We combine mechanical compression testing of bone samples with high-speed photography, Atomic Force Microscopy and Scanning Electron Microscopy to capture and assess general failure behaviors in bone in real time. We also use these combined methods to visualize the differences in failure behavior of bone tested under various chemical conditions, and the failure behaviors of bones from healthy and diseased human specimens. We primarily test human bone, but we also test bovine bone for its similarities to human bone. Not only do we actually see how different types of bone fracture, we also can measure various parameters of mechanical integrity, including strength, toughness and elastic modulus. We test the bones at rates comparable to those experienced by people during falls.

Standard mechanical tests only deliver integral information on a bone sample and no information regarding local processes experienced in the elastic, yield and post-yield region. In order to close this gap, experiments have been devised to combine mechanical testing and imaging of trabecular and also cortical bone. Most of these approaches deliver 3D information. However, they all are somewhat time consuming and thus limited to quasi-static testing and/or recording of only a few different states of a sample subjected to mechanical testing. In contrast, high-speed photography is designed to record very fast processes, which uncovers the failure dynamics of bone samples subjected to mechanical testing in real time.

We mainly test trabecular bone with these methods, as opposed to cortical or woven bone. Trabecular bone is situated at the end of the long bones and in the spinal column, where it fills all of the inner vertebral space. In the long bones it transfers loads from joint faces onto the midshaft of the bone; in lumbar vertebrae, trabecular bone carries and transfers up to 90% of the applied load. Thus a change in trabecular bone quality can have a huge impact on strength and on fracture risk in vertebrae. Trabecular bone is also the type of bone most affected in osteoporosis and in aging.

Our high-speed camera can record up to 32000 frames/s but offers limited memory of 500 MB, so we usually record images at a rate of 1000 frames/s and a shutter opening time of 1/8500 s.

The first mechanical compression tester we used was custom-made for us by physics undergraduates at the College for Creative Studies at UCSB. It was designed to test 5x5x4mm cubes of trabecular bone. The tester has since gone through several prototypes, and now, combined with the high-speed camera looks like this:

Here are examples of some of the data we have gleaned with this tester:

Table 1. Mechanical parameters derived from recorded stress strain curves, for tests on vertebral bone from three donors.
E…elastic modulus, εy…yield strain, σy…yield stress, εf …failure strain, σf … failure stress, Ef…energy to failure

We are also developing a mechanical testing device for single bone trabeculae to operate with the high-speed camera, and another mechanical testing device to crush even smaller pieces of bone while being imaged under the Atomic Force Microscope.

So far, we have observed several fracture behaviors of bone that we are exploring in further detail, including:

- In compressed bone, apparent strain is made up of greatly differing strains at the local level as can be seen in high-speed photography movies.

- Highly deformed and microcracked bone regions of whiten, like the crazing seen in rigid inorganic materials when they deform, and similar to the stress whitening in polymers. The more strain that is applied to trabeculae, the more they whiten. This allows for easier detection of microdamage. We determined this with a motion energy filter coupled to our assembly, and by viewing the whitened areas with the Scanning Electron Microscope. Interestingly, this effect has been more pronounced in the healthy and osteoarthritic bone samples compared to the osteoporotic bone sample that we have tested so far.

- The primary failure mode in trabecular bone is delamination. This squares well with our general understanding that mineralized collagen fibrils are held together by a glue, through which cracks travel under high enough strains.

We have so far published one paper (Thurner et al. Mater. Res. Soc. Symp. Proc. Vol. 874, 2005) and have had two papers accepted for publication as of January 2006 on high-speed photography of bone fracture:

Thurner et al. High-Speed Photography of Compressed Human Trabecular Bone Correlates Whitening to Microscopic Damage. Submitted in January 2006, Engineering Fracture Mechanics.
Thurner et al. High-Speed Photography of the Development of Microdamage in Trabecular Bone during Compression. Accepted in January 2006, Journal of Materials Research