Objects interacting with fluid are of high interest to visually present in three-dimensional applications, such as computer games and virtual environments. For presenting the interactions with high correctness, dynamic rigid body simulation may be used. This paper presents methods for efficient, physics-based real-time simulation of fluid-immersed rigid bodies, where the correctness of the simulation is maintained. Simulated forces include gravity, buoyancy, thrust, drag, and lift. To have the simulation run efficiently in real-time, discretization of the simulated rigid body is made by applying mentioned forces to a user-defined number of particles, sampled pseudo-randomly within the rigid body.

No description available.

This thesis adds to the knowledge base of constant-force mechanisms (CFMs). It begins by reviewing past work done in the area of CFMs and then develops new nondimensionalized parameters that are used to simplify the calculations required to design a CFM. Comparison techniques are then developed that utilize these non-dimensionalized parameters to compare mechanisms based on stiffnesses, percent constant-force, actual lengths, normal displacements, and feasible design orientations. These comparison techniques are then combined with optimization to define new mechanisms with improved performance and range of capabilities. This thesis also outlines a design process, methods to identify mechanisms that are suitable for a given design problem, and relationships and trends between variables. The thesis concludes by discussing the adaptation of CFMs for use in electrical contacts and presenting the results of a design case study which successfully developed a constant-force electrical contact (CFEC).

A mathematical dynamic model is derived for the compliant constant-force mechanism, based on the pseudo-rigid-body model simplification of the device. The compliant constant-force mechanism is a slider mechanism incorporating large-deflection beams, which outputs near-constant-force across the range of its designed deflection. The equation of motion is successfully validated with empirical data from five separate mechanisms, comprising two configurations of compliant constant-force mechanism. The dynamic model is cast in generalized form to represent all possible configurations of compliant constant-force mechanism. Deriving the dynamic equation from the pseudo-rigid-body model is useful because every configuration is represented by the same model, so a separate treatment is not required for each configuration. An unexpected dynamic trait of the constant-force mechanism is discovered: there exists a range of frequencies for which the output force of the mechanism accords nearer to constant-force than does the output force at static levels.

This thesis presents research on the force and deflection behavior of beams with rectangular cross-sections undergoing lateral torsional buckling. The large 3-D deflection path of buckling beam tips was closely approximated by circular arcs in two planes. A new chain algorithm element was created from pseudo-rigid-body segments and used in a chain calculation that accurately predicted the force deflection relationship of beams with large 3-D deflections.

Visual effects for film and animation often require simulated granular materials, such as sand, wheat, or dirt, to meet a director's needs. Simulating granular materials can be time consuming, in both computation and labor, as these particulate materials have complex behavior and an enormous amount of small-scale detail. Furthermore, a single cubic meter of granular material, where each grain is a cubic millimeter, would contain a billion granules, and simulating all such interacting granules would take an impractical amount of time for productions. This calls for a simplified model for granular materials that retains high surface detail and granular behavior yet requires significantly less computational time. Our proposed method simulates a minimal number of individual granules while retaining particulate detail on the surface by supporting surface particles with simplified interior granular models. We introduce a multi-state model where, depending on the material state of the interior granules, we replace interior granules with a simplified simulation model for the state they are in and automate the transitions between those states. The majority of simulation time can thus be focused on visible portions of the material, reducing the time spent on non-visible portions, while maintaining the appearance and behavior of the mass as a whole.

No description available.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第25279号 / 工博第5238号 / 京都大学大学院工学研究科機械理工学専攻 / (主査)教授 黒瀬 良一, 教授 長田 孝二, 教授 平山 朋子 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

Advancements in computational capabilities have allowed researchers to turn towards modeling as an efficient tool to replicate and predict outcomes of complex systems. Computational models of the musculoskeletal system have gone through various iterations with early versions employing dramatic simplifications. In this work, a three-dimensional computational model of the wrist joint was developed. It accurately recreated the skeletal structures of the hand and wrist and represented the constraints imposed by soft tissue structures like ligaments, tendons, and other surrounding tissues. It was developed to function as a tool to investigate the biomechanical contributions of structures and the kinematic response of the wrist joint. The model was created with the use of a commercially available computer-aided design software employing the rigid body modeling methodology. It was validated against three different cadaveric experimental studies which investigated changes in biomechanical response following radioscapholunate fusion and proximal row carpectomy procedures. The kinematic simulations performed by the model demonstrated quantitatively accurate responses for the range of motions for both surgical procedures. It also provided some understanding to the trends in carpal bone contact force changes observed in surgically altered specimens. The model provided additional insight into the importance of structures like the triangular fibrocartilage and the capsular retinacular structures, both of which are currently not very well understood. As better understanding of components of the wrist joint is achieved, this model could function as an important tool in preoperative planning and generating individualized treatment regiments.

Musculoskeletal computational modeling is a versatile and effective tool which may be used to study joint mechanics, examine muscle and ligament function, and simulate surgical reconstructive procedures. While injury to the elbow joint can be significantly debilitating, questions still remain regarding its normal, pathologic, and repaired behavior. Biomechanical models of the elbow have been developed, but all have assumed fixed joint axes of rotation and ignored the effects of ligaments. Therefore, the objective of this thesis was to develop and validate a computational model of the elbow joint whereby joint kinematics are dictated by three-dimensional bony geometry contact, ligamentous constraints, and muscle loading.Accurate three-dimensional bone geometry was generated by acquiring CT scans, segmenting the images to isolate skeletal features, and fitting surfaces to the segmented data. Ligaments were modeled as tension-only linear springs, and muscle were represented as force vectors with discrete attachment points. Bone contact was modeled by a routine which applied a normal force at points of penetration, with a force magnitude being a function of penetration depth. A rigid body dynamics simulator was used to predict the model's behavior under particular external loading conditions.The computational model was validated by simulating past experimental investigations and comparing results. Passive flexion-extension range of motion predicted by the model correlated exceptionally well with reported values. Bony and ligamentous structures responsible for enforcing motion limits also agreed with past observations. The model's varus stability as a function of elbow flexion and coronoid process resection was also investigated. The trends predicted by the model matched those of the associated cadaver study.This thesis successfully developed an accurate musculoskeletal computational model of the elbow joint complex. While the model may now be used in a predictive manner, further refinements may expand its applicability. These include accounting for the interference between soft tissue and bone, and representing the dynamic behavior of muscles.