The thesis at hand is part of a research project that attempts to study
and develop vibrissa inspired tactile sensors for object and fluid flow
detection. The main focus of the thesis is on the development of a model
for a vibrissa-like sensor for obstacle contour recognition under fluid
loads. To this end, a mechanical model -- based on the non-linear
Euler-Bernoulli beam theory -- is established. The model includes the main
characteristics found in a natural vibrissa, such as elasticity of the
base, that acts as the vibrissa follicle; the intrinsic curvature; and
conicity. The characteristics are represented as parameters of the model.
The model is subjected to a contact load and a fluid flow load, represented
by a concentrated load and a distributed load, respectively. Then, the
model is transformed into a dimensionless representation for further
studies to achieve more general assertions. A variation of the magnitude of
these loads, as well as the vibrissa parameters is also analyzed. A direct
numerical approximation using the finite difference method, along with the
shooting method, is used to obtain a solution of the model. Subsequently,
the model is used to simulate an ideal contact between an obstacle and the
vibrissa. This simulation considers a quasi-static sweep of the artificial
vibrissa with the contour of a profile, while measuring and recording the
forces and moment at the base. This procedure is then repeated in
combination of a distributed force acting on the vibrissa, simulating the
effect of a fluid flow. Two types of contact phases are identified and the
conditions for each one are set. Finally, the measured quantities, which
represent the observables an animal solely relies on, are used to obtain
the magnitude of the fluid load and to reconstruct the profile contour of
the obstacle. The developed model is used again for the reconstruction, an
analysis of the observables is performed to identify and predict which
contact phase the vibrissa is in. The results successfully show
identification of the fluid flow load as well as reconstruction of the
profile, the difference between the reconstructed profile and the original
profile is then calculated as a measure of reconstruction quality.