Ancient bear had strongest the bite (or is the analysis flawed?)

[Lee Margetts]

Here's a interesting story from the BBC website.

The largest bear that ever lived also had the strongest bite of any land mammal, say scientists. Agriotherium africanum was a giant short-faced bear that became extinct five million years ago. Reconstructions of the carnivore's skull revealed that it was well adapted to resist the forces involved in eating large prey. By comparing the skulls of several species, scientists also found polar bears to have surprisingly weak bites.

Click here to view full story

Its another example of the qualitative use of numerical methods: a technique called "comparative finite element analysis". As an engineer, I'm really not sure about the scientific validity of this technique. It basically involves comparing peak von mises stress values between different models of bones that are constructed as follows: (i) The bones, here a skull, are meshed as solids. (ii) The microstructure and possibly macroscale voids are completely "filled in". (iii) Then the bones to be compared are rescaled to have the same external surface area. (iv) Futhermore, the meshes comprise a single material that is given isotropic elastic material properties.

I'd be interested in the views of others. Does this sound unsatisfactory to you? Has anyone ever seen a paper published that looks at the validity of the technique or is it something that has caught on by accident? Using this technique, I would suspect that the bone with the largest volume of material would be the strongest and that no finite element analysis is necessary to reach the conclusions made by the authors.