I've been asked, "Do you ever fail to find the answer to a case?" "Do you ever blow an analysis?" The answers are: "all too often" and "seldom, but not seldom enough." In this case I almost accomplished both.
Some 15 years ago a tractor-trailer rig was traveling on an interstate highway north of Boston. A dual wheel came detached from the trailer and played havoc with other vehicles. An innocent life was lost. All ten of the studs holding the wheels to the truck had fractured just inboard of the inner wheel. I was hired by a plaintiff's lawyer as a metallurgical expert to figure out why the studs had fractured.
The wheels were held on by so-called "Budd Nuts," which are a double-nut construction that holds the wheels together even if the studs fracture. As such, the outboard part of the fractured studs remained with the wheels, rather than coming loose at the time of fracture and probably being lost.
The state police had the wheels at a barracks north of Boston, so my scanning electron microscope expert and I drove up to take a look. The wheel was stored outside, probably for space reasons. It could have been stored in two ways: With the fracture surfaces down and protected from the elements, or up and exposed. The latter was the case and the fracture surfaces were badly rusted. Rust is the bane of failure analysis of ferrous materials.
We were not allowed to remove the wheel from the barracks, so we "replicated" the fracture surface with a dental rubber normally used in taking oral impressions. We mostly replicated rust, some of which came off with the rubber.
We took our impressions to the lab, shadowed them with an electrical conductor, and studied the surfaces under the scanning electron microscope (SEM). Many readers may be familiar with the SEM, the rest are referred to earlier "Calamities" columns, in particular "The Case Of The Misplaced Trust" (http://rbi.ims.ca/3851-534).
Most dynamic failures are by fatigue, which is fracture under cyclic loading below the one-time fracture stress of the material. The studs were thus good candidates for fatigue failure. We wanted to find out if the metal surface showed the striations usually characteristic of fatigue. What we mostly saw in the SEM was the replica of the rust.
My client understandably wanted to know the failure mechanism, but all I could do was hem and haw. The SEM had showed some vague hints of fatigue, but not nearly enough to form an opinion, certainly not one that would stand up under a rigorous cross examination. By this time I had run up a fair-sized bill but had little to show for my efforts.
Then luck struck. Someone had removed the inboard ends of the studs from the brake drum and protected them from damage. The figure on this page is a photograph of one of these fracture surfaces at about 1:1 magnification. Beach marks are visible above and below the oval region in the middle. These marks represent pauses in the cracks as they propagated. The central oval region failed suddenly and in a ductile manner after the strength of the studs had been reduced by the fatigue cracks. The other nine fracture surfaces showed similar marking. The studs had failed by fatigue in reversed bending.
But why would all ten studs undergo fatigue failure? Operating overload or very rough terrain could have caused the failure, but there was no record of such use. Then my client asked if I had noticed the crack from one bolt hole to the center hole of the wheel. OOPS! I was so concentrated on the fractured studs, I had completely missed the crack.
Crack arrest "beach" lines are clearly above and below the oval region in the middle of the SEM image of one stud's fracture surface.
There was no way the nuts could be tightened properly on the cracked wheel. The wheel should have been pulled out of service the last time it was dismounted. I rather suspect that a series of service people missed the crack, much as I did.
I gave an extended deposition, but the case settled before trial. I believe that the heirs of the deceased got a substantial amount in settlement.