View Full Version : steel chioce for shaft making

01-04-2010, 12:45 AM
What type of steel would you guys make these shafts or arbors out of. They are 12 inches long and .570 dia at the largest and .430 at the smallest. I need them to be pretty stiff and have as little runout on the ends as possible. I made a 1/2 inch one out of stainless but was a real tough one to thread. The very ends that are cut down is where the bearing sits. These are used for balancing snowmobile clutches. I have even had trouble with the tapers not being perfect. I used a hydrulic cylinder shaft for the right O.D and left the chrome plating on because it was the right size, drilled a smaller hole and then bored to size for a nice sliding fit. When I turn the shaft the taper has a wobble to it which in turns make the clutch stop in the same place every time . I can spin the clutch on the shaft and keep the taper in place and respin and it stops at the same place. It always stpps on the high side. I did a dial test and found about 8-10 thousands variation on the taper flat section . Tested the ends of the shafts and they were 1-2 thousands. http://i144.photobucket.com/albums/r171/newfy_2006/ClutchBalancerpics2001.jpg

01-04-2010, 01:56 AM
If you want it real close why not use drill rod cut a bit large, harden and grind to size?

01-04-2010, 02:20 AM
For that application it makes no real difference what you use. All steels have the same stiffness regardless of hardness or treatment. Hardening such a part will serve to warp it. The only other purpose hardening will serve is to prevent wear in use. If that is an issue then turn a shaft from hot rolled steel and case harden it.

01-04-2010, 02:26 AM
Machinery Hand Book recommended 4140 as the best shafting material.

01-04-2010, 02:28 AM
Why not use tgp then- turned ground polished. Without costing an arm and a leg, that's about as close as you'll get- from what I've been told anyway.

01-04-2010, 02:44 AM
Another thought,
I would turn the shafts between centers using a lathe dog to drive them. This would give you a more concentric shaft.
It would also be nice if you could rig up 2 live centers on your balance fixture.
Little Machine shop has tail stock quills with a # 2 MT for mini lathes you could use 2 of these to hold the centers.

01-04-2010, 02:55 AM
What type of steel would you guys make these shafts or arbors out of?
A straight piece, obviously. :D

But seriously, exactly what is the problem? You state you have "wobble" but that isn't caused by the material you have used, it's caused by either a bent arbour or by failure to machine concentrically. If you machined the taper angle and bored the hole in one operation, they shouldn't be the problem. When you turned the ends of the shaft down, did you use a four jaw chuck and DTI to center the shaft? You can also hold one end in a chuck and use a center rest to hold the outer end. http://www.tpub.com/content/engine/14076/img/14076_176_2.jpg That way the turned down end will be concentric to the original surface and is sometimes easier to setup to do and is about the only way to do it if the through hole in the spindle is smaller than your workpiece.
BTW, I think that's a very nice little setup you have for balancing the clutches. :)

01-04-2010, 09:43 AM
1- your not going to get straight steel from a producer.

2- I like 1045 and 4140 but I don't think it really matters as long as it's not trash steel.

3- machine them between centers. I have made mandrels and cones like those and used 1045. Like this for instance.


01-04-2010, 11:20 AM
I can't remember what the shaft material was that I used but it appears to be straight enough for the job it's the taper I seem to be having the most trouble with. What I used was a shaft from a hydrulic cylinder that is 1.5" dia, the size I needed. I cut off 2.5" pieces in a band saw and then chucked them in a 3 jaw to face the ends. When I was done facing the second end it never left the chuck until they were finished. I center drilled the piece and then drilled it undersized with a drill bit in a quick change tool post holder. Next I bored the hole to the size needed . I took small cuts untill the shaft slid into the taper piece still in the chuck. I kept the bore as tight as possible to keep them from tipping on the shaft. After I was satisified with the fit , I turned my compound to 45 degrees and made the taper cuts by hand cranking the slide back and forth until I was near the edge of the bored hole.

This was the prototype that all others were made from. I kept them as close as possible in length to within thousands of a inch in order to keep the weights the same . When they were all said and done I tried them on their proper shafts and they seem to be turning on a angle . Not visible looking at them on the shaft only when checked with a dial indicater or spinning them fast. It seems like the bored hole was drilled at a angle which doesn't seem possible with the methods I was using..

