Would a shrink fit be feasible?
Printable View
Hi Michael
my 'get around' for this problem is to throw away any tungsten carbide tooling and feed and speed charts then grab a piece of HSS and sharpen it well. Grab an oil stone and just wipe its nose to remove the sharp point (not too much) then slow the lathe almost to its lowest speed.
I had this exact problem at the local sawmill once when I was asked to machine an infeed roller. They couldn't get rid of the chatter. The machinist guy there couldn't believe that I wasn't using the recommended surface speed for the material. :rolleyes:
I think you might also find that the gear on one end may work as a damper and remove some of the harmonics.
Using this method you could make a small undercut in the middle section of the mandrel and machine a transition fit for the washer (bore and shoulder) and shoulder for the other end.
Phil
Sometimes you just have to. I've quite a collection of mandrels used once.....
If you make the mandrel with a waist, as I would, you only need good bearing on say 20mm each end of it. That makes it easier for the workpiece to slide on. As Phil says, a nut to apply pressure will stop it moving. The old way of doing it was a very slow tapered mandrel and an arbor press. You've got a cylindrical grinder, consider it a challenge.....
Look at it this way. You'll have a nice mandrel ready to machine down to the next smaller size next time you need one. Which will likely be never but - shrug.
PDW
Hi Michael, Guys,
I'm in agreement with the mandrel ! I would make one with a close sliding fit and then use a trapped, squashed "O" ring on one end to lock the two together. You could also mount it between centres done like this.
- No heat treatment required
- Shrink fit would be nice but then getting it off again...
I was thinking of the old fashioned tapered mandrels, but with a 4 1/2" long part, that may not work too well, apart from not having a tapered mandrel that diameter.
So far the plan seems to be large diameter mandrel, flange on one end with nut on the other. I have some 2" hollow bar that will probably do. I walked past a limits and fits chart today, and clearance between the parts is only about a thou, so my O-ring plan may not work that well. Have to think of a way to expand the mandrel slightly (I already have 3 methods I could use, all being highly over complicated...)
I agree about speed Phil - needs to be right down. While my big carbide tipped boring bar is a thing of wonder and the normal tool of choice, I think I will end up making a HSS one for this job - although as the bore is large I might end up with something 1" or 1 1/4" in diameter to push the rigidity factor up.
Michael
Yes rigidity is your friend here. I've got a 32mm TCT boring bar I needed for making new rudder bushings for a 45 tonne steel motor-sailer - similar dimensions to what you're talking about length-wise but 75mm bore. I have a few big HSS boring bars as well that run 38mm diameter but can get a better finish with carbide on the lathe (the HSS is great on the Kearns HBM). Then again the Monarch CY lathe has the mass & rigidity your CVA and my Chipmaster lacks when you get into the bigger parts (and the same size/weight & low top speed makes it a PITA for small stuff).
With a length-diameter ratio of only a bit over 2 I still think doing the OD then dialing it in using a good 4 jaw and boring out the ID would be fine. Especially if you're going to insert bronze bearings and leave the centre part a slightly larger diameter, which is what I'd do. There's always a number of ways to skin this sort of cat.
PDW
I might not have interpreted you requirement correctly, but something I've seen in the past seems like it might be an option.
The machinists at work would use a steep taper mandrel (taper like a pipe live center) to finish turn the OD of bushes.
One taper was part of the mandrel, the other screwed on onto the mandrel on a fine pitch thread. I'm picking the one mandrel could accommodate a range of sizes and would be self centering.
Edit: photo might make more sense.
https://uploads.tapatalk-cdn.com/201...09ca9baf90.jpg
Steve
Steve, that's probably similar to what Pipeclay was talking about. While worth remembering, I'm not sure it would work well enough for this job. The cones would need to be sharper as otherwise with a long part like this I would be concerned that it could shift.
Not sure it adds much at this stage (probably wouldn't have done before either - I thought I described the basic geometry pretty well)
Old part on the right (with mangled gear), soon to be part on the left.
Attachment 379932
The key thing for me is that I need a centre on at least the gear end of the blank so I can index around the part axis.
Just about all the methods talked about so far would work. You can see how it lends itself to being bored on the mill. Similarly I could turn the outside then bore the inside; my concern would then be accurately locating a centre to get the gear concentric.
From experience I know that I need positive location when cutting the gear at least. I'd take a chance on friction if they were straight teeth but these are helical (which reminds me - anyone got a use for a 12DP gear, 86.25 teeth :B).
That's a keyway hanging on the side too. Not sure what it is keyed to as that wall is not very thick at all (OD of that section is actually 2.125", so my 1/8" earlier on is really 3/16" - a whole 50% more!)
So - bore out with one end in a 4 jaw and the other in a fixed steady. Once to size, set up on a parallel mandrel, around 1.75" diameter, flange on one end, washer and nut on the other. Mandrel on centres (note to self - check alignment), turn the outside + gear blank to size. Transfer to mill, cut the teeth, back to the lathe for the oil grooves and some clean up. Easy really...
Michael
That blank is over-length. Leave it that way until you've finished everything else and cut a notch in the 'spare' bit or drill a hole in it to lock the work piece to the arbor so as to positively prevent rotation. A grub screw would be fine I'd think.
I've the odd spur gear with an unusual tooth configuration too.....
PDW
So I ended up making a mandrel out of a solid lump of steel to run between centres. Welded a 'washer' onto on one end and then cut a 1" UNF thread on the other to take a couple of nuts. As best I could measure the mandrel was 1.7495"; with a bore of 1.7500" (likewise), it was a tap on fit. As an aside, my lathe dog is a quick as any thing home made jobby and looks a bit that way. Has anyone got a good design for one?
Attachment 381017Attachment 381019
And the end result...
Attachment 381018
Michael
In this case you would have been camped here for several days...
Michael