Brass Fitting Failures

[nlj]

I'll apologize for the very long post, but I've been working on this failure for a few months now and have gathered a lot of data.

We are experiencing a large amount of brass fitting failures as of late. These fittings have been in service since the 1980s and have been C 36000 brass from the very beginning.

The recent failures are intergranular in nature. The material meets requirements for C 36000 chemistry. The hardness is in the typical range for the material.

The failures have been coming from trucks that have been in outdoor storage for close to one year. The interesting thing to note is that there have been hundreds of trucks in the same storage location but only a random sampling of the fittings are failing (which leads me to believe that environment is not the root cause of failure). We also have fittings that have been in the field for 10-15 years that have not experienced cracking. We don't believe the fittings are being over-torqued as we have torqued some fittings to 3 times the print max and ended up stripping threads prior to replicating the failure.

Because we have had success with the material for years I have been trying to collect data comparing
1. Failed fittings from 2008-2009 time period
2. Non-failed fittings from 2008-2009 time period
3. Non-Failed fittings from the 1995 time period
So far I've only identified two main differences, beta phase presence and distribution and lead distribution.

I've now been trying to broaden my knowledge of brass. I've learned that the beta phase is often present in leaded free cutting brass and the percentage of beta phase increases with higher zinc content. I've also learned that the beta phase is harder, stronger and more brittle than the alpha phase. I'm seeing the beta phase in a network or conglomerate in the sample, not randomly dispersed, so I'm not sure how to measure the percentage of beta phase present.

I'm seeing the lead present on the grain boundaries, which I believe is typical. But the failed fittings have a different lead distribution when compared to the non-failed fittings.

Can anyone provide insight on how the beta phase or lead distribution affects the strength of the brass?

Also, if anyone has other suggestions for possible failure modes; I'm all ears. I've attached some micrographs of the fittings.

 

[TVP]

Intergranular cracking in brass is an indication of SCC, which also fits with your description of intermittent failures. SCC requires three things: a susceptible alloy, a source of stress (residual or externally-applied), and a corrosive environment. Take a look at the following links for more information on SCC of brass:

http://www.reitzmetallurgy.com/downloads/Aug05CaseHistoryReitz.pdf

http://www.copper.org/applications/industrial/designguide/scc03.html

http://www.google.com/search?hl=en&lr=&q=brass+thread+intergranular+SCC+filetype:pdf&aq=f&oq=&aqi=

 

[metengr]

Agree with TVP above. However, more investigation work is required because leaded brass can also develop intergranular failure from improper forming. I would look closely with the SEM and evaluate of the intergranular fracture looks to be from incipient melting due to lead along grain boundaries.

 

[nlj]

Thank you for the comments. I agree that at first glance, SCC would be the likely and I first assumed that. But why are some trucks, parked in the same lot, for the same amount of time, exposed to the same environment failing, while other truck parked right next to them don't fail?

I'll look more closely at the grain boundaries in the SEM.

Thank you!

 

[EdStainless]

I would not rule out SCC. These trucks may have been cleaned by different individuals, who used different chemicals. Or differences in microstructure (heat treatment) may make some fittings much more sensitive to it.
Keep looking for chemical contaminants in the cracked ones.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[swall]

Is it possible than transients are using the parking area for a bathroom? Decomposition of the waste can produce ammonia,leading to SCC of brass.

 

[TVP]

Keep in mind that it takes the combination of all three variables to produce SCC, so just because the vehicles are presumably experiencing the same environment, does not mean that they have equal SCC susceptibility. As EdStainless mentioned, microstructural variations due to heat treatment or residual stress differences due to processing (forming/forging, straightening, machining, etc.) may be the more important factor. Is it possible to perform a SCC test of the brass fittings from different lots, to determine if the SCC susceptibility has considerable variation?

My initial look at the microstructure images did not lead me to conclude that the beta phase amount & distribution were the proverbial smoking gun, but I am not a brass expert, so I won't take it personally if you choose to discount that statement. I would have some residual stress measurements made using x-ray diffraction before I would focus solely on the microstructure. Proto in Ontario, TEC in Tennessee, and Lambda in Ohio are all very capable companies that could provide this type of service.

 

[nlj]

Thank you all for the information. This definitely gives me more aspects to test. I'm going to test per ASTM A154 and ASTM B858 for SCC susceptibility. And potentially some x-ray diffraction for residual stresses.

Please see file for updated images.


Thanks!