Tips for collinear center drilling

[jscpm]

When turning on centers, a source of error is non-collinearity of the center holes.

Most people, myself included, just chuck up each end of the bar and drill it from the tailstock, but the problem with this strategy is that if the bar is not collinear (which it never is) then the centers will not be collinear either. So, the question is: is there a better mousetrap for drilling centers?

If you think about it, one idealistic idea would be to first face both ends of the bar and then lap them parallel on a surface plate. If the bar was then indicated in the chuck so that the face was orthogonal to the ways, then you would at least be guaranteed that the centers would be parallel. They might still be offset, but would be parallel.

To make them aligned X-Y as well, would be much more tricky. It's hard to see how to do that easily.

 

[Gordon Heaton]

In practice, when would it matter?

 

[jscpm]

In practice, when would it matter?

When the degree of the resulting taper is unacceptable. For example, it can readily be seen that the X-Y offset will translate directly into a taper, because it is as though the tailstock was set over. So, if the axial offset error is, say 0.002" on a 12" rod, then that will be the degree of taper. If there is an angular error, then not only will the taper be worse, but the rod will be out of round as well.

 

[sfriedberg]

I had this at the bottom of the message, and am moving it up to the top. If you turn between centers, the cut is going to be concentric. It does not matter if the center axes are misaligned. If you can hold the work between dead centers, the cut will be concentric. The bad that happens with misaligned center axes is that the tip of the dead center is going to tear up/deform the female drill center cone, because it is bearing only at one point rather than making full (conical) contact. You are not going to turn a taper. You are going to start cutting one side at one end and the opposite side of the work at the opposite end of the work. When the cut is deep enough to cover the whole work, and if your tailstock is aligned, the result is a cylinder. If it's a taper, it's because the tailstock is off (or the bed is twisted), not because the initial center holes were misaligned.
I was taught to drill the centers, then turn or grind the OD beween (dead) centers. The result is concentric for sure, and taper is easily checked (and corrected) during the turn/grind. This will tolerate a modest (few thou) misalignment of center axes and a minor non-parallelism of center axes, and if you flinch at that thought use Bell center drills so the work can line itself up exactly. (We did have someone on PM in the last couple of weeks reminding us about turning multi-ton workpieces on centers and how quickly that will eat up center holes without good conical contact on the centers. Acknowledged.)
If you absolutely must drill your center holes without trimming the OD later, I don't know of any technique other than testing on bench centers and then tediously lapping or scraping the center holes to correct alignment of the centers.
There is a machine technique for cleaning up center holes after they've been abused, or get rid of angular misalignment. There are center lapping machines which look like skinny drill presses that hold the work vertically. Register one (beat up) center on a male center below and bring down an abrasive cone into the other (beat up) center hole. Flip the work and repeat. This will improve the center hole surface and effectively eliminate angular misalignment between center axes. Snow used to make machines like this, maybe they still do.
Given a choice, I hold the bar in a collet and face the ends before drilling. But if I'm working on HRS material, the OD scale and surface roughness is going to dominate concentricity error. Even if I am using TGP material, the collet will have some eccentricity. If I want best concentricity, I plan to turn the OD between centers, after drilling the centers.

 

[eKretz]

It's actually pretty easy, depending on whether you are able to turn the OD. As noted in the other recent thread, it can be done on an HBM with a rotary table fairly simply to a decently close level.

It can also be done in a lathe with a roller or steady rest. Set the chuck end true, run the tailstock end in the rest and drill or tool the centers with the compound and a small boring bar. Turn the part around and set the chuck end true again, run the tailstock end in the rest and repeat. The centers will be concentric and inline. If the OD isn't true and round to start, you'll need to cut spots first for the rest to run on or use a cathead.

As sfriedberg noted, the centers being out of line is not the same as a tailstock offset. Centers out of alignment can certainly still cause problems, especially with small diameter, long work. Those can include taper (but for a different reason than tailstock offset - more related to center pressure pushing the part or center itself one way or the other), issues with part straightness after releasing center pressure, concentricity problems and roundness problems.

 

[EPAIII]

EDIT:
After reading and posting on the other thread about this, I probably should erase this post. But I am not doing that. But please read what I said there before blasting me. I was not understanding exactly what was going on. As I see it now, the best tip for getting the two center holes on the same line (co-linear) is to use a milling machine, a RT, and some vee blocks as described there.
End EDIT.

As I understand it, the whole thing about turning between centers is that the center holes are the VERY FIRST feature that is made to the RAW STOCK.

Then ALL the other features are turned while the stock is supported by the centers. This ensures that all the other/SUBSEQUENT features are concentric with each other as well as with the original center holes.

The main point here is that the two center holes come FIRST, not some other feature.

If you are starting with a piece of ground rod as your stock, then a good quality collet would be a better choice for work holding as it can guarantee a high degree of concentricity with the ground OD of the stock. This is the whole point of collets.

If you are adding center holes to ground stock and then adding features that must be concentric with that ground OD, you are just fighting the system. But if you must, then you should either use a DI to center the stock in a four jaw chuck (or other adjustable chuck) or use collets to hold the stock while center drilling using a bit in the tailstock.

