Location
Northwest Indiana, USA
Therefore, to do the most accurate work possible, you want to drill the shallowest holes possible that will keep the work in place on the lathe. Depth, then depends on the size of the work and how agressive you plan to cut.

Nope, that won't work either. Part of the stiffness and rigidity of the center is getting a good solid seat over a large enough surface area. As you mentioned, one also needs to have enough surface area to support the weight of the work and the force that will be applied to it. And yet another consideration is wear. If you make the centers very tiny, they will come loose pretty quickly due to wear. Best to just use the usual size centers and get them inline!
 

michiganbuck

Diamond
Joined
Jun 28, 2012
Location
Mt Clemens, Michigan 48035
I haven't read all the posts but am I missing the obvious here?

When a shaft is turned between centers wont it be as round/straight/true on the diameter as the machine is capable? Even if the ends are not drilled exactly in the center of the stock?
That is what I figure, if you have stock to clean up each end will be true to that ends center. A really poor center, out of round or something, too small too shallow, not enough pressure may cause some wobble at the/one/both ends, that is why a center hone/lap can help a high precision part the run at its best. and you can take it out of the machine, and put it back in, even at a different 180 and it will still run true.
 
Last edited:

guythatbrews

Hot Rolled
Joined
Dec 14, 2017
Location
MO, USA
Don't forget that if you are removing a lot of material you will almost certainly have shaft movement due to stress relief. At that point, depending on how much things move it may be a good idea to recut the centers if you need a good straight, round and true finished part. It really doesn't take very long to do.
This is a good point. If the part moves a lot during roughing it may be essential.

Made lots of transmission main shafts about 21" long facing and centering just as you suggest. 2.375 stock both ends turned to 1.375 with a hub closer to one end. No trouble holding .001 TIR on finished part. Subsequently case hardened, centers lightly lapped, and straightened to .0015 TIR.

I havent used a face driver on 6" material but I assume the axial force required to drive this 21" long chunk is pretty high. If the face is not perpendicular to the axis formed by the centers you may have a problem, unless the face driver can align itself to the crooked face. If the finished part has a thin head and small shaft ( like a nail) you'll have a lot more issues than a thick diameter part.

If the part is 21" long and the centers are within .010 TIR I think each center will be tilted by less than 1 minute of arc. Pretty close. The long part helps a bunch with this. A short part will not be so forgiving.

Make sure your fixture is on the rotary table center and on the spindle centerline and you should be able to be much closer than .010 tir. If the od runs out a bit no problem assuming enough stock to finish. Unless the stock you are centering is already very round and straight it's very difficult to check the coaxiality of two centers 21" apart within 1' of arc anyway so rely on a good setup.

The part is bigger so you will probably use a bigger center hole. Bigger centers cause more problems. We made 21" impression cylinders about 6 feet long IIRC and those centers were big and bored with the compound set with a sine bar to ensure 60*.

If the part is to be heat treated and subsequently ground the centers may go out of round or taper and need lapping.

There is really no one size fits all solution. I would never put in a center by hand (emergencies excluded) but I also wouldn't spend a lot of time (money) making perfect centers if they aren't required by tolerance/processing demands. Do the best job you can aligning the fixture and you'll do fine.
 

jscpm

Titanium
Joined
May 4, 2010
Location
Cambridge, MA
Non-collinearity of centers is a source of error. What happens is that the cones are not aligned, so one of the centers rides cockeyed, which causes a taper in the work. The question is "How much?" and this depends on how bad the lack of collinearity is. If the bar is a real banana, then the centers will probably be worse.

One strategy is to drill one center, then mount the bar on that center, put the bar in a steady rest, then drill the other center. This can potentially greatly increase collinearity.
 

EPAIII

Diamond
Joined
Nov 23, 2003
Location
Beaumont, TX, USA
I do have a lot of respect for your advise and I am dead certain that you have more experience than I. That applies especially to large parts where I have almost none. A 14" lathe was the largest I ever worked with and even then the parts I made were under 5" diameter.

In any case, where we disagree, my advise to the OP is to go with your wisdom first and only consider listening to me if that does not work.

I do see and agree with your point about the accuracy of the center holes being important. I just did not consider the degree to which the errors might extend. You talk about the misalignment between the cone of the center and that of the center hole and that certainly can be a problem. The center at the head stock revolves with the work so the problem is not there. Or do you see one there also. But I wonder, and don't blame me for wondering, if a live center at the tail stock might be a good idea. There could still be wallowing, but it would end most of the scraping. But then, I don't even know if you can get one big enough for the OP's part.

