4140 HTSR TGP

[Hertz]

Hi guys. Quick question about this material relieving. We purchase 4.5" diameter Turned, Ground, and Polished because the application requires it to stay straight within a couple thou. They are about 5 ft long. We have to cut some of the top and some of the sides to create like a square boss to fit into a mating slot. This square boss is along the whole length. The problem we're facing, is that the parts seems to be warping up to .008" when we're done. It's not heat, because the part is very cool to the touch and all the heat is in the chip when milling. Is this material still relieving and if so, is there a good way to combat this?

 

[david n]

Rough cut.........send out for stress relieve............finish cut...............or a huge straightening press??

 

[4GSR]

Even after stress relieving and you start to mill on it again, it's going to walk on you. The best bet after stress relieving is to surface grind by flipping until the size you want is met. It should be dead straight by then.

 

[Hertz]

Even after stress relieving and you start to mill on it again, it's going to walk on you. The best bet after stress relieving is to surface grind by flipping until the size you want is met. It should be dead straight by then.

Even after stress relieving and you start to mill on it again, it's going to walk on you. The best bet after stress relieving is to surface grind by flipping until the size you want is met. It should be dead straight by then.

Unfortunately it's round material with only one side getting milled. The round part is essentially bowing and without affecting the diameter, I can't really machine or grind.

 

[4GSR]

Agreed. I was aware that you started out with round bar.

 

[Mtndew]

Either send it out for straightening or buy larger stock.

 

[Milland]

It's the one-side milled that's the issue, and I agree with those that say even a stress-relief may not be enough to prevent warpage when down to final size.

What "might" work if you make enough of these to be worthwhile is to make a robust fixture that this shaft clamps to that has a purpose-made bend along it's length, that runs counter to the typical "after milling" bend of the shaft.

Presuming a VMC doing the cutting (rotate as needed if HMC), place shaft in fixture, tighten ends or middle such that the shaft is pre-stressed, then profile a curved cut with a large round tool (so the geometry of the curve matches what will be a flat cut when removed from the fixture).

This is not guaranteed to work, but could be tested with, say, a 1" diameter bar proportionally long of the same material in an ad-hoc fixture to see what relative pre-curve might do to the final, relaxed cut part. If you can find a relationship of pre-curve amount and a functionally good final part, then try the large fixture with the correct size shaft.

Again, can't say for sure it'll work, but trying it with a smaller bar and similar fixture may be cheap-fast enough to test. I do not recommend trying to tighten directly to the bed of the machine, let the fixture take the strain and the machine only holds for cutting forces.

 

[William Lynn]

Does your shop have the capability to grind or turn (assuming finish is acceptable)? We had almost the same problem except the shaft was bowing once a keyway was cut. Tried stress proof steel with the same result. Ended up cutting the keyway deep, putting a sacrificial key in the keyway, then turning to finish OD. Shaft then was straight and serviceable.

 

[gbent]

If you can make a 5' long cut on one side and they are only out by .008 you are starting with good material. You need a good man on a straightening press.

 

[Rickyb]

From an engineers point of view I see no viable option other than straightening post machining. warp will depend on depth of hardness and will vary significantly depending on material chemistry and the heat treat batch. It’s a safe assumption that the supplier does not tune the quench to the material jominy, so case depth of the martensite will vary dramatically for each batch. This is the part of the shaft you are machining away causing imbalance of stresses resulting in warp. You see 0.008” now but the next batch will be 0.004“ or 0.016”.
yes 4140 is through hardening but not at 4.5” and not without a tuned quench per batch.
4340 will quench, at least partially, to the core and better distribute internal microstructure and stresses. Heat treater will still cause variability so 4340 will minimize your problem but not eliminate it. This takes you back to a straightening process.

 

[guythatbrews]

If you can tolerate some marks on the side opposite the flat I think peening will straighten this pretty quick. I've never done anything that big a diameter. Maybe a needle scaler?

 

[hvnlymachining]

Flame straighten? Seems like a good candidate considering the light bow. But the actual end use will determine if the heating is even allowed.

