The Case of the Buckling Brace
A titanium brace succumbs to fatigue failure, reminding engineers that “one-size-fits-all” isn’t always a good strategy when designing parts
By Contributing Writer Ken Russell
Titanium is a design engineers’ dream and a process metallurgists’ nightmare. The strengths and ductilities of titanium alloys are comparable with those in steels, but the density is over a third less. This weight savings is crucial in aerospace applications where high-strength titanium forgings made jumbo jets possible.
Titanium, like aluminum and chromium, reacts with air to form a very thin protective oxide film. The resulting corrosion resistance makes titanium alloys competitive with stainless steel for many applications.
Unfortunately, titanium is nearly a universal solvent and dissolves almost everything it contacts. This reactivity makes it very difficult to process. Even though titanium ores are fairly abundant, the refined metal is expensive. Titanium costs about as much as premium steak; stainless steel costs about as much as peanut butter.
This case involves the failure of a titanium alloy leg brace. The alternative materials would have been stainless steel or aluminum alloys. Apparently, the lighter weight of titanium made it preferable to stainless steel. Titanium is much stronger than aluminum, so it could be used to produce a less bulky brace. The brace broke at the knee joint, causing the wearer to fall and suffer serious injury. I was retained by one of
the defendants in the case. We were limited to strictly nondestructive testing, and not even a minimally invasive hardness test was allowed.
I suspected fatigue failure, but the parts of the brace were too large to fit in the scanning electron microscope (SEM). (See “The Case of the Jelly Roll Blues.) We had to use replication of the surface to study the fracture. A plastic dental replicating compound was spread on the fracture surface and allowed to harden. The replica was then shadowed with a very thin layer of metal to make it electrically conducting, and studied in the SEM. Study of the replica revealed the fine striations characteristic of fatigue failure.
Fatigue failure occurs under cyclic load. A crack starts at an irregularity known as a stress riser and proceeds little by little with each stress cycle. Finally the crack weakens the piece enough that sudden catastrophic failure occurs. In this case, walking provided the cyclic stress and the sudden final failure caused the injury.
This case was peculiar in that both sides of the brace had undergone the same amount of fatigue damage. Most fatigue failures are started by nicks or dings that are unlikely to occur on both sides of the brace. But the brace clearly failed, and my job was to figure out why.
Could the brace have been made of faulty titanium? The The brace was made of rolled strap, and rolled material is rarely defective. The material in the brace was a very small part of a large batch of strap. If one piece was defective, the whole batch probably was. It appeared that the brace may have been under-designed. I asked my client if the plaintiff was large. The answer came back yes, she weighed some 300 lb. She was one of the last of the pre-Salk vaccine polio victims.
The case settled shortly after my study and I never learned any more about the matter. My suspicion, though, was that the brace was designed on a “One Size Fits All” basis and was suitable for some “average” 150-lb user. Such a brace would not be suitable for a user twice this weight. On the other hand, a 90-lb user would be carrying around a heavier brace than she needed.
In an earlier investigation, The Case of the Lamed Lothario, the plaintiff had a reputation as a bedroom athlete. One defense lawyer suggested that amatory activity might have caused the failure. I concluded that such was likely to be the case then. Such a cause is even less likely now in that the plaintiff is a woman.
Contributing Writer Ken Russell (kenruss@mit.edu) is Professor Emeritus of Metallurgy and Nuclear Engineering at MIT. He specializes in physical metallurgy, forensic metallurgy and failure analysis. Cases presented here are drawn from his actual forensic files.
Mahlon R. Hoover commented:
There are a lot of disigners out there that do not understand S=P/A; and they do not have the slightest understanding of S=MC/I.
just me commented:
What this realy amounts to is some computer jocky doing limited tests, and thinking simulations were the same as real world testing.
A brace designed for a 150# person shouldn't fail until many times the 150# bodyweight is aplied to it.
In truth a heavyer persom mite actualy put less of some types of stress on the leg brace as they would most likely move slower and not walk as far.
In a real world case, many people who need braces are people such as those who have had a stroke, etc. ware they may move in other than normal ways and may catch themselves on the brace rather than falling, so it would need to hold much more than just thier bodyweight, just like a persons legs and feet are exposed to several times bodyweight when walking, climbing stairs, etc., plus there are torsional stresses and various angular stresses, etc.
Also the brace is very likely to be bumped against all sorts of things, so that should have also been taken into consideration.
I'm sure what realy caused the failure is some "comp-u-neer" cutting the specs way too close to some unrealistic computer simulations that don't take into acount all the factors involved.
He was probably making the accountants happy by cutting costs on a product that was probably sold for many hundreds to thousands of dollars, and actualy cost realativly little to build (I would guess like $100 or less, as there ain't all that much metal in a leg brace, I'll also bet it was made of scrap titanium, which should also be taken into consideration because of posible impuritys).
Also since no destructive testing was done on the brace after failure, which would determine the metal in the braces quality, and could have deturmined if the design was valid in a real world use, I suspect the manufacturer already knew of the problems before hand and thats why they wouldn't allow proper testing.
terdburd commented:
As a very sturdy (280 lb, 6' tall) and extremely active user of a prosthetic leg, I can attest to the fact that the industry doesn't provide for people heavier than 250 lbs. It is a sad commentary on an industry that accomodates an artificial "norm" and ignores a growing segment of the population that could lead fuller and more physically active lives at expense of several ounces of precious metal.
I too, have experienced a similar failure to a 330mm titanium tube, but fortunately, only suffered injury to my dignity (my foot fell off in mid-stride). Suing the prosthetist wasn't the right thing to do, but the woman did deserve to have her brace sized properly.
JT commented:
It seems to me that there's too little mat'l here for an article.
dw commented:
This entire post and "analysis" seems to be a case of "facts not in evidence", with a huge dose of conjecture. What was the general failure rate for the product? What, if anything, did the specifications about applicability for obsese persons? Who much did she weigh when the brace was first fitted? Someone from MIT should be very aware of bad science and the limitation of a sample size of one. MAYBE it was a bad design. There is NOTHING here to "prove" that.
JS commented:
The next to last sentence: "I concluded that such was likely to be the case then." Should that be "I concluded that such was unlikely to be the case then." ? The next sentence "Such a cause is even less likely now..." seem to follow that better.
tiguy commented:
Just curious... Do you know where the raw titanium material came from?