I have no formal training as a machinist, no mentor to guide me other then these web forums. I started out from scratch less then a year ago. Basically everything being made for this clutch balancer needs to be as close to perfect as possible in order for it to work correctly.

This might be the same problem that the factory has but doesn't care. It seems that just about every clutch is out of balance a bit. This is the same set up used at the factory where the clutches being put onto sleds. The picture posted here is what a friend of mine has made and he sent me the measurements and pictures. I will post a picture of mine later with some minor tweaks of my own. Thanks Guys

Chip Soles
01-04-2010, 01:11 PM
Hello, This is my first post, hopefuly it is not screwed up. Nearly all of the cylinder manufacturers use a SAE 1045 which I bieleve has been pre heat treated to the 32-34 Rc range, for rods. Without ht the extra carbon will give you a better shaft. McMaster-Carr will sell in 1&6 foot lengths w/o a saw charge. It is all about the end cost and starting as close to finish as possible.
On shafting, & log mold cores, I have had less warpage w/ more shallow cuts
and raising the rpm, to pickup time, than deeper cuts.

01-04-2010, 04:22 PM
nitris223, When I did them I cut the angled face and flat face while the part was mounted on a mandrel. To be balanced you would also have to cut the OD on the mandrel. The bores on the ones I did were 3/4" I think, I would have to go look at the drawings to be sure.

It looks like you may have a 1/2" bore, is that so? Drilling and boring that hole may not have it end up as straight as you think it should be. The only way to be sure is to drill the hole and then ream or bore it and then mount it on a mandrel, either solid or expandable, and turn all the faces. At that point it will be as concentric with the bore as you will get it.

You will have to mount the mandrel between centers with a dog to drive it for the best results. Also, expandable mandrels are hard to machine on because the work can get loose but solid mandrels seldom if ever slip the work. You can't take heavy cuts on work on a mandrel.

01-04-2010, 06:58 PM

I have one shaft that only has 2 thousands run out which is fine I would think. It seems to be the tapers giving me the most trouble. I am going to change the size and dia of the tapers . Try and make them as small as possible reducing the amount of weight should help some. I will make a mandrel or jig of some sort like what you made and do all the steps from there. I pissed around with the tapers today a bit putting them between 2 centers and got them to within 3 and 5 thousands but it still throws the clutch off when I try and balance it. Everytime I spin the clutch and it stops I mark it with a fine marker. I then respin to double check . Keep one taper in place and rotate the clutch about 20 degrees and respin and the marker point moves. I marked each taper with a high side and low side mark ,also did the same on the clutch high and low. I then tried different combinations of High, Low, high , low. High, low, low, high , etc. Every time I changed something the marker point changed. Nothing was consistant which is what I will need in order to say that where I am going to remove material is in fact the right place.

I might be after something impossible because there are to many points that can affect the outcome. From the shaft, the bearings, tapers and the shaft itself on the clutch all have to be pretty close in order for it to work right. I might be getting to picky but I would hate to mess up a $500 clutch or possibly take out the crank on my sled.

Might have to try a different design and eliminate some of the variables.

01-04-2010, 09:06 PM
Not sure, if I understand what you are trying to achieve. The clutch looks long enough to benefit from dynamic balancing, which requires a special machine. Static balancing will not eliminate the vibration due to unbalance in such part.

Balancing tolerance depends on the aplication, speed and weight. The higher the speed and the lower the weight, the tighter tolerance is required. The quality of the balancing arbor and the way how it is centered on the clutch bore becomes absolutely essential with tight balancing tolerances.

If you want to use the arbor as a tool to balance multiple parts, you want it to be reasonably hard (let's say HRC 45-50), with minimum runout (ground is the best) and balanced. 4140 steel can be a good choice.

What is the max speed and weight of the cluch?

Glenn Wegman
01-04-2010, 09:23 PM
Sounds to me like the reason you say that the clutches are all out of balance from the manufacturer, is possibly because of your balancer.

It sounds as though you are making tapered ends that slide onto a shaft and you are turning them on a lathe and fitting them by feel. Way too much room for error. I have a balancer that is a shaft with removable discs and as well as being all ground, the fit is such that you can just push the tapered discs onto the shaft and move them by hand. Very close fits.

