Thread: whos the best at shaft straighting

if you had a cross feed shaft for a lathe that has a bend in it noy bad bud wobbles, who would you send it to if you wanted it right . jim
its around 18 inchs long and acme . with turned ends . have pics will post here .

I'm pretty good at shaft straightening. I've done plenty of it anyway: little bitty one to great big one and lots of ones in between.

I'd roll the major diameter of that lead screw on the granite flat with the thrust collar just hanging off and use the info I gained there to smack the high spots. Straighten it spanning a couple of wood blocks with a hardwood mallet or a wood block protecting the work with a from a steel hammer. Practice on something about the root diameter of the thread.

This is where a solid controlled bump from a heavy hammer has greater success than a less than controlled blow from a light hammer. "Don't force it get a bigger hammer" is not an ironic jest: it's Revealed Truth in many machine shop settings.

After you get the threaded part straight and if the ends are still bent, dial them in separately and look for origins of the bend. Locate the max position of run-out and grip in in protected vise jaws. Apply enough force to the screw to just short of bending and whack it close the to vise to make the adjustment.

Straightening is an art. Do not practice on finished parts. It's very easy to make an un-recoverable kink in an Acme screw thread. Make a mock-up and straighten it. When you can straighten a mock-up reliably you can straighten the real thing.

I look at it this way: Steel has "memory" and is more than happy to go back the way it was (straight). The trick is to coax it in the right direction. Usually this means applying varying degrees of pressure at the highest points along with varying the spacing of the supporting blocks.

Forrest and Mike are right on. Vertical shaft lawn mowers frequently suffer from crankshaft bends when people hit concealed objects. Most of the time, the mowers get trashed because it is not worth the time and effort to put in a new crank. I learned long ago to put that sucker up on it's side, remove the blade hub, and smack the high side of the run out with a heavy hammer. A lead hammer or something in between to protect the shaft works well for me. The shaft has a good memory, and amazingly good concentricity can be achieved. An old body and fender man told me long ago to take the dents out in the reverse order of how they happened, and you will minimize stretching the metal. The custom barrel makers who wanted to put all their bullets in the same hole carefully sighted the bore, and wacked the barrel with a lead hammer. I guess experience counts for a lot, so send it to Forrest, or pratice a lot on some other stuff. There is a great satisfaction in being able to learn and do something yourself, so don't ever be afraid to try something new.

Jim

Well, you've heard from two of the best already but I will toss in a method that worked for me on the worm shaft of a 12" rotary table. I got it cheap because the shaft had a wicked bend in it from being dropped. My plan was to cut the worm off the shaft and mount it on a nice new chunk of shafting. My father in law suggested I try the following. As he pointed out, if it didn't work I could try my original method. His idea was to mount one end of the shaft in a four jaw chuck and dial it in. Using the largest lathe available, the far end was supported loosely with a big,rugged live center. As the shaft was slowly rotated a torch flame was played on the area of he shaft where it was bent. The shaft was only heated to 400-500 as determened by a temp stick crayon. The tailstock spindle was tightened and clamped. After the still-turning shaft cooled down to room temperature and the tailstock removed it was almost perfectly straight! Maximum runout at all checked points along the shaft was within .001". When the shaft was reinstalled, it worked great. Perhaps this was a fluke, and not somthing to try on the best lathe in the shop but it did the job quickly and did not require the skill that other methods do.
Let us know how you do,
hms50

We do lots of little straightening jobs, but always use a hydraulic jack in a press to do them with. This includes lawn mower cranks, but we make them strip the crank out and bring it to us Usually we get them within .002" before we quit and that accuracy would be checked with the part on centers in the lathe.

We make up aluminum V-blocks to press with. Lead is too soft, but aluminum seems just right. These do not need to be anything fancy as they will displace over time.

For straightening shafts, detection of where the bend really is, is important. A straight edge or surface plate as Forrest mentioned is indispensible.

Equally important is how far apart you place the v-block supports beneath the shaft. Say that that 18" long screw has a gradual bend in it, generally looking like it is maxed out at the middle. Some guys would put the vblocks at each end and press the middle down. Congratulations, you have just created a crankshaft

The optimum spacing for the support v-blocks would vary somewhat according to the severity of the kink in the shaft. But, for straightening a gradual bend, I would estimate that the v-blocks should be 5x to 6x the shaft diameter apart.

That does not mean that only pressing in one location is sufficient. A gradual bend needs to be removed by pressing in several locations. Although the v-blocks should stay about the same distance apart, you can press at points spaced whatever...2 inches apart along the shaft.

I favour a hand operated hydraulic jack in the press. Typically, we use a 20 ton, because it moves slow. When enough pressure has been applied to exceed the elastic limit of the material, every thousandth past that adds to the new bend.

Another benefit of using aluminum V blocks (even gross, crude mushed over blocks) is that a round shaft will rotate quite steadily in position. This makes it possible to use a dial indicator right at the press to check the progress. Put the indicator beneath the part, directly under the press ram. Press down, watch exactly how far you go, then release the pressure and rotate the shaft. Note the runout, and whether you changed it any. After just a very few trials, you will note what reading the indicator has when 'real bending' is taking place. Then it is a simple matter to add so and so many thousandths to get just enough new bend.

You should still work with a straightedge check as well. It is easy to become too focussed on one area with the v-blocks and get yourself in a real mess.

The 'top' of a bend in a good steel shaft is pretty tough to push compared to bending it back in the same direction as the original bend. You can really feel this with a hand operated jack. Take lawn mower crankshafts for example, they are quite tough to straighten, and if you go too far, and start to push it back, it just moves way too easily. A straightened crank is forever prone to having a weak side. Same goes for soft 'CPO' hydraulic cylinder ram shafts.

