Welding a larger screw

[Robo48]

I've got a larger 1 1/8" screw I need to weld back together. About 3" of it broke off the end. I believe it to be mild steel (pipe wrench bit into it when extracting it) with perhaps a bit of nickel in it? It doesn't appear to be all out SS but hasn't rusted in its lifetime either? Its for a larger butterfly valve I am attempting to fix at work. Of course it has to be somewhat straight when its all said and done (really straight would be nice).

What I am planning on doing is beveling down the two ends and clamping the ends in a piece of angle iron to maintain alignment during welding and cooling. I will weld half cool and then reposition and weld the other half. Where I have to weld it I won't have to worry about the threads being there. When its all said and done will I have to heat treat this to relieve stress and maintain strength?

Any input on this would be much appreciated.

[Jim-Tx]

I welded a shaft on a combine that was about that size. I did what you are talking about.... clamp in angle iron. I used 7018 and started at the bottom of the bevel. I welded, turned and reclamped, and welded more. I didn't really let it cool any. I just didn't get in a hurry and just kept turning, welding and chipping flux. When done, I sprayed the hot metal with WD-40. I think this was more ceremonial than anything else and have no idea whether or not it helped. The smoke made it look pretty kewl though.

This was basically an "emergency fix" that we hoped would last until the new shaft arrived in a day or two. This was on a custom harvester's machine and they left the area a day or two after I did the welding. The next year I asked how long that welded shaft lasted. They said they didn't know since it was still on there months later when they completed the harvest and traded that machine off. The new shaft was never installed.

I'd say that you can weld the screw successfully if you do like I did. WD-40 is optional.

[Robo48]

Jim
Thanks for the reply. This screw will be under alot of pressure at first as I fear the valve overtightened (hence the broken screw) once its cracked open there shouldn't be much force on it.

When you welded the shaft did you weld the whole thing in one pass? or several?

[Jim-Tx]

I just kept turning the shaft and widening (slowly weaving) the bead as I went. It was close to 1 inch wide at the top when I got done. It was hot and the 7018 ran good. 7018 loves heat. I probably took 15 or 20 minutes to weld it up counting the turning and reclamping. If you use stick, be sure to clean the flux/slag before going back over it. I didn't have a MIG at the time so stick was the only option.

This worked for me. I'm sure that there are guys on here that do this sort of stuff routinely and know much more about it.

[Robo48]

Jim
I have a 187 I am planning on using as the mig at work doesn't have enough @ss to do. Hoping the 187 will git r dun. Thanks again for the replies!

If there are any others who have any input about this job please help me out.

[enlpck]

Don't know how I missed this thread....

Not always an easy job. If the length is critical, much more difficult. Jim-Tx's method will work to keep it straight, as will clamping. The key thing is to check for straightness frequently, and, if it starts to go out of line, adjust. As you add more metal, there is more resistance to bending, but metal farther out from the core will pull more. It is a balancing act.

About the best way I know of is to prep one piece flat on the end, and put a cone prep on the other piece for about a 45 deg groove all around. The root is then very little metal, and goes real fast. By rotating, you keep the strain reasonably uniform all the way around. I have seen it done using a lathe-type setup that holds both parts rigidly in line (not just one end) to help prevent bending at the weld. Stop frequently to check for straightness, and if it looks like a whip is developing, deal with it immediately.

If you go with a hunk of angle for alignment, a chisel-edge prep is probably easier. Root it, and rotate to root the other side. Rotate every pass to keep the sides even as you fill. Check frequently for alignment, and, again, deal with it immediately. Again, the narrower the groove, the easier it s to maintain the alignment.

If you arn't sure about the material, I would suggest about 150F to 300F for preheat, and 500F interpass maximum. If it goes near 500, slow down to keep it below. This is a good way to go for many steels, but not all. The preheat will help reduce distortion and help keep the weld material ductile, if it is a plain carbon and for many alloys. A few alloys will have problems if they are not cooled very quickly.

