This dissertation presents several dynamic models in order to evaluate the quality of vibration isolation in a motorcycle. Alternate meta-models are also developed from theoretical models for model simplification and reduction of computational effort. All models presented herein are used to solve the engine mount optimization problem so as to minimize the forces transmitted to the frame while meeting packaging constraints, and using the mount system parameters as design variables.; There is a significant amount of published literature in the area of motorcycle dynamics and motorcycle modeling. However, there is a lack of published literature in the area of vibration isolation modeling and optimization of a motorcycle engine mount system. This dissertation tries to fill in the gap by developing several generic models, consisting of various structural sub-systems, to encompass different motorcycle layouts used by current manufacturers.; This dissertation addresses three main problems. The first problem addresses a step-by-step development of mathematical models in order to capture the quality of vibration isolation in a motorcycle system. This involves formulating several models with different underlying assumptions. The influence of the structural stiffness of the frame and swing-arm assemblies on the vibration isolation problem is incorporated into the model.; The second problem this dissertation focuses on is optimization of a mount system, using the above-mentioned theoretical models, in order to minimize the forces transmitted to the frame through the mount system as a result of engine excitation loads and road load inputs. The engine mount parameters are used as the governing design variables and displacement constraints, static and dynamic, are placed on the power-train assembly to account for packaging constraints and to prevent premature snubbing of the engine mounts. Comprehensive examples are provided to demonstrate the validity of all the models for solving the mount optimization problem.; The third problem this dissertation addresses deals with building meta-models from the theoretical models in order to simplify the governing model and also reduce computational effort required for solving the mount optimization problem. These alternate models are then used for optimization as well. The two meta-modeling techniques that have been used in this dissertation are Response Surface Methodology and Kriging. Design of Experiments techniques have been used to formulate the experiments for developing the meta-models. (Abstract shortened by UMI.)
展开▼