I also wonder if it would be helpful if we knew more about the final part he is making. And the actual level of accuracy he needs to attain. And just how rough his rough stock is.

Also, for the life in me, I can not see how a horizontal milling machine would solve the problem. A large one may be able to face both ends of his stock at the same time, ensuring parallelism, but that does not provide two aligned center holes. I guess I am missing something.

Here I go again, but it seems to me if you want to drill two aligned center holes, a vertical milling machine with a right angle head (to drill horizontally) and a table large enough so that a rotary table can be mounted with the vee blocks on top of that. Set the RT at zero degrees and drill one end of the stock. Then rotate it 180 degrees and drill the other. If the RT is aligned below the spindle and the right angle head is aligned horizontally then the two holes would be in line to the accuracy of the set-up. And there may be no need to face the two ends before mounting it in the lathe. It seems to me that would be the fastest way, but it depends on having all of the above.

OK, now I see it. The same thing could be done on a horizontal mill and without the right angle head. And the RT would be in front of the horizontal spindle.



Most shops these days do not often turn between centers. When they do, it's usually for OD grinding work. And if I'm not mistaken, the OP is not asking questions about a workpiece that he can fit in a collet, and almost certainly using a live center at one end and a chuck at the other...

And you are wrong about the centers not needing to be accurate. They absolutely do need to be - if the work to be done is to be of good quality. I've already explained that. Have you ever actually worked in a professional machine shop environment? Sometimes you seem to attempt to say things with an authoritative manner ("the only requirement is"...") that are very much at odds with my experience, anyway.

In addition - if you DO run between dead centers it's still important. Between dead centers it's less likely that center pressure will flex a shaft, but center misalignment means the center will contact on one side near its point and the other side near its base. Then you're going to revolve it around for thousands upon thousands of revolutions as that point contact pushes all the center lube out of the contact zone... Doing that over and over again will wear both your center point and your workpiece's center all to hell and gone. Getting things right from the start means more accurate work and longer lasting, more accurate tooling. Up to you if that's not an important factor in doing quality work.
 

eKretz

Diamond; Mod Squad
Joined
Mar 27, 2005
Location
Northwest Indiana, USA
I do have a lot of respect for your advise and I am dead certain that you have more experience than I. That applies especially to large parts where I have almost none. A 14" lathe was the largest I ever worked with and even then the parts I made were under 5" diameter.

In any case, where we disagree, my advise to the OP is to go with your wisdom first and only consider listening to me if that does not work.

I do see and agree with your point about the accuracy of the center holes being important. I just did not consider the degree to which the errors might extend. You talk about the misalignment between the cone of the center and that of the center hole and that certainly can be a problem. The center at the head stock revolves with the work so the problem is not there. Or do you see one there also. But I wonder, and don't blame me for wondering, if a live center at the tail stock might be a good idea. There could still be wallowing, but it would end most of the scraping. But then, I don't even know if you can get one big enough for the OP's part.

I also wonder if it would be helpful if we knew more about the final part he is making. And the actual level of accuracy he needs to attain. And just how rough his rough stock is.

Also, for the life in me, I can not see how a horizontal milling machine would solve the problem. A large one may be able to face both ends of his stock at the same time, ensuring parallelism, but that does not provide two aligned center holes. I guess I am missing something.

Here I go again, but it seems to me if you want to drill two aligned center holes, a vertical milling machine with a right angle head (to drill horizontally) and a table large enough so that a rotary table can be mounted with the vee blocks on top of that. Set the RT at zero degrees and drill one end of the stock. Then rotate it 180 degrees and drill the other. If the RT is aligned below the spindle and the right angle head is aligned horizontally then the two holes would be in line to the accuracy of the set-up. And there may be no need to face the two ends before mounting it in the lathe. It seems to me that would be the fastest way, but it depends on having all of the above.

OK, now I see it. The same thing could be done on a horizontal mill and without the right angle head. And the RT would be in front of the horizontal spindle.

Yes, your vertical mill approach is basically the same exact thing as using a horizontal mill... The important part is getting the alignment correct. For that you need a machine that is in good shape, and also a rotary that's in good shape. The table must be set with its axis of rotation square to the mill, which is sometimes a trick on old worn out equipment. A cylindrical square is very helpful for that. Once that is set, the horizontal's head doesn't have to be moved up or down, it stays right on center. The x-axis may need to be moved if the part hasn't been set dead center of the rotary axis.

