Session F22: Biological Fluid Dynamics: Locomotion Swimming - Fishes II

Fish locomotion studied using an active-swimming soft-robotic model

By attaching soft robotic actuators (pneunets, ‘PN’) on a flexible passive foil, we constructed an active swimming model for fish locomotion called the ‘pneufish’. Pneunets are a series of hollow, connected chambers, molded from silicone rubber (Dragon skin 20, Smooth-On Inc., Easton, PA 18042, USA). When pressure is increased, each chamber expands and pushes against its neighboring chambers, resulting in a net lengthening and curvature of the PN. We assembled several pneufish and suspended them in a recirculating flow tank attached to an ATI 6-axis force-torque sensor. We measured thrust, lateral forces, and amplitude of trailing edge oscillation, investigating the swimming performance of the model at a variety of frequencies, flow speeds, foil stiffnesses, and air pressures. Results revealed the pneufish generates positive thrust at several parameter combinations, and while stiffness and activation frequency are important parameters, the interaction of the two are more important. Additionally, we have expanded the model from a simple two-pneunet apparatus (duo-pneufish) to a four-pneunet system consisting of two consecutive PNs on each side of the foil (quad-pneufish), and plan future experiments to evaluate the performance of this multi-segmented system that is more fish-like.   

Optimal Caudal Fin Kinematics at moderate Re numbers

We investigate the optimality of caudal fin kinematics inspired by certain fish species such as tuna.  We consider caudal fin moves that are out of phase with the anterior section of the body. We investigate the benefits of spatially discontinuous fin-kinematics at moderate Re, by coupling three-dimensional simulations of fish-like swimmers with the Covariance Matrix Adaptation (CMAES) optimization algorithm. The optimization suggests kinematics similar to those observed in natural swimmers at higher Re numbers. At moderate Re, optimal kinematics lead to improved performance with regard to both the peak and average swimming power. However, a similar advantage is not observed for fish swimming at low Re, where a continuous sinusoidal wave is found to be preferable.    

Study of the thrust--drag balance with a swimming robotic fish

A robotic fish is used to test the validity of a simplification made in the context of fish locomotion. With this artificial aquatic swimmer, we verify that the momentum  equation results from a simple balance between a thrust and a drag that can be treated independently in the small amplitude regime. The thrust produced by the flexible robot is proportional to A^2 f^2, where A and f are the respective tail-beat amplitude and oscillation frequency, irrespective of whether or not f coincides with the resonant frequency of the fish. The drag is proportional to U^2, where U is the swimming velocity. These three physical quantities set the value of the Strouhal number in this regime. For larger amplitudes, we found that the drag coefficient is not constant, but increases quadratically with the fin amplitude. As a consequence of this correction, the Strouhal number becomes an increasing function of the fin amplitude as well.

  

How fish power swimming – a 3D computational fluid dynamics study

In undulatory swimming of fish, muscles contract sequentially along the body to generate a bending wave that pushes against the water and produces thrust. Here we use 3D computational fluid dynamics coupled to the motion of the fish with prescribed deformation to study the force, torque, and power distributions along the fish's body. We find that forces on the bodies of both the anguilliform swimmer and the carangiform swimmer are dominated by reactive forces, and furthermore, the force on the caudal fin of the carangiform swimmer is dominated by drag-like forces. The torque exhibits a wave pattern travels faster than the curvature wave in both the anguilliform and carangiform swimmers but the wave speed is even higher for the carangiform swimmer. The power output for the anguilliform swimmer is concentrated on the anterior half of the body and is significantly negative on the posterior side of the body. In contrast, most of the power is generated by the posterior part of the body before the peduncle for the carangiform swimmer. The results may explain the differences in the observed electromyography patterns in fish with different swimming modes.   

