Automotive Failure Analysis

[leafsprung]

Hello all. I have been making custom transmission mainshafts for an adaptor I have been having them inertia welded. Recently one of the shafts failed. Obviously Im looking to prevent this from happening in the future and would welcome any analysis on the mode of failure as well as steps I could take to rectify this potential weakness.

 

[ad101]

Hi Leafsprung...

Where is the weld you mentioned in that picture? What is it made from, what conditions did it operate at and how long did it last..! Looks a right mess!

 

[dbooker630]

Leafsprung,

When you say inertia weld, is that like a friction weld where a rotating and a stationary piece are joined together? Please confirm the joining process.

If so, I would suggest that the fracture be examined by a Metallurgist. There may be oxides in the fusion zone that were not purged during the upsetting step. The met exam can also provide hardness and microstructure.

Also let us know if the shaft was mechanically straightened after welding.

 

[leafsprung]

ad101- I dont have the shaft back yet. Im hoping to get it back and have it examined. It was in a customers vehicle, I think it lasted 2000 miles or thereabouts. No real hard use. Obviously they were not happy and of course neither am I. From what I can tell it failed at the weld itself.

dbooker -The shaft was friction/inertia welded and your description of the process is accurate. Thanks for the tips, I will inquire about those steps and see if those could be an issue.

Thanks for all of your help and suggestions.

 

[Metalguy]

Are there any bending loads on this adapter? Do you have a local met. lab. that can perform the analysis?

Gold is for the mistress - silver for the maid
Copper for the craftsman cunning in his trade.
"Good!" said the Baron, sitting in his hall
But iron - cold iron is the master of them all.
Rudyard Kipling

 

[leafsprung]

The bending load would be applied as the gear on this shaft drives the intermediate or idler gear. However, the shaft is supported at both ends. Here is another picture of the shaft. I have not yet received the shaft back but this picture may give someone more knowledgeable than myself some clues. I am beginning to suspect that my supplier may not have inertia welded this shaft. Can anyone provide me with some thoughts that may support this hypothesis?

-Thanks
-Ike

 

[dbooker630]

A good way to prove your hypothesis is to have the shaft sectioned longitudinally through the fracture. As metalguy suggests, a local met lab can polish and etch to the section to look for evidence of a heat affected zone.

Also, can you tell us whether the shaft is heat treated after friction welding, if so to what hardness, and what type of finishing was performed - grinding, hard turning, phosphating, etc. Anything you can provide will help.

 

[leafsprung]

Looking at the first image I posted the shaft was apparently machined for alignment. It appears to have been welded conventionally as there is not penetration into the center of the shaft. Seems to me the weld would be 100% with friction welding and there would be no need to machine the shafts for alignment as the fixture would provide the alignment.

 

[dbooker630]

Yes, with a proper friction weld 100% fusion will be achieved. You would have to remove the curl if it was friction welded, and perhaps drill the hole after FW as well. There still is the possibility of mismatched components if the fixture is not set up correctly.

 

[redpicker]

[Minor nit-pick]

Friction welding is a non-fusion welding process. It is technically impossible to achieve 100% fusion if you, technically, aren't fusing the material.

[/Minor nit-pick]

But, yes, you should achieve complete bonding with a friction weld. It if obvious that the failed part was fabricated with a partial penetration fusion weld. Based on the apparent porsity visible, I'd guess FCAW, but could be SMAW or GMAW. Obviously, the weld was not up to the task.

If a friction weld isn't avialable, I'd recommend a TIG weld, with a filler metal to match the base material and a PWHT if needed to bring the HAZ back down to the base materal hardness. You'd also want to locate the weld at least 1/2" away from the end of the splines to keep the stress concentration from the splines away from the weld.

rp

 

[Metalguy]

The original photo shows good macro evidence of fatigue cracking with multiple initiation points--that's why I asked about bending stresses.

Be real interesting to see the results of a microhardness traverse along the shaft, including areas beyond the weld area in both directions. What is the shaft alloy and hardness?

Gold is for the mistress - silver for the maid
Copper for the craftsman cunning in his trade.
"Good!" said the Baron, sitting in his hall
But iron - cold iron is the master of them all.
Rudyard Kipling

 

[patdaly]

Been a while since I worked on a failure, but I have to go along with Metalguy. From the photos, it appears as if the first shown has multiple initiations, and is 180 out from the second photo.

We had a rash of shaft failures immediately above a spline very similar to your photo, turned out to be field guys turning the air pressure up 2X the recommended.