That's the way the text books teach it. But I am sure you will find folks on the shop floor who do it otherwise for one reason or another. The good ones know what they are doing and produce good results. The not-so-good ones are just doing a bad imitation of what they THINK they saw the good ones doing. Their results, of course, will vary.

 

[Bill D]

As long as the raw stock is slightly oversize it will be turned into a cylinder regardless of misplaced centers. It is much easier and cheaper to buy larger stock then arttepmt to mount it perfectly.
Bill D

 

[jscpm]

Now that I think about it, the only way to ensure collinearity is to have setup that can drill from both ends. If both your drills are collinear, then your centers will be collinear no matter what the shape of the workpiece is.

On the other hand, if you have just one drill position and move the workpiece, such as by flipping it, then all bets are off, because the result will depend on the shape of the workpiece and how it interfaces to the fixture holding it. In fact, if the workpiece is symmetrical with respect to the fixture, then flipping it will actually work, even if the bar is a banana. The bottom line is that you are rolling the dice if you do not have dual drills that are proven to be collinear.

So, the question is, "How do I get dual inline drill positions?" and one obvious answer is to just provide collinear bushings. So, it is actually kind of easy, if you think about it. You just get or make a test bar and use it to line up and fix two bushings. Then you position the workpiece between the two bushings, and center drill through each of the bushings and voila, guaranteed collinearity (at least to the degree of the linearity of your test bar).

 

[Mtndew]

What I've always done:
Face and center drill
flip
Face,center drill and turn the od for about 1/4"
Hold on to that 1/4" with soft jaws, engage the tailstock and turn your part.

 

[DDoug]

Now that I think about it, the only way to ensure collinearity is to have setup that can drill from both ends. If both your drills are collinear, then your centers will be collinear no matter what the shape of the workpiece is.

On the other hand, if you have just one drill position and move the workpiece, such as by flipping it, then all bets are off, because the result will depend on the shape of the workpiece and how it interfaces to the fixture holding it. In fact, if the workpiece is symmetrical with respect to the fixture, then flipping it will actually work, even if the bar is a banana. The bottom line is that you are rolling the dice if you do not have dual drills that are proven to be collinear.

So, the question is, "How do I get dual inline drill positions?" and one obvious answer is to just provide collinear bushings. So, it is actually kind of easy, if you think about it. You just get or make a test bar and use it to line up and fix two bushings. Then you position the workpiece between the two bushings, and center drill through each of the bushings and voila, guaranteed collinearity (at least to the degree of the linearity of your test bar).

You really need to get out more:

Hey shaft facing and centering machines at DuckDuckGo

DuckDuckGo. Privacy, Simplified.

duckduckgo.com

 

[michiganbuck]

If one set one center on a tail center (dead is best) and laps the front center with the likes of a Norton point mounted lap hone held in the headstock (or the other way around)
Then flips the part to hone lap the other then both will be near perfect in a minute or a few.
In close grinding, we had to repair crunched centers and even restore them back to center, no big deal at all.
And some shops have an Excello center lap machine, just as good or better.
A standing drill press (floor model) can be easily made into a center lap machine.

 

[eKretz]

If one set one center on a tail center (dead is best) and laps the front center with the likes of a Norton point mounted lap hone held in the headstock (or the other way around)
Then flips the part to hone lap the other then both will be near perfect in a minute or a few.
In close grinding, we had to repair crunched centers and even restore them back to center, no big deal at all.
And some shops have an Excello center lap machine, just as good or better.
A standing drill press (floor model) can be easily made into a center lap machine.

That works very well indeed for small parts. Not gonna fly for the hundred pound and above category though.

 

[EPAIII]

Do I have to draw pictures?

Read the second half of my last post (#47) in the other thread. SLOWLY.

Accuracy of centre drill holes

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...

www.practicalmachinist.com

 

[eKretz]

Do I have to draw pictures?

Read the second half of my last post (#47) in the other thread. SLOWLY.

Accuracy of centre drill holes

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...

www.practicalmachinist.com

Who are you aiming that comment at?

 

[guythatbrews]

I really haven't seen a problem with this for work as close as .0002 as long as reasonable care is taken with alignment, no matter how the centers are created. I can't speak to more precision and giant parts which surely have their own set of problems.

Uneven center wear and eccentric centers seating in during machining aside, use reasonable care and you are fine for general work.

As pointed out center lapping machines are used to align centers and improve the form. But only for smaller work.

Facing and centering machines may do a good job or not depending on setup alignment. Seems like maybe their main function is increasing production rate, which they do well.

Collets are great but not foolproof. You don't really know what the free end is doing back in the spindle, unless you put a bushing back there. Same with a chuck.

And how do you check center alignment on a piece of rough stock? It's next to impossible or at least very time consuming.