To have any accuracy at all the parts should all be ground or honed and be very close tolerance fits.

I both static and dynamic balance thngs that weigh 450 lbs and are 11' in diameter, and when balanced, a dollar bill with throw them off.

01-05-2010, 02:59 AM
I have not used this balancer yet. I am trying to get it finished so that I can start to use it. No one around here can balance a clutch so I need my own plus can do this for other people.. I got the measurements from a friend that has the same thing and he will balance the clutch in the same manner. The picture is actually his tool that I used to show what I was talking about.. He will run these clutches on a 290 hp sled with no problem. He knows his balancer is good because he has a custom made billet clutch that will not stop in the same place twice which tells him that it is balanced. My tapers are bored about 3-4 thousands larger then the shaft , just enough to make them slide on with some friction. I will keep playing with it until I get it figured out. Need to learn somehow. The steel shop has lots of material on hand .

01-05-2010, 09:34 AM
nitris223, I have used a static balance like the one in your photo and they have issues. Balancing a long assembly on them is not good, the weight may be close side to side but you don't know if is heavy on one end or the other. A static balancer only tells you if it is close anyway.

The balance shaft has to be very straight, close is not good enough. the cones have to be exactly bored for a SLIP FIT not with 3 to 4 thousandths fit. The cones have to be exactly concentric with the bores, that is, no run out at all or at least only .0002".

The tighter the tolerances are the better you can balance the clutch but a static balancer is only good as a starting point for a dynamic balancer. There is no way you can balance the clutch in the photo on a static balancer and guarantee it is balanced. It may show balanced on the static balancer but if the weight that makes it show it is balanced is on opposite ends 180 deg apart it will be out of balance at speed.

About the only thing that can be balanced without that happening is a disc. You may think you have the clutch balanced with a static balancer but you don't really know for sure.

You don't really think the factory just static balances that clutch assembly do you?

01-05-2010, 09:46 AM
I went back and took a long look at the balancer in the photo and I have issues with it.

1- the V's in the uprights could be out of align with each other.

2- the bearings used to roll the shaft and clutch in are subject to friction load.

The static balancers we used had dry balls running in the races and the surface the shaft layed on was the outer race and had a radiused edge or a knife edge. The balancer in the photo would have a lot of residual friction in my opinion.

Every static balancer I have seen uses a knife edge or a radius where the shaft lays and there is a roller on each side of the shaft at each end. The two rollers form a V to lay the shaft in.

I can see why your not getting repeat readings.

Here is the type balancer we used.

http://www.otswire.com/92%20Small%20Static%20Balancer%20$600.JPG You will have to highlight, cut and paste and google it if you want to see what a real balancer looks like.

01-05-2010, 12:27 PM
This balancer is a copy of the one used at the Polaris factory. The guy that gave me the photos went on a tour of the Polaris Factory and he said there is a lady sitting at a bench doing this all day long. They will balance the stationary sheave first and then assemble the spider and cover and then rebalance the whole clutch. Arctic Cat will balance each individual piece of the clutch there for you can but all the different pieces of the clutch. Polaris on the other hand will only sell you a complete clutch..

As for the v grooves the were cut together and then press fit into t- slots in the base and then the final cuts were taken to make sure they were the same on both sides. They bearings are run dry to cut down on the resistance from the grease. I am going to try and make the tapers again in a 4 jaw chuck to see if I can get them tighter.

Last night was the first time I installed the 4 jaw on the lathe. I figured I would give it a try. Guys say that they like them better once you get use to them. Well I installed it and when I turned the lathe on the whole lathe shook back and forth. I quickly shut it down. only 200 rpm but I could see there was a bad wobble in the mating surfaces to the head stock and to the back of the chuck.The way I seen it, I could try and fix it myself and make it better or screw it up totally and have a fancy looking boat anchor. After 4 hours and dozens of lite cuts on each flat mating surface I now have a nice wobble free chuck. This backing plate must have been made on a Friday or monday because it was nasty. I took at least 1mm off in places to square everything back up. Couldn't return the chuck under warranty because the company House Of Tools is out of business.

Here is another type of balancer sold for doing snowmobile clutches. The shaft taper is 30mm 5%. That is exactly what the stub of the crank shaft looks like and how it is mounted on a sled.


01-05-2010, 12:32 PM
Good luck.