Simulating dinosaur trackway formation

Palaeontologists recognise that they have a tough challenge when faced with determining how dinosaurs went about their day to day activities, for example how fast they could run and whether bipedal examples walked with the upright gait typical of museum displays. Apart from the occasional pile of petrified bones, the only other clues available are track-ways or fossilised footprints. Even these are contentious, with disagreement as to whether track morphology can be used to identify different species. As well as the morphology of the footprint itself, in the three-dimensional zone beneath, sedimentary deformations provide a snapshot of how the ancient sediment responded to loading. The nature of these plastic deformations depends on many factors - sediment composition, humidity and how the loading was applied. In soil mechanics, there are many plasticity models that could be applied to help back-figure the prevailing conditions and thus how the dinosaur walked. In essence, a footprint simulation would be no different from one carried out for a foundation design. However, there are many features that would make it perhaps more computationally demanding: Firstly the loading will have been applied obliquely and secondly the fine scale structure of the soil needs to be adequately represented. Both these factors mean that the problem is irreducibly three-dimensional. Capturing the level of detail required to compare field, laboratory and computational studies, means that plasticity problems with tens of millions of degrees of freedom must be solved. This paper looks at how parallel processing techniques developed for large-scale geomechanics problems may be used to provide further insight into dinosaur locomotion.