Most lathe work these days in a working professional shop will be done in a chuck with a live center. Even so, a dead center in the headstock can still cause some alignment issues - think of it this way: your center cone in the workpiece is at an angle to the face and you push it hard against the dead center with the center at the other end. The work will tend to try to twist or kick over to realign that cone. On a very substantial workpiece it probably won't have much luck. On a thin long one however, it will almost certainly put a bow in the part as it flexes. This is less of a problem with dead centers since they're generally run at lower pressures, but it's still there.

My advice to any reader is to listen to all advice offered here or anywhere else and then make a few choices based on experience then to try a few different experiments with techniques for themselves. One should be capable of listening to advice with an open mind and trying a few things, then making the decision on what will work best in each specific situation.
 
Last edited:

dian

Titanium
Joined
Feb 22, 2010
Location
ch
Material is 160mm (6.3")diameter and 550mm long (21.5") cold rolled bar. About 1000 of them in few variations.

It's too big to be chucked deep inside chuck so i would like to get it face and center drilled on horizontal mill.I was thinking about putting v-blocks on top of palette i can rotate 180 degrees .


My real question was what KIND of problem i can expect. If centers are coaxial but not In center of stock then raw outside diameter will have some runout but if you have enough stock that's not a big problem.
But what happens if centres are not coaxial? Or the are not parallel?
Will it affect reatability of clamping between centers?
I'm wondering how accurate my fixture has to be guarantee coaxial and parallel of centres. But of course same applies in drilling in lathe chuck and rotating especially if holding on rough stock.
sorry for not having read the whole thread. there are center drills with rounded flutes just for that (usually marked with "r"). no idea if rounded nose centers exist, but usually you use a bearing ball and a hollow center. i have addapters i can stick on a regular center. not that hard to make.
 

michiganbuck

Diamond
Joined
Jun 28, 2012
Location
Mt Clemens, Michigan 48035
Did the op say how much stock He has to take? flipping the part 180* as held in a V block fixture with a horizontal mill seems a failure is like to happen. Hand scribe crossing lines and hand punch, then look with a loop and Chicago the pinch mark and get to <.005 error may be the way to go.
 
Last edited:

implmex

Titanium
Joined
Jun 23, 2002
Location
Vancouver BC Canada
Hi All:
I have read this thread with great interest; here's my take on the subject.
Facing/centering machines like those shown by Emanuel Goldstein have two opposed spindles that are supposed to be exactly co-axial.
The shaft never moves...only the two heads move toward one another.
So no matter how rough or bent the stock is, the centers will always be concentric to one another even if they are not concentric to the OD of the shaft.
This is the desired state for all of the reasons EKretz points out.

You CANNOT achieve the same thing as soon as you move and re-clamp the stock; I don't care if it's on a horizontal, or a vertical mill with a right angle head, or a lathe with a steady.

If your bar is pretty straight and pretty round you can get decently close, but it's ALWAYS a crap shoot.

For most general turning, you can fudge it a bit and do OK enough that no one is going to reject the job.
But if you need to split tenths, you need better than you can get with a bodge, which, as some have pointed out is why facing/centering machines and center grinding machines and center lapping machines even exist, and it's why volume manufacturers are willing to drop millions into this capability.

So in order of desirability:
1) a facing /centering machine
2) A Horizontal big enough to swing the shaft on a rotary so it never has to be un-clamped.
3) A bridge mill big enough to swing a right angle head and still get to each end.
4) All the methods that accept drilling one end, un-clamping the blank, flipping and re-clamping it, and drilling the second side.
This includes lathes, horizontals, drill presses vertical mills, etc etc.
Way way down on the list is a center punch mark and a pistol drill...but in a pinch for a low tolerance part or a shaft that can have the centers re-trued in the lathe with a steady and a boring bar before finish turning, sometimes you do what you have to do.

Having said that, I wouldn't bother centering both ends...If I need really good centers, I always center one end, grab the other in a 4 jaw chuck, turn two spots for a steady, (one close to the chuck, one close to the tailstock.)
I mount the steady and taper bore the center true, then re-set the center and skim the area I had the steady on to make it as concentric to the taper turned center as I can.
I skim the other steady rest area again too, just to be sure all is concentric, and to make the diameters identical (makes setting the steady easier)
Then I flip the bar and dial in the headstock end on the area I had the steady running on originally using short soft pads in the 4 jaw chuck so the part can align itself to the steady without bending.
I support the new tailstock end in the steady, then center drill and taper bore the second center.
I live with the outcome knowing it is pretty good considering I violated a basic rule and flipped the bar to get the second center in.

This is all a lot of fucking around, but it does a pretty good job of making co-axial centers that are properly conical and can support precision work.

Cheers

Marcus
www.implant-mechanix.com
www.vancouverwireedm.com
 








 
Top