Really the only center problems I can recall relate to poorly drilled centers out of round or poorly chased centers incorrect angle or clearance hole not deep enough and point hitting. The worst was a family of parts about 5" long 5/8 dia with a 3/8-24 hole in one end and a key seat. Shaft was casehardened and the big center(which is bad) would go out of round (worse). The centers had to be hard turned.

My bottom line is make good round centers and do your practical best to ensure alignment.

 

[eKretz]

I really haven't seen a problem with this for work as close as .0002 as long as reasonable care is taken with alignment, no matter how the centers are created. I can't speak to more precision and giant parts which surely have their own set of problems.

Uneven center wear and eccentric centers seating in during machining aside, use reasonable care and you are fine for general work.

As pointed out center lapping machines are used to align centers and improve the form. But only for smaller work.

Facing and centering machines may do a good job or not depending on setup alignment. Seems like maybe their main function is increasing production rate, which they do well.

Collets are great but not foolproof. You don't really know what the free end is doing back in the spindle, unless you put a bushing back there. Same with a chuck.

And how do you check center alignment on a piece of rough stock? It's next to impossible or at least very time consuming.

Really the only center problems I can recall relate to poorly drilled centers out of round or poorly chased centers incorrect angle or clearance hole not deep enough and point hitting. The worst was a family of parts about 5" long 5/8 dia with a 3/8-24 hole in one end and a key seat. Shaft was casehardened and the big center(which is bad) would go out of round (worse). The centers had to be hard turned.

My bottom line is make good round centers and do your practical best to ensure alignment.

It's easy to get alignment on rough parts with an HBM and the rotary table. When the table is set correctly, center height in the y-axis doesn't change when the table is spun around. The center of the table in the x-axis also doesn't move when the table rotates... If the part is on center the machine doesn't need to be moved at all to center both ends. Only the rotary table needs to be spun 180°. If the part is off center, the machine only needs to be moved in the x-axis equally in either direction relative to center.

This can easily be accomplished in the lathe as I described earlier also. Cut some spots to either run a rest on or simply to indicate to, then center both ends with those spots running true. The centers will be concentric and colinear. If there is much stock to be removed, especially on longer parts or those that will be heat treated and brought back to the machine, it can be beneficial to retool centers at some point.

It is never a good idea to center parts that are seated deep in a chuck or collet when you can't see what the far end is doing unless you are confident that your workholding device is in good shape and holding the part true. Centers being within a few thousandths is certainly good enough for most any work but the most fussy. I have seen some chucks in "professional" working shops that hold work WAY farther out than that. And of course, for larger work, that is not even an option in the first place.

 

[guythatbrews]

It's easy to get alignment on rough parts with an HBM and the rotary table.

Agree completely. Sometimes the only and best way. Faced and center drilled many horrible rough forged bars on HMC.

We relied on good setup and machining for alignment and never confirmed on the part. And never a problem, at least that we knew of.

 

[henrya]

I think y’all are missing an important point - what feature are you trying to preserve or measure from?

The top guys who do high end rifle barrel work generally use a spider on both ends of the spindle bore. The barrel is indicated in at the bore at both ends before any work starts. Of course, other than the happy accident, the bore isn’t perfect (it snakes and won’t be perfectly parallel or collinear to the outside) but again at the high end its really really good.

The set up uses a gimbal type of holding the barrel at each end (sometimes just a piece of copper wire wrapped around the barrel to let it pivot in the jaws) to allow for any misalignment that would tend to bend the work piece. At that point you have your part suspended stress free and ready to start cutting. That first involves facing the end. For some it means next cutting a center by boring.
This is as good as your machine and your willingness to measure and adjust. Typically, .0001“ is the level of measurement used to get the part running as true to the bore as possible. This has to be checked repeatedly during the work or you don’t know what you are cutting.

For a part where the outside features are rough, cutting a center then a journal then recutting the center by boring is how to make it as good as it will get.

 

[eKretz]

I think y’all are missing an important point - what feature are you trying to preserve or measure from?

The top guys who do high end rifle barrel work generally use a spider on both ends of the spindle bore. The barrel is indicated in at the bore at both ends before any work starts. Of course, other than the happy accident, the bore isn’t perfect (it snakes and won’t be perfectly parallel or collinear to the outside) but again at the high end its really really good.

The set up uses a gimbal type of holding the barrel at each end (sometimes just a piece of copper wire wrapped around the barrel to let it pivot in the jaws) to allow for any misalignment that would tend to bend the work piece. At that point you have your part suspended stress free and ready to start cutting. That first involves facing the end. For some it means next cutting a center by boring.
This is as good as your machine and your willingness to measure and adjust. Typically, .0001“ is the level of measurement used to get the part running as true to the bore as possible.

For a part where the outside features are rough, cutting a center then a journal then recutting the center by boring is how to make it as good as it will get.

No, for a rough part, you are aiming for centerline. That's why you need to cut the spots to create one to aim for first. On a lathe using that method, it's pretty difficult to miss center.

 

[henrya]

I’ll accept that you are probably aiming for a centerline but you will also need some feature beyond the centers unless the entire work is done between your perfect centers. We’re kinda saying the same thing, maybe.