Two-phase flow large-eddy simulations of jumping archer fish hydrodynamics and aerodynamics using detailed motion kinematics

The archer fish’s impulsive vertical jumping behavior is associated with rapid upward thrust, produced from an initial stationary position below the free surface, followed by gliding. We herein integrate detailed jump body and fin kinematics with two-phase flow large-eddy simulations to investigate the hydrodynamics of oscillatory body kinematics and the aerodynamics of C-shaped body bending after water exit. Our simulations enable accurate prediction of forces exerted by the fish and elucidate the emergence of intricate 3D vortical structures. In water, we investigate the effect of the Reynolds and Strouhal numbers, which cannot controllably vary in experimental studies on live archer fish, on the shape, size and strength of the emerged turbulent coherent structures. We further compare the vortex loops generated by the caudal fin during in-water oscillatory motion. Braided hairpins, whose heads and legs appear to have similar thickness, constitute patterns that are also found in the wake of in-water C-start accelerations. Rich vortical structures and square-shaped vortex rings are formed after the fish leaves the water.   

Ontogeny and scaling of burst-and-coast swimming in zebrafish

Swimming kinematics of small fish such as zebrafish are characterized by intermittent sequences consisting in an active swimming phase directly followed by a passive coast phase. These specific sequences are based on a coupling between sensing and decision: fish use the passive time to sense their environment and prepare their next move. Fish essentially use vision and the lateral line system to see and sense their surrounding environment. The mechanisms that govern this “sensing” to “decision-making” (S2D) process are still to be understood and detailed. 
This work is an attempt to characterize these sequences using several archetypal
model experiments and models gathering hydrodynamics, statistics and behavioral sciences.
We will focus on new results obtained from real fish experiments in free swimming and forced gait configurations (using a controlled swimming channel). We will show how the statistics of
these S2D sequences evolve with the conditions of the experiment; the important parameters
being here the external flow conditions and the size of the habitat, but also the species and
the maturity state of the fish (larva, juvenile or adult). We believe that these results will have
direct implications on the design and implementation of biomimetic robotic systems.   

Three-dimensional Tracking of Fish inside a Large School at the Rotterdam Zoo

The three-dimensional motion of schooling of fish is a startling display of collective behavior that continues to intrigue many scientists including fluid-dynamicists. Our current understanding of the physical and social principles underlying to the emergence of coordinated group motion is limited due to a lack of three-dimensional experimental data. In this work, we perform three-dimensional tracking of a school of 1500 Harengula Clupeola (false herring) at the large-scale ocean aquarium of the Rotterdam Zoo. This aquarium is a unique facility that reproduces a tropical ocean ecosystem where the fish swim in an unconstrained environment that includes multiple species of predators. From our tracking data, we gain detailed knowledge of the kinematics and spatial organization of fish inside the large school. Using the measurement of the position and velocity of the individual fish, we study the morphology and group dynamics of the school. Our data provides new insight into the unconstrained and three-dimensional dynamics of schooling fish; we will discuss its relevance to other active biological systems.   

Flying underwater in the optimal Strouhal number range

The Antarctic shelled pteropod (Limacina helicina antarctica, or “sea butterflies”) swims with a pair of parapodia (or “wings”) via a unique flapping propulsion mechanism that incorporates similar techniques as observed in small flying insects.  It is highly unusual to observe a “flying” aquatic organism, especially in the intermediate Re regime.  The pteropod achieves this locomotive behavior by hyper-pitching its body to effectively generate upward thrust during both the power and recovery strokes.  For the specimens observed in this study, the shell size range is 1.8 – 4 mm, the swimming velocity is 14 – 30 mm/s, and the shell pitches forward-and-backward at 1.9 – 3 Hz.  The non-dimensional variables characterizing the motion of swimming pteropods are flapping, translating, and pitching Re (i.e. Re_f, Re_U, and Re_Ω).  The flapping and translating Re may be combined to form a Strouhal number based on the wing stroke amplitude, A.  The results reveal that L. helicina antarctica swims in the optimal power efficiency range of St_A between 0.2 and 0.4, which is consistent with many taxa.  L. helicina antarctica are threatened by ocean acidification, and the concern is that their shell degradation may adversely affect their ability to swim in the optimal St_A range.