Titolo della tesi: Mutual cooperation of experiments and numerical simulations for the design of fish-like swimmers
Aquatic animals have evolved a diversity of propulsive mechanisms to locomote effectively through
water. These mechanisms are a result of a long evolution whose requests have acted on propulsive
systems and has generated an array of novel anatomical and physiological responses to the problem
of moving through water.
Over the years, adaptive behaviours have developed in response to environmental requirements
where aquatic species live, resulting in engineering-level performances that are unparalleled. For
instance, consider the challenges of migration, such as the need to reach specific locations for spawning
and navigating with different water conditions. This demonstrates their efficient energy-saving
capabilities, allowing them to swim continuously for hundreds or even thousands of kilometres.
Conversely, when fleeing predators, some fish can reach speeds of up to 20 body lengths per second,
showing their ability to maximize efficiency as needed.
For these reasons, humans have begun studying fish locomotion due to the fascination with mimicking
marine animals. Understanding fish hydrodynamics has become of great interest in the last
50 years. This knowledge could potentially be applied to develop efficient bio-inspired watercraft
capable of specialized tasks such as reaching remote locations or aiding in aquaculture. Furthermore,
understanding aquatic animals’ locomotion mechanisms offers an appealing alternative to
conventional propeller models, which can be overly noisy and ecologically damaging.
Despite significant advancements in the study of fish self-propulsion in recent years ranging from
simple mathematical models to complex numerical solutions and bionic robots, some of the fundamental
mechanisms governing fish locomotion remain not purely understood and they certainly
require further investigation.
In this thesis, fish hydrodynamics is approached from both experimental and numerical perspectives.
Numerical analysis is based on an inviscid panel method and on a laminar Navier-Stokes
viscous code. The model initially considered consists of a 2D airfoil, either rigid or deformable, in
classic incoming flow conditions. Moving towards more physically meaningful conditions, we are
also considering free-swimming, where the fish-like body is able to move freely in the surrounding
fluid under the action of forces exchanged with it. Despite the simplicity of the model, this approach
qualitatively explains many kinematic and energetic aspects. These properties have subsequently
been used in the development of experimental models. We present two platforms, one to be attached
to a fixed measurement system and one fully autonomous, tested respectively in a recirculating tunnel
and in the naval basin of the CNR-INM. Although such prototypes are an approximation of
reality, they provide further insights with the respect to simple 2D numerical models, offering the
opportunity to better understand fish hydrodynamics and quantify their performance. By measuring
speed, forces, efficiency, and consumption, these experiments can ultimately serve as a basis for
practical applications in underwater robotics.