1045 stress corrosion cracking

[Tmoose]

No mention of 1045 being susceptible to SCC in MIL HDBK 5.
A lab has proclaimed a Normalized 1045 shaft failure as SCC.

Are we justified in going farther?

thanks

Dan T

 

[unclesyd]

Can you comeback with some environmental conditions and physical properties if the metal?

 

[CoryPad]

This is a good document regarding SCC susceptibility:

http://standards.msfc.nasa.gov/documents/detail/3314945

1045 may or may not be susceptible - it depends on the material and environment details.

 

[unclesyd]

My experience with 1046 CS has been limited to drill bits but have a lot with 1035 CS. We used 1035 for many mechanical components on our site. We suffered several failures from using said components in handling high molecular weight Amines along with failures from a NaOH dry down problem in steam service. The failures were both SCC or in case of the steam a mixed brew. The resolution to these problems was to require a high temperature stress relief on all components made from 1035 CS regardless of intended service.
Based on the similar chemistry I would definitely expect 1045 to respond the same way.

 

[Tmoose]

The fracture occured on the clean, oily side of the seals in a pillow block housing for a spherical roller bearing. Circulating oil system.
The shaft was a bit rusty, which was surprising. No available eye witnesses for the condition of the bearing.
Strong solutions or atmospheres of caustic etc very unlikely. It >should< be a pretty nice place for a shaft to be.

thanks

Dan T

 

[CoryPad]

How did the shaft rust? What is the composition of the oil? They can have Cl^- and other corrosion promoters in them.

 

[metengr]

Umm... roller bearing, shaft rotation, pillow block, sounds like a mechanical fatigue failure of the shaft to me.

 

[Tmoose]

It's motor driven with a Falk Grid type coupling. The only reversing radial load (relative to the shaft) is coupling hub weight, plus whatever restoring force the coupling exerts when transmitting torque misaligned radially a few thouands of an inch. Most folks consider the torque nearly constant.
No varying forces, no fatigue.

Too make this tougher, two different metal labs have looked at the fracture surfaces. One says a brittle fracture with no signs of fatigue. One said stress corrosion ( some texts describe the fracture surface from SCC as the same as a brittle fracture). I would not have come to that conclusion, but I'm no metallurgist. A third lab will be checking it out next week with an SEM. I have asked that one of the principal analysts do the peeking.

 

[swall]

First, determine origins and type of fracture, i.e. torsional, tensile or bending. Then determine fracture morphology at the origin and else where. Then look for notches and flaws.Once you have this, you can begin to formulate a failure hypothesis. I once saw a CV joint stem fail due to rotary bending fatigue--there should be no way in heck for that to happen, but it did. Sometimes you will get an inexperienced person at a lab that will rush to put something on the SEM, get some fracture morphology shots and jump to some weird conclusion.

 

[Tmoose]

A long time experienced manager here offered this before any analysis was even done. " 4 metallurgists, 4 opinions"

 

[swall]

Which is why you have to ask the questions suggested by my approach. As an in house metallurgist, I have to refute stuff all the time from inexperienced metallurgists from outside labs.

 

[unclesyd]

I think we should add some opinions to the mix, can you post a picture of the fracture surface?

 

[CoryPad]

And the material condition - hardness, surface treatment, etc.

 

[metman]

I once saw a CV joint stem fail due to rotary bending fatigue--there should be no way in heck for that to happen, but it did

A CV joint does apply a bending load to the shaft(s). It might be a little difficult to visualize but the torsional load is transmitted thru an angle which of course is the whole purpose of the joint.