BTW, use a third aluminum V-block under the ram to protect the shaft. Pressing with a steel ram directly onto a steel shaft will dent the surface. The V-block spreads this force out over at least twice as much contact area.

Roller rests, small oxy-act torch and tip, cooling air/oil mister and a dial indicator and stand. Can be gotten to within a couple thou. JRouche

I have straightened a lot of shafts for boats and machinery. HuFlung outlined some good ways to do it. I always prefered to use a lathe with dead centers in headstock and tailstock and a bottle jack between the part and the lathe bed. You put the bend down and jack it up and hit it with a lead, brass, deadblow or ballpeen hammer as required to keep from damaging the shaft. You can move along the part and straighten as needed using a dial indicator on the carriage to check straightness. Hammering the part relieves the tension in the bend. On SS boat shafts we used a ballpeen as the SS is hard to straighten and the marks can be filed out without affecting usability. Heat is sometimes needed to release the stress in the bend but be careful. Using a press takes to long in a job shop and I seldom did it that way.

Carl,
Remind me not to buy your used lathes

We do have a captive fixture for a bottle jack to use on the lathe as well. Again, this is a set of V blocks in a frame that straddles the shaft. The bottle jack is inside the frame. The lathe itself takes zero stress whenever the straightening is done.

Flame straightening is indispensible for tubes, but we don't often do it for shafts. Hidden stresses in the shaft can end up pulling the thing the wrong way. I know there is no alternative at times. Flame straightening does induce 'upset bumps' so its not without drawbacks.

We use one of these

If we can't get within 0.001" over about 4', it's junk. I can't think of anyone in the area who's a job shop that has one. You might call Eitel and ask them.
JR

I do not claim to be an expert, but I have had good luck straightening a steel shaft and also some rolled aluminum plate using a 25-ton hydraulic press. The hard part for me is judging how far the stock is going to spring back. Smaller sections will spring back quite a lot and will yield without the pressure gauge needle ever moving.

I found that I could make small and controlled corrective bends by attaching a flexible ruler next to the ram on the press. Taking care to position the work the same way each time, I increased the stroke length by 0.1 or 0.050 until I could measure the piece not springing back the whole way. Then I kept reducing the increment to the limit of what I could judge with my eyes. The final presses were repeats of the same stroke length.

I checked the parts on a surface plate between each press and marked the high and low spots and where to position the supports with a Sharpie. For the shaft I put the supports to either side of the area of the bend rather than the ends.

I don't trust myself enough with a hammer to use that method. [img]smile.gif[/img]

What everyone else said, except...

If you hit some big, immovable object with your brand new Honda mower, don't wail on the stub of the shaft with a hammer. The bearings don't like it. Use a really long, tubular lever and a strong friend. Works like a charm.

We often straighten shafts for the electric motor company located across the street. The same procedure is followed for every shaft, which is:

Secure one end of the shaft in the 4-jaw chuck (use brass or aluminum pads), rest the other end in the steady. Indicate the 4-jaw to within .001 & rotate the chuck by hand to visually locate the high/low spot. Set up a mag base & indicator & rotate the shaft until the bend is at the bottom & apply heat to the side opposite the bend.

The heat will start straightening the shaft, continue heating until the shaft is fully straightened plus .010 more. Let cool, check the runout, most likely you'll need go through the process four or five or six times to get it finished.

The easiest shafts to straighten are the long, thin ones (like yours). The hardest are the ones bent right at the end & three or four inches diameter

----------------
Barry Milton

We built a rig we used to straighten the raw stock for conveyor shafting we made. (.002 max tir)
Shafts were various lengths worked great for everything.
It clamped the shaft ends wherever you placed the clamps on a beam (machined v-block with clamp) It pushed the shaft up from the bottom (hand lever), and had a gage on top that was just a graduated piece with a slider that moved with the part, and the witness mark was spring loaded to keep it at the highest point. This gave a ref. point, so you could keep trying till you got it right.
It had a swing away travel indicator for inspection. This, and the bend gage slid on a rail, so you could find the low spot, and check tir at different places.
If you went to far, rotate the shaft and go a little less than when you over bent.
Worked out with the cycle time of the cnc lathe.

Check this out, put pressure on a bent shaft in a press and while its under pressure strike it with a hammer. Do this until its true.
Then with no pressure on the trued shaft, strike it again with a hammer, test it again and see what happened.
One can learn somthing about the nature of metal this way.
Part of my making a living is straightening out multi piece crankshafts, a complex bending problem.

Oh man. Straightening is a reaccuring nightmare for me.

Everyday I get to play with 10' by 8" warped and kinked shafts that only need a proper taper and thread chased. THere are two bushing surfaces that matter too, and they're about 35 inches apart.

The customer of course supplied the material, and needs them all done YESTERDAY.

So.... I spend 1/2 a day straightening with a crane + hydrojack getup, drop it on rollers and check concentricity, push a little more, check, push, check, you get the idea.

Then it's all of MAYBE 2 hours worth of actual cutting. We have a decent enough 20'+ lathe that can rough 1" on a side with the right machinist workin it.

So I spend 80% of my time straightening, and 10% actually cutting, another 10% fighting for crane time.

If I were king, I'd allocate a corner of the shop for a I-Beam and 30Ton hydra, nice rollers and a rolly-swing crane to do it all, but alas....

Like JRIowa we use an Eitel. 3/4 inch to 8 inch, 18 inches to 20 feet, and enough indicaters to check.

Also handy to fix chuck wrenches.