Site Summary

Researchers at Virginia Tech, the Stevens Institute of Technology, and the University of California-Irvine have worked together to improve methods for modeling and controlling locomotion in water and air using methods that mimic biological locomotion. The effort proceeded in two phases.

For the first phase (NSF Grant No. CMMI-1435484), the over-arching goal was to discover and disseminate new methods for optimal, integrated design of the morphology and gaits of biologically inspired atmospheric and aquatic vehicles. Specific objectives supporting this goal were to:

  • Identify and develop dynamical system models of appropriate complexity -- those which are general enough to represent a large class of biomimetic vehicle systems but which are amenable to geometric control and averaging methods -- and assess the validity of these models by comparing them with diverse examples from biology.

  • Construct a taxonomy of design optimization problems for biomimetic locomotion in terms of relevant metrics (e.g., maneuverability; robustness; power consumption) and design parameters (e.g., shape, placement, and articulation of appendages; shape, frequency, and amplitude of internal waveforms).

  • Using analytical techniques from geometric control and generalized averaging theory, within a symbolic computational environment, address selected design optimization problems within the given taxonomy. These problems will be selected to support comparisons with data available in the current biology and engineering literature.

  • Within the symbolic computational environment used to address the objectives above, create tools for visualizing the optimization process and points in the high-dimensional design space (i.e., specific vehicle morphologies and gaits). Working within a mature, widely used software environment will increase the accessibility of these tools among the research community and will increase the available options for exporting to web-based environments.

In the second phase (NSF Grant No. CMMI-1635143), the major goal was to incorporate an experimentally validated unsteady flow model into a geometric control framework for the design and control of pisciform swimmers. The collaborative effort built upon expertise and resources of the partner institutions to strike a balance between fundamental analysis and physical and computational experiments. Specific objectives supporting this goal were to:

  • Develop a reduced-order, state space model for the unsteady hydrodynamic force generated by a flapping appendage (e.g., a caudal fin) for use in a geometric control framework.

  • Use the geometric control framework, including the unsteady force model, to derive optimal gaits for select configurations: minimum power propulsion, maximum thrust propulsion, and maximum turning moment.

  • Design a modular biolocomotion emulator (MBE) for use in the Virginia Tech and Stevens tow tank facilities and execute an experimental test program to support the following three tasks:

    • Assess the utility of the unsteady vortex lattice method (UVLM) for predicting unsteady forces and moments.

    • Identify and validate the state space unsteady hydrodynamic model over a range of geometric and motion parameters.

    • Validate optimality of the gaits obtained through analysis for selected configuration.

Low-order, control-oriented models of biomimetic locomotion support rapid vehicle and gait design, providing suboptimal solutions that may be refined within a slower, more expensive design process that incorporates higher fidelity multi-physics models.