Try a practice piece first so you can get a feel for how things will move. Pulling it back to line with weld metal is usually better than by applying external force. If the weld is in a place where you can avoid grinding it after welding, then you are less likely to have whip, as you won't grind out trapped stress that is maintaining the alignment. Most of the trapped stress that wil be an issue will be near the surface.

As to heat treat: If you don't know what the material is, don't heat treat it. You my do more harm than good.

If you can ID the alloy, you will be best off. Try the manufacturer, and, if they won't help, see if you can find someone with a PMI unit. Many scrap yards have them these days, as do pretty much all NDE testing services, many university and college labs, some fab shops, etc. I've had a few pieces shot as favors, or for a couple beers, as long as I didn't need paperwork to verify the analysis. This WON'T tell you everything about the material (in particular, about heat treat, etc), but will give you an idea of what you are looking at, and let you know if it is something where you need to worry about how to keep it from breaking in the HAZ.

[enlpck]

Also, it sounds like one part is still in the valve body. Unless you are replacing the valve, get the stem loose first, preferably remove it completely. The heat of welding may bind it tighter, or even distort or crack the valve body or butterfly. These forces are only limited by the yeild strength and cross section of the metal, so you could easily see 50K to 100K pounds pushing one way or another (depending on the configuration, of course)

[Robo48]

Also, it sounds like one part is still in the valve body. Unless you are replacing the valve, get the stem loose first, preferably remove it completely. The heat of welding may bind it tighter, or even distort or crack the valve body or butterfly. These forces are only limited by the yeild strength and cross section of the metal, so you could easily see 50K to 100K pounds pushing one way or another (depending on the configuration, of course)

It is out of the valve already. I wouldn't have tried to fix it in place as it was broken right at the top of the housing of the workings of the valve. Kind of a nice design for working on it as the whole mechanics of the butterfly actuator are in a water tight cast iron "box" and very accessible after removing the cover bolts.

[Robo48]

Don't know how I missed this thread....

Not always an easy job. If the length is critical, much more difficult. Jim-Tx's method will work to keep it straight, as will clamping. The key thing is to check for straightness frequently, and, if it starts to go out of line, adjust. As you add more metal, there is more resistance to bending, but metal farther out from the core will pull more. It is a balancing act.

About the best way I know of is to prep one piece flat on the end, and put a cone prep on the other piece for about a 45 deg groove all around. The root is then very little metal, and goes real fast. By rotating, you keep the strain reasonably uniform all the way around. I have seen it done using a lathe-type setup that holds both parts rigidly in line (not just one end) to help prevent bending at the weld. Stop frequently to check for straightness, and if it looks like a whip is developing, deal with it immediately.

If you go with a hunk of angle for alignment, a chisel-edge prep is probably easier. Root it, and rotate to root the other side. Rotate every pass to keep the sides even as you fill. Check frequently for alignment, and, again, deal with it immediately. Again, the narrower the groove, the easier it s to maintain the alignment.

If you arn't sure about the material, I would suggest about 150F to 300F for preheat, and 500F interpass maximum. If it goes near 500, slow down to keep it below. This is a good way to go for many steels, but not all. The preheat will help reduce distortion and help keep the weld material ductile, if it is a plain carbon and for many alloys. A few alloys will have problems if they are not cooled very quickly.

Try a practice piece first so you can get a feel for how things will move. Pulling it back to line with weld metal is usually better than by applying external force. If the weld is in a place where you can avoid grinding it after welding, then you are less likely to have whip, as you won't grind out trapped stress that is maintaining the alignment. Most of the trapped stress that wil be an issue will be near the surface.

As to heat treat: If you don't know what the material is, don't heat treat it. You my do more harm than good.

