PhD defence by Remzija Cerimagic
12.09.2022 kl. 13.00 - 16.00
Remzija Cerimagic, AAU Energy, will defend the thesis "Isogeometric Tribodynamics: A Generic Framework for Modeling and Simulation in Tribology"
Isogeometric Tribodynamics: A Generic Framework for Modeling and Simulation in Tribology
Professor Torben Ole Andersen
Associate Professor Per Johansen
Professor Rudolf Scheidl, Johannes Kepler Universität Linz
Associate Professor Kim Sørensen
Associate Professor Kim Sørensen, Aalborg University, Denmark (Chairman)
Professor Andrea Vacca, Purdue University, Indiana, USA
Professor Huayong Yang, Zhejiang University, China
This thesis is part of a research project, focusing on the development of digital displacement fluid power machines for wind turbine transmission systems. The use of hydraulic transmissions within wind energy increases requirements for efficiency and reliability of the digital displacement motors and pumps in order to compete with the existing solutions. The power losses in hydraulic motors and pumps depend on the operating conditions and the design of sliding and sealing surfaces, which are considered critical design elements in hydraulic machinery. The study of friction, lubrication and wear inherent in the sliding and sealing interfaces is called tribology. Therefore, suitable tribological tools and design methods is essential in the development of digital displacement fluid power machines for wind turbine transmission systems. As part of this design and development process, analysis and optimization of admissible design topologies is absolutely key. The main focus of the thesis is on the development of a suitable and generic tribological tool for optimal design of digital motors and pumps for hydrostatic transmissions.
Since the applicability and efficiency of displacement units depend on the selected design topology, the need for a generic tool and methodology with more generalized parameter assumptions suitable for design and optimization of various topologies, is therefore justified. The ambition of applying tribodynamic models for design optimization purposes is to a great extent restricted by joint specific analytical geometric and kinematic models, and unacceptable simulation duration. This is certainly challenging the practical application of the advanced state of the art models for design optimization purposes.
A method based on the use of Computer Aided Design (CAD) parameterizations to generalize tribodynamic modeling is presented in this thesis. The basic idea is to use isogeometric analysis to combine the tribological models at the lubrication interface with the rigid body motion analysis. Isogeometric analysis combines the ability to analyze complicated engineering problems from the finite element method (FEM) with the ability to effectively represent complex geometries from CAD. The method seems particularly attractive for shape optimization purposes due to the tight connection between the geometric model and the analysis model; and has also been found to be effective in fluid flow modeling due to the high built-in regularity of the pressure and velocity approximations. Thus, it seems very natural to pursue isogeometric analysis in tribodynamics modeling.
A main challenge to the problem of generalizing tribodynamic modeling is the seamless integration of the physical model from CAD with tribological models. Two important contributions have been developed in this context. As the first contribution, a new kinematic formulation based on an implicit matrix method to transform CAD parameterizations into implicit functions is presented. A detailed description of the construction of the implicit matrices is given and explicit expressions of the thickness of the lubricating film, its velocity and the relative velocities of the sliding surfaces are presented. The second contribution shows how the new kinematic formulation, together with isogeometric analysis, can be used in a generic tribodynamic computational framework. In addition, the thesis touches upon some issues involved in tribodynamic virtual prototyping. In this context, it is shown how mechanical contact with endogenous air bubbles in the oil can give high parameter sensitivity. Why optimization with tribodynamic models require special consideration.
THE DEFENCE will be IN ENGLISH - all are welcome.