Welding on a case-hardened gear?

[isuzu14b]

I have a case-hardened gear from an automobile transmission that i have modified, however because of the modification, i now need to re-weld the two parts of the gear back together. I say "reweld" because these pieces were welded from the factory (mitsubishi) and the modification caused the weld to be machined off. Obviously mitsubishi case hardened the gear after the welding process, but now i dont have that same luxury. The outer, hardened layer of metal has been machined off at the location of the new weld, and since i would be welding the "core" metal i am wondering what results i can expect as far as strength is concerned. Also, i do not know what alloy was used to make the gears, and if anyone has a good idea of what it may be i would be interested to know. I was planning on TIG welding the pieces back together without any filler rod. During the welding process i need to keep the gear as cool as possible since i do not want to loose temper/harness. At what temperature will the temper become compromised? Someone suggested packing the gear in glass bead or sand as a heat sink, do i have any other good alternatives?


I have attached several images of the modified gear, as well as an unmodified gear for comparison.
The synchronizer cone has been machined off to accept a double-synchro assembly, as seen in the images.

Thank you very much for any information you can give me!

 

[CoryPad]

You ask about what to expect for joint strength - in torsion, tension, bending, other? Do you plan to do any post-weld heat treatment? Your joint geometry is unclear, so it is difficult to provide guidance. For parts like this, static strength is not critical, fatigue strength is. In industry, we test this kind of thing rather than rely on any formulae.

The alloy(s) for the gears likely are something like JIS SCM 415 (similar to SAE 5120), or JIS SNCM220 (SAE 8620) or JIS SNCM420 (SAE 4320).

If the gears have controlled sulfur concentration for machinability, then you may experience cracking if you do autogenous welding (no filler). High purity filler (low sulfur) reduces cracking tendency.

You will lose hardness on the teeth if they go above 150 [°]C. The core will lose strength is it reaches somewhere around 400 [°]C to 500 [°]C.

Regarding the beads/sand, they would be used to provide a uniform temperature over the entire part to reduce distortion. These are gears, so their shape is critical.

You mention several images, but I can only view two images, and I cannot determine all of your geometry details. Caution is warranted on this weld.

 

[Wrenchbender]

I was unable to view the pictures. My semieducated guess at the material would be that its a low alloy, low carbon steel such as 8620 which is suitable for carburizing. The base material is easily weldable, but the presence of the high-carbon surface layer may cause post weld warpage, and cracking problems. There may be inhomogeneous areas of decarb’ed material, untempered martensite (brittle) and high stress gradients, all of which depend on part geometry, cooling rates and other difficult to control factors. Check the thread in this forum from 19 Aug 2009 concerning welding on case hardened steel for more info.

So how to mitigate these problems - I suggest
(1) Experiment - Weld up a few simulated mock ups to see what to expect.
(2) A heat sink should be a conductive material, not a thermal insulator as had been suggested. Try clamping in Al blocks or something that will remove heat rapidly.
(3) Use intermittent heat input. E.g., weld for several seconds; cool for an hour.
(4) Otherwise, look into a preheat and post heat; others in the forum may be able to advise better.
(5) My seat of the pants limit on stress relief/tempering temperature is ~500 deg F.

 

[mfgenggear]

would suggest making a simple drawing to show your weld joints, if possible take hardness readings in the weld area.


all the 9310 steel carburized gears i have mfg are tempered @ 275 Deg F, so preheat & stress relieve sould be not more than 300 deg F.

to prevent cracks it is crucial to preheat the gears to 300 Deg, F prior to welding. if welding in a carburized area it's very very bad.

if it's a non functional area(no carburize) then it may not crack.
Heat effected zone from welding will soften your core & surface hardness.
re-heat treating (subcritical anneal, austentize quench & temper)is not recomended because it will distort your gears & any critical diameters.

after welding I would highly recomend NDT Magnetic Partical Insp. for cracks.

It is not normal practise to weld after carburize, all tho it can be done. if done correctly.

