Bronze ring failure - cracking

[kclim]

To do with the fabrication of a valve seat ring

The material in question is C62700 (admiralty bronze)

Approximate composition
Cu 80%, Al 10%, Fe 5%, Ni 5%, Zn 0.2%

All other elements below 0.1%

The product was in round bar form (extruded) with 39 mm diameter

The bar was then rolled into a ring, approximately 46 (~1200 mm) inches in diameter, with the ends being welded together.

The following describes the (post weld) heat treatment the whole ring was subject to
- Held at 650 C for 1 hour
- Quenched in water – 600 to 300 C (presumably quench to prevent gamma 2 phase formation)
- Slow cool (air) from 300 C to room temp

At this point cracks were found in the ring upon closer inspection.

Unfortunately, we could not determine when the cracks had arisen as no inspection had been performed up to that point.

The cracks were oriented circumferentially (around the larger ring, not the cross section) and went through the welded region but did not appear to be associated with the welding.

Tried performing some metallography – took a section which profiled the crack depth (which was more than 10 mm deep). The sample was etched in a 5% FeCl3/HCl/ethanol solution. See linked images:

General Microstructure

Showing cracking

Close up of cracking

From what I have researched & observed,

Microstructure appears to be consistent with aluminium bronze – alpha phase with interspersed kappa phase.

Cracks appear to be predominantly integranular (intercellular? between the islands of pure alpha phase).

What I am trying to determine is:

How long have the cracks been present?
Is it possible to tell, just from the microstructures? Were they present before the quench?

What is the damage mechanism?
I’m leaning towards either a quench associated damage mechanism (hydrogen embrittlement?) or a rolling defect-associated mechanism.

Any leads/comments would be greatly appreciated.

 

[Goahead]

To determine if cracks were present before quenching, usually one has to open them and visually determine the presence of oxides, from heating in furnace (provided it is an air furnace). If no oxides are present one deduce that cracks were not present before quenching.
A rolling defect should be present in the section also where the bar is not cracked.
Are the cracks near the weld? Was it flash butt welding?

http://www.welding-advisers.com/

 

[EdStainless]

Look in areas away from the weld. Do they look the same? Any cracks?

Looks a lot like SCC or LME.

Cut out a crack and open it clean. Look at it in an SEM and see what is on the face of it.

= = = = = = = = = = = = = = = = = = = =
Corrosion, every where, all the time.
Manage it or it will manage you.

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[swall]

kclim--Some bronze "bar" is actually cast bar rather than rolled or extruded. Are you sure you actually have extruded bar? If by mistake you got the cast variety, that could be the source of your problems.

 

[kclim]

Thanks for your replies. Wasn't sure I'd get a response given how long my initial query was?

In regards to your queries:
- the sample was taken well away from the weld, about 300 mm from the HAZ
- the weld was a double J prep, although I'm leaning away from weld related failure as there was an isolated crack completely contained in the parent metal
- the certificate claimed it was extruded; I just assumed the heat treatment lead to a cast type microstructure. I might see if I can get my hands on original material

Some SEM work sounds like a good idea.

Once again, thanks for your thoughts.