What's new
What's new

Help turning long SS square bar

Froneck

Titanium
Joined
Dec 4, 2010
Location
McClure, PA 17059
My customer uses 1-1/2" SS 304 square shafts to drive conveyer belt sprockets. OAL 47", one side is turned 1.250 X 5" (no problem) other side same diameter X 15" (Problem) I set them up in P&W 16" newer Model C lathe, I chuck the square in 12" 4 jaw chuck, measure back 15" and indicate the flats of the square best I can due to material is not perfect square. Center drill the end. Cut to about 1-3/8 diameter with CNMG6** insert. No problem but when I remove the center stresses in the shaft move it sometime over 1/8" or more. Sometime the stress is so great it moves so that the square side I just indicated is running quite a bit out. (1/4" TIR.)
The lathe has the power to cut the square round in one cut and done on the short side. But on the long side I take a light .050" and when I notice the square running out remove the shaft. Chuck the the 15" side in another 24" lathe and straighten until it runs reasonably well. Re-chuck in P&W take another few cuts and have to straighten again. Seems like every time I straighten then re-cut it returns to where it was prior to straightening. After getting to 1-3/8" OD I put the shaft between centers and find it running out quite a bit and straighten again. taking .010" total cuts, I get 2 passes before the shaft about 2" from the live center is running out. Cutting with DNMG 55° insert. Again every time I straighten the shaft it returns back to the same amount of run-out after .020 removal. Thought now the square shaft is running out due to holding between centers. Is there anything I can do to lessen the run-out without sending the shafts for stress relieving?? Being it is a poultry plant shaft requirement is NOW! They need it yesterday! The Idler shafts 137" length cut round 1.250" X 5" both sides are easily done with no need to straighten.
 
Do you think a steady rest in the center of the bar would help. I know the bar is square so you would have to make a round bushing with a square hole in it and set screws on the sides of the square.
 
I should suggest to change design. To have square discs or the like on a round bar, secured by through dowel pins, nice and classical.
 
How about a simple box tool with the rollers set on the turned surface ("behind") ?

Granted, you will need to use you existing turning tool to get the starting surface for a short distance (to get the rollers a place to ride), but the box tool should eat that material off in qty (1) pass.
 
This sort of issue is why you have to be damn careful quoting work. Nothing sucks more that having stock that wants to move as you cut it.

[OK, lots of things suck worse, but this is up there]
 
I have been asked to make similar shafts, but I chose to use RHS (box section steel) and fit stub axles in each end.

I machined each stub axle from round bar, finished to size, key way if necessary, then drove them into the RHS. Make a groove in the stub axle if the RHS weld interferes. Weld them into place.

The main axle (RHS) needed no machining, just cut to length. I did drill some holes in the RHS so the stubs could be plug-welded in several places down their length. The main weld was on the end of the RHS.

I had no problems with run-out, made sure the stubs were tight and went a good distance into the RHS (approx. 5 - 6").

These were for plastic modular belts (slat belts) which use square-bore plastic sprockets. For those who haven't seen them, they have several sprockets down the length of the shaft, from memory they are free to float on the square shaft. FWIW, it was 40mm square.


Regarding the run-out of the square 304 shaft during turning, I have had a lot of trouble recently when machining steel pipe for conveyor belt drums. The local steel mill shut down its pipe production line last year (too many cheap imports I guess). The NZ-made pipe machined beautifully, no stress relief when machining the OD full length (crowning for belt).

I then discovered that all steel pipe is not the same, the worst was from China, it could not be used. No matter how many cuts were taken, it would distort in a new direction. Fortunately we found an alternative supply.

What I am saying is, the problem with run-out might be something you can't fix, just try a different supplier and specify a trustworthy steel manufacturer. Make a note of the manufacturer if it machines well.
 
Sounds to me like a steady can only help, ..make up a catspaw sleeve (with 2 screws per flat) and set that true on the SQ bar for the steady to run on.

BOTOH what you describe sounds a lot like crap material, .........or possibly ??? too much tail stock pressure
 
I'm not buying the material, customer supplied. I have gotten 1-1/2" square stock from McMaster but never turned one side 15" long before. McMaster material has a fine "sandy" surface finish with corners slightly rounded. This material as very smooth "cold rolled" finish with almost no radius on the corners. No not too much tail stock pressure when roughing where all the distortion happens, Check it often so as any heat don't expand the bar. Bar does not get very hot, I can put my hand on it. I told them to get me another bar so I can experiment, I picked up some large bearings at HGR cheap a while ago NIB. 75mm bore double roller bearing, probably ABEC1 but good enough, I did get a few others I'll have to hunt for.
I plan to use the steady rest idea by garyphansen and make an insert for the bearing. I don't think it will stop the distortion but the bow will not cause the square to be cut off center. After roughing bending it back to as straight as possible then ends between centers so that both rounds are concentric. As it is now the two round sides are concentric, the square runs out, they are not very concerned that the square runs out but I am. I know I can't remove all the square run-out heck the shaft is not very square but I want it better than I'm getting however at least the ends run true in the bearings and the hollow shaft gear box.
 