If you can ID the alloy, you will be best off. Try the manufacturer, and, if they won't help, see if you can find someone with a PMI unit. Many scrap yards have them these days, as do pretty much all NDE testing services, many university and college labs, some fab shops, etc. I've had a few pieces shot as favors, or for a couple beers, as long as I didn't need paperwork to verify the analysis. This WON'T tell you everything about the material (in particular, about heat treat, etc), but will give you an idea of what you are looking at, and let you know if it is something where you need to worry about how to keep it from breaking in the HAZ.

Whoa.....First off excellent answer but not sure I know enough to try this repair after reading it I intending on beveling both ends of the screw and using the "angle iron method" to keep it straight as its what I have. One thing I have working for me is there is a collar that goes on the screw right over the welded area. The collar is what traps the screw in place and forces the arm of the butterfly to move up and down the threads on the screw as its turned. I am going to turn the threads out of the inside of the collar and bevel the top and bottom of the inside of it then weld it to the screw. Most likely I will have to turn the shoulders square on it again after welding. When I say the screw must be straight it obviously isn't going to be a spindle on a bridgeport or anything.

[enlpck]

It all comes down to how straight it must be and what the material is.

I have done similar jobs, and got results straight enough not to have problems with bearings on both sides of the repair in low speed applications. For a valve, probably more than good enough if the rod isn't going through seals, or there is minimal restraint outboard of the repair (no tight guide bushings, for example)

The material question is what would worry me. If it is a nothing-lost situation, take your best shot. If it is critical it be done right the first time, you will want to gt an ID on the material and any heat treatment.

As a side note: I have several good straightedges with notches cut in them for this kind of job. The notch is for clearance over the weld cap. Same basic idea as grinding the corner off a framing square to clear the fillet on tee joints.

[hogan]

Robo48,
I have to agree with with enlpck:

If you can ID the alloy, you will be best off. Try the manufacturer, and, if they won't help, see if you can find someone with a PMI unit

Before you do anything you need to know what the material is. This will determine everything else! Especially knowing that this shows no signs of oxidation (rust). The PMI testing is cheap and will help to advise a proper welding procedure. This is still something you should be able to repair, you just need to know the material ID to start off.

[Robo48]

After consulting a local welding shop I learned the screw is nothing more than mild steel. The guy told me to prep it and have at it. I welded it the other night and all seemed to go well turned out straighter than I thought I could get it. I had it on the lathe at work turning the welded area and there was no wobble whatsoever (still think I got lucky to some extent). The welded area ended up being about 1" long and required what seemed like 100 passes to get it built up. Hopefully the weld holds once it is installed and the valve is being opened. I will keep you guys posted as to the results. Thank you for your help.

[Jim-Tx]

Sounds good. I'll bet that it holds. Got any pics?

[Robo48]

I don't have any pics (which I should take a couple before assembly). Screw is 1 1/8" diameter, 25" long, it has 4 holes drilled in it for pins to go through 2 of which are for positive stops (to tell when the valve is all the way open or closed) the other is for a pin that holds the "wrench nut" (large 2' by 2" square nut you put a "key" on and turn the valve with from ground level) and the other didn't have anything in it. There also is a groove cut near the top that an O ring rides in to seal the whole assembly up. The "collar" I welded on is what makes the whole thing work as it keep the screw stationary forcing the arm on the butterfly to thread either up or down actuating the butterfly thus opening or closing.

One thing I will add is every time I use my little 187 welder I am more and more impressed with it. Simply amazing how much power these little guys pack and how smooth they run. For most homeowner hobby type welding I don't know if there is a more perfect machine. I am trying to get them to buy one at work as we have an old miller 150 mig welder and it just doesn't have enough @ss to accomplish much. I will see if I can get some pics of this screw where I fixed it and when it is installed with the cover still off the mechanism so you guys can get a better understanding of what I am working on.

[Robo48]

Wanted to let everyone know I put the valve back together yesterday at work and it opened right up. I couldn't get any pics as it was dark in the manhole raining and not a very nice day in Michigan. Thanks to all for the help and responses.