Horizontal Milling Accessories for a Hardinge UM Milling Machine

[SightlessSeer]

A while ago I picked up a Hardinge UM Milling machine. It came with the Bridgeport H head on an arbor that mounts in place of the overarm to give vertical milling capabilities. While this is nice to have, I would also like to be able to do some horizontal milling on the machine, but I didn't get the original overarm, arbor support, or any milling arbors with the mill when I bought it. I'm planning to build my own versions of each of these. So far I have finished making the overarm and am planning out what the arbor support should look like.

The Hardinge UM's original overarm is about 32" long and about 2" in diameter (I measured the arbor for my bridgeport head at 1.999"). My plan to build a replacement was relatively trivial. I needed a steel shaft about 32" long that's either very close to 2" in diameter, or is larger than that so I could turn it down to fit. I managed to acquire just such a shaft off of a local HM member who, I also work with. The shaft I picked up was 2.25" diameter and 34" long. This is just short enough to turn on my 54" bed Atlas 10F.

Spoiler: Making the Overarm

Here you can see when I first put the overarm to into the lathe. The first order of business was to face both ends and drill centers into them. This is also when I started to think that maybe I had bitten off more than I could chew...

A difference perspective of the previous photo. You can see that I haven't even closed the jaws of the steady rest. I was definitely still in the "can I do this?" phase here.

With the tail stock removed, I was able to face the ends without issue

Facing was followed by drilling a center. One of the most frustrating parts of working with stock right at the limits of my machine's capabilities is that I had to repeatedly remove and reinstall the tailstock. Nearly every operation required either it's fitting, or removal to make space.

After flipping the stock and doing the same to the other side, I have the piece suspended between centers. It's not super obvious in this photo, but the tailstock is overhanging the end of the bed by about 1/4". The only way that I could work on a bigger piece would be by switching to a dead center and even then I'm buying about an inch. Also notice that I don't have a drive dog here. My largest drive dog is suitable for about 1.5" diameter stock. You'll see in the next photo what I came up with to solve that issue.

An old clothes line hammered flat, plus a couple holes, and a 1/4-20 threaded rod was enough drive dog to get me going. Here you can see the first appearance of my insurance board. When I was changing configurations of the shaft I tried to keep one of these under either end incase of incident. Fortunately nothing happened, but you'll probably see these appear and disappear in the coming photos.

Despite the difficulties I was able to turn the shaft. Here I'm making about a 0.06070+60
*2
p cut with a feed rate of 0.00187" at 28 RPM. While taking deep cuts I wasn't able to increase the speed any before I got insane chatter, but eventually I was able to increase the feed rate to 0.0087". This setup still chattered, but was good enough for roughing out the first couple hundred thousandths.

Pictured here is my gorgeous cutting oil feed. Every few minutes I would squirt some oil onto the brush and it kept the cutter lubed without further intervention. The milk crate in the background was my improvised stool. I'll let you do the math yourself, but a ~26" long cut at 28 RPM and 0.0087" feed rate takes a long time (this picture is about when I upped the feed rate).

Here we can see the first cut completed

After the first pass, I decided to add the steady rest in hopes that is would help limit the chatter. This definitely helped, but for the second pass, which was about 0.070" there was still plenty of chatter, but it was manageable. Also, keep in mind that I dialed 0.070" off the radius, but in practice I was taking off 0.005-0.010" less than that. I'm guessing that this was mostly due to flex in the shaft.

Now we're getting much closer to final dimension, so I've changed to a different tool to start creeping up on it.

When I got really close to 2.000" I changed to a vertical shear tool and I took the time to map out the diameter. I ended up turning the shaft down in ~2" long sections to make sure that I was between 2.000" and 1.999" across the entire length.

You can see the different finishes on the different finished sections of the shaft. I was more concerned with dimensions than surface finish, so I can live with that. In any case, the finish is very good across the length, but you can see some slight chatter in a couple spots. The chatter marks are so fine that you cannot feel them with your hand.

A closer view of the finishing passes

Finally, a nice big chamfer to finish off the first end. I don't have any pictures of it, but I stripped out the t-nut in my compound when I changed the angle of the tool post to make this cut, so I had a short side project to cobble together a new one.

I repeated the same process for the ~8" remaining on the other side, although this is the only picture that I have of it. I used some copper sheeting to protect the finished surface underneath the "drive dog" and the whole process went much faster on the shorter section of shaft.

I've very pleased with the finished overarm. The finish is good enough for my standards (the only ones that matter) and it fit perfectly into the mill on the first try.

Spoiler: Finished Overarm

Next order of business is the arbor support. I'm not exactly sure what it's going to look like yet, or how I'm going to make it, but here are my thoughts

I'd like to bore out the hole in it for the overarm on the lathe, but I think that it's slightly too small for the job, unless I line bore it.

