Size does matter: The use of fish motion for improving human swimming simulations

In this study we evaluated the length and volume inconsistencies that occur when applying simple motion functions onto surface models of swimming fish and human, respectively. The larger the size (diameter and length) of the body, the larger are these inconsistencies. To eliminate these errors, different surface correction algorithms are introduced. Additional terms are implemented to overcome structural insufficiencies through over-stretching and overlapping in sensible body parts. For human undulatory swimming these models combine the motion of rigid extremities with the smooth surface bending at joints. The advantages of this hybrid model can be crucial for Computational Fluid Dynamics, where deviation from natural behaviour results in unrealistic surface pressures and artificial vortex structures. The structural improvements are also promising for computational graphic programming, where skin-like stretching surfaces mediate a natural impression.