Khan, M Tanveer Rahman (2009) Modeling and simulation of flexible frames for proper modeling of heavy truck dynamics. Masters thesis, Memorial University of Newfoundland.
- Accepted Version
Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.
The goal of the proposed research is to determine the range of validity of rigid body models of trucks depending on the severity of road inputs, and the mass and location of the payload. A "half car" model of a Class VI truck was constructed using the bond graph modeling representation, for which power flow among elements has been employed to determine whether or not frame flexibility affects vehicle motion, or can be neglected. Bond graph based proper modeling and partitioning methods were used to systematically and quantitatively determine whether or not frame flexibility effects were negligible. A proposed algorithm based on Design of Experiments and response surface analysis showed promise in more efficiently searching the design space to generate the range of parameters with fewer simulation runs than were required with a "brute force" method. The feasibility of the power-based partitioning method and response surface algorithm were demonstrated with a simple free-free beam case study. -- Flexible frame modal parameters were calculated from theory, and with a finite element model, for inclusion in half-car (pitch plane) bond graph models. Application of the partitioning algorithm to a nonlinear half-car model with vertical and longitudinal dynamics resulted in a range of payloads and road roughnesses for which a rigid model may be assumed valid. The metric for assessing suitability of a rigid model, which was called "relative activity", was correlated with accuracy of a rigid model. Rigid models within the range of validity performed well. The results of this thesis support the use of energy-based partitioning to automate the reduction of truck models. Unnecessary complexity can be avoided while predictive ability is maintained, with less reliance on intuition and assumption.
|Item Type:||Thesis (Masters)|
|Additional Information:||Includes bibliographical references (leaves 121-125).|
|Department(s):||Engineering and Applied Science, Faculty of|
|Library of Congress Subject Heading:||Bond graphs; Trucks--Dynamics--Mathematical models; Trucks--Vibration--Mathematical models|
Actions (login required)