For the arbor bearing I'm thinking that a piece of bronze pressed into the main support body and then bored or drilled or reamed into a slight taper would work nicely as a bearing along with a simple cross drilled hole for oil delivery. I'm thinking that by having the bearing and arbor tapered I can take up any slop or wear between the two by pulling the overarm in a bit more. Using a Morse Taper seems like a good way to avoid scratching my head in a couple years and asking "was that bearing 2° or 3°?" when I'm trying to make another arbor. An included bonus is that I could use standard tooling to make a shallow taper. I'm not sure what an MT3 reamer costs, but it's definitely cheaper than a custom one.

If any of you have done similar projects please give me any suggestions that you have. I'm just an idiot who doesn't know his limits, so I'm likely to dream up more difficult ways to do trivial things. If a tapered bearing is completely unnecessary I'd love to hear that, but the original Hardinge arbors and supports are tapered, so my knee jerk is that they did it that way for a reason. I also still need to figure out exact shapes and fixturing for what is likely to be a very irregular shape.

 

[brino]

That's certainly using all of the Atlas bed length. Nicely done!

Are you suggesting a Morse taper for the arbor bushing/plain bearing? (I may be misunderstanding this.... still)
I believe the Morse taper specs were designed to make tapers for power drive and are meant to lock together.
I'd be concerned about trying to use them for a bearing surface where you want rotational sliding fit.

A simple 60 deg. centre like a lathe tailstock support would work and be useful for both turning the arbor on the lathe and then using it on the horizontal mill.

Brian

 

[SightlessSeer]

From what I've seen, the original arbors have a tapered interface between the arbor and the arbor support. In the pic below, the end of the arbor looks like a slight taper. I also found a drawing for a Hardinge arbor in the Hardinge Mills group in groups.io. It calls out that taper as 5.623° (2.8115° from center). That's more than the ~2.9° degrees of MT3, but still very shallow. I haven't seen the receptacle in the OEM arbor support before, but unless they have something special there (I'm thinking an integrated roller bearing or something) I don't think the extra 3° are going to make that much of a difference.

A 60° center might be a good option too. Are you thinking of having the male 60° point on the arbor or the support? I've seen older mills with 60° dead centers on their arbor supports, but it always struck me as a bit of a primitive way to do it.



Here's a preliminary model of what I have in mind. The yellowish part would be a pressed in Bronze insert that is then drilled/reamed to MT3.

 

[extropic]

Congratulations on completing the overarm.
Next time, take a look at linear bearing shaft for about $200. 60 HRC case hardened, straight and smooth as a baby's bottom.
Or, TP&G hydraulic cylinder shaft, often chrome plated too. I don't know a price comparison for 2" TP&G.

I'll throw out a couple of design concept ideas that you may find food for thought. Just thinking out loud here.
It sounds like you need to make your arbor(s) also.
For bearings, I'm thinking use a pair of tapered roller bearings rather than a tapered bushing/shaft. A Morse taper is NOT a self releasing taper. Too fiddly for a rotating element. Tapered roller bearings and seals can be had at any auto parts supplier. No sweat.
Use a threaded stud on the end of the arbor to pass through the bearings and a nut w/ lock nut to set the preload.
For the support arm itself, I'm thinking a three piece design. The top piece is a clamping cap using four fasteners.
The bottom piece houses the arbor support bearings and extends upward toward the overarm.
The center piece clamps to the bottom of the overarm and bolts to the lower piece.
The reason I'm suggesting the center and bottom pieces are separate is so that the exact distance between the overarm and the spindle/arbor is adjustable. The joint between the two pieces could be secured by threaded fasteners and could be fixed using tapered pins, after the length adjustment is completed.

However you decide to proceed, good luck.

I always wonder how the critical attachments get separated from their rightful home. Actually, I know how it happens. BOZO does it.

 

[brino]

A 60° center might be a good option too. Are you thinking of having the male 60° point on the arbor or the support? I've seen older mills with 60° dead centers on their arbor supports, but it always struck me as a bit of a primitive way to do it.

60 degree hole on arbor shaft and 60 degree point on the support.
Actually, then you could use a lathe live centre in the overarm support if it was MT2/MT3.

Primitive?
Maybe, but I think that's what my old Cincinnati has.... pretty sure I have used it once at least and it seemed to work,

I think I have an MT-3 reamer you could borrow.

Brian

 

[Weldingrod1]

I would make your arbor ends straight. Make the -outside- of the bronze bushing tapered so you can adjust the ID.
Though, with the adjustability of the over arm, tapered is do-able.

You want more like R8 or 5C taper; self releasing.

Sent from my SM-S911U using Tapatalk

 

[SightlessSeer]

I think that I'm going to try my idea of a pressed in bronze bushing in the support with a MT3 taper on the inside. If that fails (i.e. it ends up too grabby), I like Brino's idea of throwing a live center in the MT3 and calling it a day. It's exactly my type of sacrilege.