 

[Milland]

Flame straighten? Seems like a good candidate considering the light bow. But the actual end use will determine if the heating is even allowed.

Steel bar 4.5" by 5 foot is a lot of mass, would take a big rosebud and some patience to get that right.

I think it's either mechanical straightening or perhaps getting permission to flat the other side an equal amount. Changing the other parts to work with a double-flat shaft might be cheapest in the end.

 

[4GSR]

You could leave the OD about .030" oversize. Mill the flat, then line up and turn the OD to size. But even then, there's still going to be a chance of distortion from turning .030" off. 4340 or better yet 4330v may be a better choice of material to start off with. Either will be more stable in the long run than using a 41xx grade material.

 

[Garwood]

All the suggestions seem a lot more complex than just straightening the shaft in a big press. I haven't straightened a shaft that size, but I'd think a 200-300 ton press with some pads cut to protect the surfaces would do it.

 

[4GSR]

All the suggestions seem a lot more complex than just straightening the shaft in a big press. I haven't straightened a shaft that size, but I'd think a 200-300 ton press with some pads cut to protect the surfaces would do it.

It's a little tuff getting the last foot on each end straight in a press. I bet you they could get it within a thou or two doing so.

I've worked at a place in my early childhood where they had a 300 ton press made just for removing the kinks from bar material before they trepanned holes in it. Guys were darn good at it, get that hot rolled bar to within a 1/64 end to end except for the last foot or two. (not including the out of roundness of the bar) I recall them straightening a bent motor shaft on that press to within .003". Stuck it back in the 75 HP motor as far as I know it was still running when they shut the doors a few years back. That was 45 years ago.

 

[hvnlymachining]

I had a 4-1/2" worm gear shaft come in for repairs a couple years ago, had to take it to a shop with a 300 ton press to get it close then I finished it with flame and welding (the bearing journal had been thoroughly destroyed by a locked bearing.) I had it straightened before I did the welding In case it cracked or fractured. The very large machine shop said the 300 ton press was nearing max to straighten it. But they were only willing to get it to .007 out. It started at .030" warp, but that was all in a short distance. It left my shop at .0015".

In any case the OP has few truly good solutions to the problem, most or all will be trial and error. Re creating stress on the flat side may be easier than stretching the opposite side, small welds sunk into the flat face of the bar perhaps?

I have straightened heavy steel by placing the ends on blocks and driving an appropriate weight vehicle on it, but accuracy could be iffy.......

Hertz, please keep us posted on the eventual fix, I'm curious about this as I can see me in this same situation eventually since much of my work is in larger materials.

 

[Overland]

It's the one-side milled that's the issue, and I agree with those that say even a stress-relief may not be enough to prevent warpage when down to final size.

What "might" work if you make enough of these to be worthwhile is to make a robust fixture that this shaft clamps to that has a purpose-made bend along it's length, that runs counter to the typical "after milling" bend of the shaft.

Presuming a VMC doing the cutting (rotate as needed if HMC), place shaft in fixture, tighten ends or middle such that the shaft is pre-stressed, then profile a curved cut with a large round tool (so the geometry of the curve matches what will be a flat cut when removed from the fixture).

This is not guaranteed to work, but could be tested with, say, a 1" diameter bar proportionally long of the same material in an ad-hoc fixture to see what relative pre-curve might do to the final, relaxed cut part. If you can find a relationship of pre-curve amount and a functionally good final part, then try the large fixture with the correct size shaft.

Again, can't say for sure it'll work, but trying it with a smaller bar and similar fixture may be cheap-fast enough to test. I do not recommend trying to tighten directly to the bed of the machine, let the fixture take the strain and the machine only holds for cutting forces.

You might end up with a "straight" bar on the original surface, but surely a "bent" bar on the machined surface; a tapered bar.

It seems that the bar is being machined to fit into a slot. Is there a way that the bar could be held in the slot such that the fastenings pull it back straight ?
Bob

 

[richard newman]

Hertz, do you check the bar for straightness before machining it? .008" over 5' might be within tolerance for TG&P.