Search

Analytical models of avian flight have previously been used to predict mechanical and metabolic power consumption during cruise. These models are limited, in that they neglect details of wing kinematics, and model power by assuming a fixed or rotary wing (actuator disk) weight support mechanism. Theoretical methods that incorporate wing kinematics potentially offer more accurate predictions of power consumption by calculating instantaneous aerodynamic loads on the wing. However, the success of these models inherently depends on the availability and accuracy of experimental kinematic data. The predictive simulation approach offers an alternative strategy, whereby kinematics are neither neglected nor measured experimentally, but calculated as part of the solution procedure. This thesis describes the development of a predictive tool for simulating avian wingbeat kinematics and wakes. The tool is designed in a modular format, in order to be extensible for future research in the biomechanics community. The primary simulation module is an inverse dynamic avian wing model that predicts aerodynamic forces and mechanical power consumption for given wing kinematics. The model is constructed from previous experimental studies of avian wing biomechanics. Wing motion is defined through joint kinematic time histories, and aerodynamic forces are predicted using blade element momentum theory. Mechanical power consumption at the shoulder joint is derived from both aerodynamic and inertial torque components associated with the shoulder joint rotation rate. An optimisation module is developed to determine wing kinematics that generate aerodynamic loads for propulsion and weight support in given flight conditions, while minimising mechanical power consumption. For minimum power cruise, optimisation reveals numerous local minima solutions that exhibit large variations in wing kinematics. Validation of the model against wind tunnel data shows that optimised solutions capture qualitative trends in wing kinematics with varying cruise speed. Sensitivity analyses show that the model outputs are most affected by the defined maximum lift coefficient and wing length, whereby perturbations in these parameters lead to significant changes in the predicted amount of upstroke wing retraction. Optimised solutions for allometrically scaled bird models show only small differences in predicted advance ratio, which is consistent with field study observations. Accelerating and climbing flight solutions also show similar qualitative trends in wing kinematics to experimental measurements, including a reduction in stroke plane inclination for increasing acceleration or climb angle. The model predicts that both climb angle and climb speed should be greater for birds with more available instantaneous mechanical power. Simulations of the wake using a discrete vortex model capture fundamental features of the wake geometry that have been observed experimentally. Reconstruction of the velocity field shows that this method overpredicts induced velocity in retracting-wing wakes, and should therefore only be applied to extended-wing phases of an avian wingbeat.

Jumping is useful as a precursor for flight and arboreal locomotion, as well as locomotion through intermittent terrain. This thesis aims to answer the questions: What insights can be gained through the application of predictive modelling methods to novel contexts in jumping dynamics problems, and how should the fitness of different predictive modelling techniques be assessed? This thesis uses the metrics: predictive power, computational cost, and intellectual investment, to assess predictive modelling methods for use in jumping research. Analysis of model equations is rarely found applied to multiple degree of freedom jumping models in the literature. This work explores such applications of analysis and determines that the method is capable of providing insights into low and high complexity models. The work identifies static optimisation for the resolution of second order kinematic redundancy and reinforcement learning as predictive modelling methods which are not currently used in jumping research. The results of this work imply that static optimisation could be made viable for providing insights into jumping dynamics problems. However, the static optimisation method is demonstrated to produce results which are comparable to experimentally measured data found in the literature and may be a useful method in jumping applications. The method is limited in that it does not accommodate the physical constraints of jumping systems. Deep Deterministic Policy Gradients, a reinforcement learning method, is shown to also violate physical constraints. The method is demonstrated to be unreliable when applied to jumping dynamics problems with 32% of optimisation runs failing to achieve any positive rewards. Dynamic optimisation is assessed on its predictive power, computational cost, and intellectual investment and is deemed to require considerable intellectual investment to use. As well as insights into the methods used to solve jumping problems, this work provides technical contributions of: 1) demonstration of a novel method used to determine the dynamic balancing capability of a jumping system and 2) empirical demonstration of a second order kinematic simulation method with results comparable to those found in the literature.