If somehow neither of those ideas work I can always press in another piece of bronze and try again.

Regarding extropic's suggestion to make the distance between the spindle and overarm adjustable, I don't think this is necessary since I'm planning on boring the hole for the spindle in situ on the mill. I was planning to put a center drill in the spindle and free hand push the arbor back to make a center mark then use that for locating the hole in a different setup, but this video by Workshopfriend builds on that same idea by holding the arbor support in a vice while it is also hanging off of the overarm and using the Y axis to feed it into a boring head. I think this will work just fine.

I spent about 6 hours after work today squaring up the plate that I'm planning to cut the arbor support body out of. I didn't quite get it finished, but it shouldn't take too much more time. It occurred to me that a slab cutter in a horizontal milling setup would have been perfect for the job ... if only I had one. I took some photos along the way, but it's late and I can't be bothered to download them from my phone tonight. I'll upload them either when I have a chance, or once I have made more progress.

 

[SightlessSeer]

Today I "finished" the arbor support. I still want to do some finishing work to clean up some of the tool marks (I've been thinking about jeweling the front face. It's front and center every time I use the mill plus it might be fun) and corner radii, but functionally it is done.

I went with a Morse taper 3 live center for the support bearing (thank you to @brino for loaning me both his boring head and an MT3 reamer) and a simple slit clamp for mounting to the overarm. The clamp is tightened with a shop made pair of 3/8" square headed 3/8-16 bolts (they definitely didn't take me 3 hours to make... ).

I don't have a whole lot else to say about this at the moment. Time will tell how well it works once I make some arbors and get a couple of cutters, but it certainly looks the part. I'll add more details on the build process in the photo dump. Might have to split it across a few posts depending how many pictures HM allows.

Spoiler: Photo Dump

This is the drawing that I was working off of. The actual finished dimensions varied a bit from this because a lot of the features are purely aesthetic, so instead of worrying about hitting an exact number on everything I tried to make sure there were nice surface transitions and decent surface finishes. The only really critical dimension here is the overarm bore diameter. The other bore was reamed to size, so as long as I could start the reamer more or less anything would do. That being said I ended up making the pre-reamed size a good bit bigger than shown here, so the angles of the sides and the diameter of the end round at the bottom changed along with it. The second critical dimension is the distance between the two bored holes. In the drawing I have it marked as 4.75", but that was my best guess based of eyeballing it with a pair of calipers. The hole for the Morse Taper was bored with the support hanging off of the overarm. This is what set the actual distance between the bores.

Here we see the lovely process of cleaning up the random plate of steel that you bought from the local scrap yard. This piece is far larger than needed for this job, but I was tired of fighting my way through this crusty scale every time I was to do a project, so I cleaned up as much of the piece as I could.

I thought my clamping solution here was pretty clever with the plunged pockets in the side.



This endmill was pretty thoroughly trashed after cleaning off all of the scale from this piece, but I think it was well used. I won't be contending with that again any time soon.

I cleaned up the rough tool marks the now-destroyed endmill left with a couple fly cuts.


Finally, I cut off the section that I couldn't reach with my mill's table travel on the bandsaw and cut out a 3" x 8" piece for the overarm.

First order of business is to drill a couple of clearance holes for the boring head to come into later and a pair of holes for the transition between the clamping feature and the rounded section around the overarm bore.

I don't actually remember why I have the workpiece setup on 123 blocks here. I think I might have been flycutting the face to make it parallel along it's length? I'm not sure. Parallelism doesn't even matter in that dimension, but it's still probably a good idea incase I try to use the faces as reference surfaces later.


The boring part begins (yes, I do think I'm funny)
The keen eyed among you will notice that I partially drilled the top hole in the wrong place (read the wrong dimension on the drawing). The partial hole was about 0.150" deep 3/8" in diameter, and centered on where the edge of the correct 0.5" diameter hole should have been. I ended up fixing this by filling it in with a bit of weld then re-machining it. If you look for the spot it is still visible on the finished support, but it's not too obvious.

Before I started boring, I turned up several aluminum disks to use as gauges. This the final one for the overarm bore. I might have been a bit optimistic on the precision of it seeing as I'm in a non-climate-controlled garage and turned the disk several days in advance of using them.

In any case, the 2" disk would go in both sides and get tight in the middle of the bore. I think this was because the disks are only 1/4" thick, so they have a tendency to get misaligned and jam. I don't think there is a significant difference in the diameter of the bore across its 2" depth.

The overarm is a very snug fit in its bore. At the end of the 34" overarm there was maybe 1/8" of wiggle. If I cared to I could have probably use that to do some math and find the clearance between the arm and the bore, but I don't care enough. suffice it to say that it was tight. When I was showing this picture around at work, one comment I received was "you really trust those clamps".