Can you cut 15" off and drill the problem end to thread in the round? Maybe something like a 1" or 1 1/8" thread? Then the round end would just screw in straight.

I missed this one, that seems a good idea, though I am not sure the thread would give you a true running shaft.

I would go for a plain bore in the square 304 and use several set screws, well-seated into the shaft. Use say a 1" reamer in the square shaft and then turn the extended end of the "15 inch" shaft to be a good tight fit in the reamed bore.

15 inch of turned shaft seems an odd design, far too much sticking out for a bearing and shaft mounted gearbox. Watch that gearbox wobble.

I have my doubts a steady is going to help if the square shaft is stress relieving every time you take a cut.
 
I missed this one, that seems a good idea, though I am not sure the thread would give you a true running shaft.

I would go for a plain bore in the square 304 and use several set screws, well-seated into the shaft. Use say a 1" reamer in the square shaft and then turn the extended end of the "15 inch" shaft to be a good tight fit in the reamed bore.

15 inch of turned shaft seems an odd design, far too much sticking out for a bearing and shaft mounted gearbox. Watch that gearbox wobble.

I have my doubts a steady is going to help if the square shaft is stress relieving every time you take a cut.


It would need to be better than just a threaded hole in the sq bar. One way would to ream the end of the hole so the shaft would center nicely, like a thread on lathe chucks pilot. Kind of like your suggestion, just with threads in the bottom of the hole instead of set screws. Another way might be too have a flange on the end of the round bar, could even be like a shoulder bolt threads, might be ok with his 1 1/4 shaft to use a 1" fine thread and a 1/8" shoulder all around to square it up.
 
The square stainless material, being customer-supplied, is something the OP is stuck with and has to work with. I've read the other posts, and an idea came to me:

-set the bar stock up in the 4 jaw chuck, indicate it, and drill/ream 1" diameter x perhaps 2-3" deep at each end.

-turn about 2"- 2 3/4" long x 1" nominal diameter on pieces of 1 3/8" diameter stainless round stock long enough for each of the projecting sections. I'd
machine the 1" diameter for a light drive fit in the 1" reamed end holes in the square bar.

-machine an extra 3/8" at 1" diameter so that when the 1 3/8" round bar is driven into the end of the square bar, there is about 3/8" width of 1" diameter exposed.

-GTAW (TIG) weld the 1" diameter are to the end of the square bar. Use small sections of weld, and 'quarter' the weld to try to balance weld stresses. Keep going
until you have built up a fillet weld the meets the 1 3/8" diameter round bar.

-chuck the work in the lathe and re-indicate off the square section. Turn the 1 3/8" round bar section to final diameter.

The idea here is to avoid what is known as 'opening locked in stresses' in the stainless square bar. A reamed hole in each end of the square bar should not produce much warpage as it is symmetrical about the centerline and not too large an area. The 1 3/8" diameter stainless steel 'spud' has an extra 1/8" of stock to allow for truing and turning to final diameter.
 
The first thing I will say is cutting cold rolled is always a formula for parts twisting out of shape, but it is what it is. I was also going to suggest making a round bushing that you can clamp on the square stock and run in a steady rest.
 
I really don't want to use a threaded round into the square style. As they are the shafts are quite durable in a place where the most used tool in the maintenance tool box is the hammer. Plus Keyed long shaft I'm not sure which direction it rotates, it will depend on which side the gearbox on the conveyor it's mounted plus the possibility that mounting side might change. On previous shafts the keyed gearbox side shaft was only 10" long so I didn't have much problems but I'm told some type or transducer is also to be added to the assembly requiring additional length. Plus connecting 1-1/4" shaft to 1-1/2 square don't leave much wall thickness even if coupling diameter is 1".
I might try using the 24" Hendey lathe, it has 20" 4 jaw chuck so jaw and chuck are quite a bit longer than the 12" on the P&W. I don't think my spindle bore is 2-1/8" to fit 1-1/2" square shaft, I think I did check some time ago nor do I know what the spindle taper in on the 24" but tail stock is 5MT. I might be able to hold more than 10" in the chuck of the 27-1/2" of 1-1/2" square shaft. Plus mount another large 4 jaw chuck in the 20" and be able to hold quite a bit more than 10" so that the steady rest is not needed. In addition to that make a live tail stock support using a 4 jaw chuck that way I don't have center hole that I'm stuck with after the shaft bends. After each cut adjust the support chuck to accept the bend after most of the 15" is roughed out and bend will allow finishing machining to 1-1/4" diameter then drill center hole and mount shaft between centers.
 
On previous shafts the keyed gearbox side shaft was only 10" long so I didn't have much problems but I'm told some type or transducer is also to be added to the assembly requiring additional length.

Froneck,
Could you suggest they stay with the previous 10" long shaft and consider some other method of mounting the transducer? E.g. supply them a separate short length of shaft, drilled through, so you could bolt it onto the (tapped) end of the 10" long shaft? Presumably the transducer/encoder would have no load on it.
 
At one time I was the Engineer at the company, I left. Now they have none so making suggestions to most there is like telling them their wrong! Most have the opinion that what they thought of is correct and want management to continue thinking that way! I do make suggestions but most are not used.
 








 
Back
Top