 

[TVP]

Have you considered arc brazing instead of welding? The process is also known as TIG brazing or MIG/MAG brazing, depending on setup, it may be a better option due to the significantly lower heat input into the parts. The process uses a lower melting point filler rod, typically a silicon bronze alloy, and deposits a "weld" bead that is very similar to a TIG weld, but with even lower heat input. The process is often used to join thin sheetmetal parts, especially zinc-coated steel sheet, but is equally suited to non-coated applications. Here are a couple of links, but you can google "TIG brazing" or "arc brazing" or something similar for more information:

http://www.handyharmancanada.com/TheBrazingBook/contents.htm#Index

http://www.ewm.de/content/application/x-pdf/WM031801.pdf

 

[Manifolddesigner]

I second the notion. Silicone bronze would be the only way to go.

 

[isuzu14b]

Thank you everyone for all of the great information!

CoryPad
Im not sure of the correct way to word the stresses on the weld. What comes to mind is to say Radially Shear. That may not be technically correct, but to elaborate, the "dog ring" will be fixed to the rest of the gear and the stresses on the welds would be in the direction that the gear turns.
I am not sure why you cannot see all of the images, i can see them all on several different computers.

Bestwrench
I have been experimenting with some unusable gears. I have been able to get harness info this way as well as to test the weldability of the core material. What i have found is that the metal at the surface of the dog teeth and gear teeth is hardened to nearly 60 rockwell. The core material is about 30 rockwell.
As for the heatsink, i have considered using BB's or lead (or steel) shot or something along these lines. I am definitely open to suggestions in this dept.
Also i am in total agreement with the intermittent weld technique. Obviously keeping the gear cool is critical.

mfgenggear
Unfortunately I dont have a weld drawing to show you.
There will be no welding in hardened areas, though the welds will be close, maybe 0.250 inches away from hardened material.
Preheating sounds like an excellent idea.
"magna-fluxing" the gear afterwords should not be a problem.

TVP/Manifolddesigner
I really like the idea of using brazing, however i am unfamiliar with the process for this particular application. I am also particularly unsure about how to use the TIG process for brazing, as i have never heard of such a thing. I will do some research though, and if it is a viable option, it will be among the top of my list of choices.


Thankyou again everyone for all of your helpful information.

I have done a few preliminary tests on some test gears and so far it seems that i can keep the hardness up on the dog teeth and gear teeth if i allow a significant cooldown period. I am unsure of how to test the strength of my welds to gather any useful information, i have a hydraulic press, but no way to put radial torque on the pieces to check them for real-life stresses.

For my next test piece i plan to use a preheat, with the gear packed in BB's and i will only be applying heat in short bursts and allowing considerable time for thermal stabilization between welds. Then i will NDE the test piece to check for cracking in or near the welds. Hopefully everything works well.

 

[CoryPad]

isuzu14b,

Do you have the images stored on different servers/Internet sites? I was able to see all of them from a different computer, but still only see two images from this computer.

The stress state you are concerned with is torsional shear created by twisting the shaft during normal use, plus any bending induced along the shaft. Radial shear is if you put the shaft in a guillotine and try to cut it into two pieces.

Preheating is good for unhardened parts, but you don't want to temper the surface back to a lower hardness. I suppose you could use a preheat temperature that is just below the tempering temperature (~ 150 [°]C), but that isn't much of a preheat.

 

[isuzu14b]

CoryPad

Ahh yes, torsional shear makes sense. That is definitely the primary stress the welds will have to withstand.

The images are hosted on image-shack. Perhaps your company has the site blocked. I'll look into other hosting options.

Thanks for your reply!

 

[mfgenggear]

Corypad & isuzu14b

My experience with parts of this nature with welding.
parts would crack without preheating at 300 deg F
this temperature will not effect the parts.
This is a stress relieve temperature.
it is stress at this temp. day in & day out with out issues.

you may may or may not induce cracking with out preheating.
I would recomend in favor of preheating.
it will not add much cost to do so.

 

[Robertmet]

I would suggest preheat and postheat at 400 F.The original part may have gotten a 400 F temper. In any case 400 F shouldn't have any detrimental effect .