Hi Sean,
I'm not sure if these comments will be enough to narrow down the cause but here goes some brain-storming. A bit of hand-waving can be expected.
-Did the material properties measured fall within the "typical" values? Not asking whether they met minimums, but whether they were remarkable. The recent discussion of unusually high tensile values in a 15-5PH comes to mind. Are there tracking statistics on this material, historical data to compare with?
-When a tensile bar necks down quite substantially there is a very high state of triaxial stress. I don't recall ever seeing quite as regular a set of 90 degree splits but I think I saw some with internal nearly vertical flat faces within the fracture zone, sometimes called a star pattern. Which leads me to wonder if there could have been a plane of weakness along the material axis due to an alignment of inclusions or perhaps an unusual microstructural feature from an imbalance of composition or heat treatment. A planar arrangement of ferrite, or nonmetallics, where the high triaxial stresses then caused a tensile failure across such a plane that had lower ductility (local / directional fracture toughness, essentially)? An unusual microstructure from a borderline failed heat treatment, or an improper temperature during forging?
-Another commenter asked about machining. The near-surface residual stress state, if there was some abnormality in the machining process, can be affected. Could the surface have ended up being "burnished", similar to a residual stress improvement treatment, due to problems in an automated specimen prep routine? Or the opposite, an excessive depth feed or bad machining tool contour causing micro-tearing that wasn't apparent after the final specimen was polished down with emery prior to testing?
It depends on how far you want to go. You could look for asymmetry in the specimen's hardness by doing a ring of indentation hardness tests around the specimen circumference, that would be pretty inexpensive. Magnetic particle inspection to see if there is a continuation of any sort of feature beyond the cracked lines, fairly inexpensive. You could look for microstructural issues by doing metallography, at a slightly greater expense. A quick review of furnace charts and chemistry would seem warranted, particularly if the tensile results were remarkable in any way.
If you come to any conclusion, it would be interesting to hear about it.
------------------------------
Paul Tibbals
------------------------------
Original Message:
Sent: 05-12-2021 11:42
From: Sean Piper
Subject: Unusual fracture pattern in F91 tensile test?
We recently experienced an interesting fracture pattern on a tensile that we pulled, see attached pictures. Overall it was a typical, ductile cup/cone but the strange thing is that the perpendicular fracture face included a second, X-shaped fracture parallel to the direction of loading. As you can see, it's fairly centered in the tensile and the fractures forming the X are essentially perpendicular to each other. My lab actually took some pictures of this tensile on the machine before it broke where you can see the fractures breaking through to the surface just before rupture.
I suppose what I am wondering is why the stresses would distribute such to cause a fracture to open up parallel to the direction of loading like this. I understand that ductile shear tends to happen at 45 degrees and brittle fracture tends to happen at 90 degrees, but to me this implies that the greatest stress (at least relative to the material's strength in that direction, which could be anisotropic) was circumferential tension, like hoop stress. Shouldn't the surface be in circumferential compression?
A few pieces of background information on this:
- My lab tells me that this is not the first time they have seen this on F91, or for that matter F22. The reason they took pictures and brought it to my attention this time around is that the sample did not break on its own; they needed to manually bump the force up a smidge at the end to get it to rupture. My presumption here is that, since the machine is constantly adjusting its force to produce a constant crosshead speed, it probably overshot a bit and started applying too little force to get it to break.
- I have no reason to believe that there is an issue with our tensile machine where it's exerting a significant torsional component or anything like that. We get it calibrated regularly and we test plenty of other materials on it which don't show this fracture pattern.
- The material is standard ASTM A182 F91, normalized and tempered. We forge a test bar to about 2:1 reduction and take the tensile in the longitudinal direction at T/4. Aside from the aforementioned behavior at rupture, the stress/strain graph was perfectly normal, with typical values for YS/UTS/elong/RoA. I have no reason to believe there were any gross inclusions or anything anomalous like that in the material.
------------------------------
Sean Piper
Product / Process Metallurgist
Ellwood Texas Forge Houston
Houston TX
773-524-8985
------------------------------