constrained joints

[crzdriver]

I'm welding a four inch solid shaft from an 18-wheeler hub to a half inch mild steel plate. There is a hole in the plate, and it's a butt weld around. This thing lasts about two months before the cracks are visible again. It keeps breaking at the interface with the steel plate. I'm welding it with 7018, preheating it to death, etc. My vote is that the geometry of the joint is causing the cracks to start as soon as the material starts to cool and has no place to loose tension. It's obviously not lamilinear tearing because the stress is in the direction of the grain in the plate. Anyway, is there any good info on contrained joints on the net? I know pipe flanges are welded all the time all day long without cracking, and its a very similar joint. Any ideas?

[dabar39]

Huh???????????

I'm not sure I'm getting the right thoughts (not that I ever do) in my head by what you describe. PICTURES and lots of them, any pics you can give us will help in trying to come up with a cure for you. Dave

[crzdriver]

I can't figure out how to upload a picture, so I'll do my best. Take a piece of plate and lay it on a table. Cut a four inch hole in the middle and remove insert. Lay a shaft inside the hole and butt weld the back side. The design of the part makes it impossible to access the back for a fillet weld. Basically, the part makes a T and I'm welding the top of the T with the saft being the vertical part of the T.

[weld_13]

i guess ur usin 7018 1/8 rod? what amp setting? im not shure i know of any info for that but ill look on the net.

[Fat-Fab.com]

Are you beveling the end of the shaft?

What is the shaft made of?

What is the function of this?

Below is me procedure for situations like you find your self. This only applies for non life safety situations.

If the weld or joint fails and people could get hurt, then you are on your own.



I would bevel the heck out of the shaft, deep enough so that I could do several passes up to five no less than 3. I would remove any existing weld. I would preheat the shaft and plate after tacking. I would place welding blankets around the weldment so that cooling would take as long as possible.

I would use 8018 or 10018, or I would use something with nickel in it. Coreshield 8 if I were running wire.

Please let me know the answers to the above questions as thet could change my answers a lot.

[crzdriver]

More data! I finally made contact with the manufacturer. The shaft is 4041 surface hardened to I don't remember what. It's a spindle off of an 18-wheeler. The plate is plain old A36. They sent me a detailed weld procedure that didn't detail filler (figures). Anyway, you nailed it, they said the cooling phase was absolutely critical and that we needed to sandbag it or any other means to prolong cooling. I actually welded it last with 3/16 7018 around 220 amps. The procedure said small beads and not too much heat was better, but I'm still not sure of the recommended filler.

Which brings me to a long-time question of mine. What is more critical when welding dissimilar metals together? Chemical or physical properties, or a combination. I know the higher alloy rods are more brittle, but are closer in chemical makeup to the higher alloy steels. However, 7018 is about ductile as it gets and makes forgiving welds much of the time in the proper hands. I come across this a lot like the shaft problem, or welding AR plates and sheets over mild steel frames. Thanks for all the help!

[JonnyTIG]

I weld similar stuff all the time, usually 4140 or 4340 to mild steel (A36 / 44W). The machinist at work has always had me preheat the shaft around 325F and do the weld with 8018 C3, or a 80ksi rod with good elongation properties. Medium carbon steel or low alloy steel electrodes work as well. We don't have a sand box, but I wrap the parts in a fire blanket to cool slowly, usually takes overnight to cool. In my application the parts I'm making are usually transmission adapters for dyno testing large equipment trannys. Lots of torque loading and vibration, never had one crack or break yet. Sounds like you're on the right track.

I hope this helps

[JonnyTIG]

Sounds like you are on the right track.

[Fat-Fab.com]

As to filler I would be looking for something with a Charpy Impact Test # of 20ftlbs at -20F. This is good rod for stable elongation at a wide temperature range.

The reason for the small beads and lots of them is to build a lot of heat in the part and the slow cooling allows the slow release of the built up stress.
I would do this only when I had the time to weld straight through to cap.

Good luck

To post pictures look to the bottom of the post dialog box find the manage attachments button and click new window opens (small) click on browse to find a picture on your system. Note they cant be very big. If your photos are too big to upload open a photo editor and save a copy under "save for the web" then publish the copy.


NOW we want pictures! lol

[MMW]

go to www.weldit.com click on products, then alloy steels, then polaris aaa. I've used this wire & stick rods for welding 4140 plate to mild steel plate many times. I pre-heat, weld then let it cool on it's own. Never had a problem. Or www.rockmountnassau.com (same place)

[Anti-GMAW]

go to www.weldit.com click on products, then alloy steels, then polaris aaa. I've used this wire & stick rods for welding 4140 plate to mild steel plate many times. I pre-heat, weld then let it cool on it's own. Never had a problem. Or www.rockmountnassau.com (same place)

Very impressive site. They've got alot of uniqe fillers, thats for sure.

[st333v]

I'm welding a four inch solid shaft from an 18-wheeler hub to a half inch mild steel plate. There is a hole in the plate, and it's a butt weld around. This thing lasts about two months before the cracks are visible again. It keeps breaking at the interface with the steel plate. I'm welding it with 7018, preheating it to death, etc. My vote is that the geometry of the joint is causing the cracks to start as soon as the material starts to cool and has no place to loose tension. It's obviously not lamilinear tearing because the stress is in the direction of the grain in the plate. Anyway, is there any good info on contrained joints on the net? I know pipe flanges are welded all the time all day long without cracking, and its a very similar joint. Any ideas?

QT-100 plate steel, bend on break press,diamond shape holes with rounded angles. stitch weld.....don't look back, ur good 2 go now.

[welder_one]

when is the last time you tried to bend T1 on the brake??? (qt100 = T1)

[hockeyguynick]

I used to work at a tractor trailer repair shop. We had a guy come in and weld replacement axle stubs on all the time when the truck blows out a bearing. This was before I really got into welding so I didnt pay a whole lot of attention to what he was doing. But I do remember him preheating, and weld a little, then let it cool a bit. When it was all done, he pointed a torpedo heater at it and let it cool very slowly. He would go nuts if we opened a garage door in the winter while he was working.

[seattle smitty]

Do you weld clear around the shaft in one pass? In school they told us to divide the circle into eight segments, and make the root pass with eight sequenced (opposite sides) welds. It's a lot more trouble, and you have to make nice clean starts and stops (and grind out bad ones as best you can), but it is supposed to reduce the stress accumulation. That said, maybe slow-cooling solves your problem without resort to this method.

I'm always interested in reading about these problems; glad you posted it.