The wind flow through an urban built environment has a significant impact on the safety and comfort of pedestrians and inhabitants. Simulation of urban wind flow presents a formidable challenge to computational engineering field due to the complex geometry of the built environment and the multiscale nature of the flow. Numerical analysis via computational fluid dynamics (CFD) based on the Navier-Stokes equations is now commonplace in the industry, usually run on high performance computer clusters of CPU nodes. However, resolution of all the turbulent scales of motion in the entire domain is likely to remain beyond the reach of such hardware for the foreseable future; thus, so called Direct Numerical Simulation is not practical for industrial applications. Moreover, the region of interest usually represents a small percentage of the total volume of the domain. The objective of this thesis is to achieve a time-dependent simulation that resolves large to medium turbulence scales in the region of interest at a significantly reduced computational cost compared to turbulent scale resolving Navier-Stokes methods. To do so, we couple a lattice Boltzmann (LB) solver running on graphics processing units (GPUs) with a Navier-Stokes (NS) solver running on CPUs. The LB solver incorporates the mean turbulent flow information from the Navier-Stokes model into its resolved velocity via a synthetic eddy method (SEM) implemented at the inlet of the LB domain. The resulting coupled Navier-Stokes lattice Boltzmann (NSLB) solver combines the accuracy and computing speed of the GPU implementation of the LB model with the stability, low memory consumption and mesh flexibility of the NS solver. Moreover, the NSLB model exploits the widespread availability of CPU and GPU hardware on desktop, and workstation computers. Validation results of the two-way coupled NSLB solver for laminar flow demonstrate that the coupled solver is able to reproduce the results of the full domain single solver (either LB or NS) independently of the position of the interface between the solvers. For the application to urban wind flow, we embedded our lattice Boltzmann large eddy simulation (LB-LES) solver within a pre-calculated Reynolds averaged Navier-Stokes (RANS) simulation of flow around a single building at Re_H = 47893. The LB-LES model accurately predicts the mean and fluctuating velocity around the building, increasing the flow data available and its accuracy with respect to the underlying RANS simulation. The RANS LB-LES coupling allows fully resolved LES results to be achieved in practical time scales with a single desktop based GPU, which shows potential to run industry applications on consumer devices. The original contributions of this work include the development of the coupling framework, adaptation of the SEM to LB and refinement of the embedded boundary conditions. The resulting solver fully realises the objective of a coupled Navier-Stokes method with a GPU accelerated lattice Boltzmann method in order to accurately simulate high Reynolds number flow at affordable computational cost.

This thesis investigates the relationship between form and function of the wing musculature in birds and its implications for locomotion style. Modern birds show an extreme diversity of wing morphologies and flight modes, from ducks that constantly flap their wings, albatrosses and large birds of prey that fly by gliding and soaring to birds such as penguins and auks that use their wings for underwater propulsion. Previous studies in birds have investigated aspects of wing shape and forelimb skeletal morphology in relation to their flight modes and evolutionary history. However, the forelimb musculature has received less attention and our knowledge on its functional anatomy still has many gaps. In this work I aimed to address these knowledge gaps. First, I investigated the use of contrast-enhanced micro-computed tomography to visualise the soft tissue comprising the wing in a model species. I developed a workflow of digital dissection and presented a threedimensional (3D) model of a bird wing that can be used as a basis for future biomechanical analysis using computer modelling. Next I compared the muscle architecture of the wing musculature between closely related species of raptors and aquatic birds because they display a variety of flight styles. I provided quantitative data of muscle architecture for the entire musculature of the avian forelimb for a total of 25 species and found that birds showing similar flight mode showed similar patterns of muscle mass distribution along their wing. I also evaluated the range of movements that birds use during wing-driven locomotion in a kinematics analysis of swimming in the Humboldt using video-based markerless photogrammetry. I described the wing and body kinematics during a variety of diving modes and velocities, and provided values for hydrodynamics forces. This, in combination with numerical data of forelimb muscles, provided some insights into the role that individual muscles play during wing-propelled locomotion and its implication with a variety of flight style.