On prominent display her (in addition to the subject of this thread) is my makeshift x axis lock. Surprisingly, Hardinge didn't see fit to outfit the early UMs with any form of x lock. When they eventually did on the later machines it was not the type which presses on the gib (like the y and z axis have, but a clamp between the t slot on the front of the table and the saddle. This strikes me as a very odd choice for what is touted as a tool room mill.

[Continued below]

 

[SightlessSeer]

Spoiler: Photo Dump Continued

Before boring the hole for the Morse Taper, I roughly levelled the mill and then setup the arbor support at the same angle. This won't make any functional difference, but if it was too far off I it would have looked weired


While boring out the second hole I feed the arbor support into the boring bet by clamping it in a vice and feeding the able in the y axis.



Towards the end of the boring I had to move the vice to one side to avoid it hitting the boring head.

I don't have a good photo of it, but this is after I plugged that partial hole with the well and I'm setting up to fly cut it flush again. I'm indicating off of the back face here because that is the one I had been referencing up to this point, so I thought it best to try to keep it as the reference.


Here you can see the repair. It's very hard to spot in person. the outline that you can see in the photos is really a slight difference in how it reflects at the edge of the weld. There is a tiny void right at the edge of it, but it doesn't really show up in photos.



With that mistake finished, back to work. Here I'm roughing out the faces of the overarm clamping feature. I cut the slots by stepping over 0.100" and plunging the mill through. Once I was had done that out to the edge I came back in and cleaned up the slot with sides of the mill.


I wasn't happy with the finish that the endmill had left on the clamp. Specifically, the transition from the flat into the radius didn't line up very nicely, so I came in with this bur, which I accidentally bought in a bag of assorted endmills. It left a phenomenal finish and since it was ball nosed also cut a perfect radius.



I didn't take many photos of this part, but I drilled, counterbored, and tapped the holes for the clamping bolts on the drill press. Drilling without a movable quill is pretty tedious and the exact location of the holes didn't matter much. I took some time laying them out and punching them before going over to the drill press and they ended up where they needed to be. On display here are my square headed clamping bolts. I'm very pleased with how these turned out. I turned them out of some 1/2" round bar and threaded them on my lathe. After that I took them over to the mill and used a spin indexer to cut the flats. Nothing ground breaking here, just some nice looking bolts.
[/spoiler

 

[SightlessSeer]

Spoiler: Still More Photos to Dump

The next order of business was to rough out the final profile with the band saw. This took about 3 hours.




With the shape roughed out I cut the tapered sides next. To get these parallel to the edges of the bores I used gauge pins against the back of the table and the inner edge of the bores before clamping the piece down. By this method, I also compensated for a 0.013" difference in thickness between the left and right webs of the overarm clamp (when I was cleaning up the surface of the clamp with the burr I had to go a bit deeper on one side than the other to get it to fully clean up). The effect of this is that the tapers are not at quite the same angle left and right. This way the round around the arbor support bore can have a constant width and still transition into both sides nicely.

The endmill I'm using here had a corner go dull on me as I was doing the finishing pass, which was really unfortunate. The finish didn't turn out bad in spite of that, but it wasn't as good as it had been before that.


Now we're up to what I was doing today. This is my makeshift corner rounding setup. Bolted to the large aluminum plate are the test disks that I used while boring the two holes. These make sure that the center of the bore doesn't move while I rotate the piece through the range needed for a given round. The worked pretty well, but I wish I could have set it up in a such a way as to side mill instead of end mill. I really like the finish that side milling gives. It's a shame that I wasn't able to use it for more faces during this project.

Showing off the bolted down 0.825" disk that I'm pivoting around




The results of the fixture aren't too bad. Definitely not as nice as a rotary table would have done, but I don't have one, so I made do.


Rounding off the last corner. The smaller pieces of aluminum here were used as spacers to keep the arbor support from tilting when it wasn't fully resting on the larger plate.


Here I've cut the slot for the clamp and it clamps nicely. Also on display are some scratches on the fly cut surface from when I was rough cutting the profile on the bandsaw.

Last thing to do is ream the arbor bore to take the MT3 live center. This was a really slow process. You'll notice the jerry rigged setup for feeding the support. I specifically didn't want to use the vice setup that I had before because I wanted to let the arbor support pivot around the overarm to follow the reamer (in my head this is an ersatz floating reamer holder). I'm not sure how much of a difference this would have made, but it seemed to work. The taper turned out nicely.


Obviously there isn't much point to this setup other than to show off my new toy. Everything seems like it will work for what I need. Next up is to start making cutter arbors, buy some horizontal cutters, and make a drive collar (to drive the arbors from the drive dogs on the spindle instead of from the 5C collet.