Tribodynamic Modeling of Digital Fluid Power Motors

Research output: Book/ReportPh.D. thesis

Abstract

In fluid power engineering, efficiency and reliability optimization have become a major objective. The interest in using fluid power transmission in wind and wave energy applications are producing requirements concerning efficiency and reliability in order to compete with other transmission systems. In fluid power motoring and pumping units, a significant problem is that loss mechanisms do not scale down with diminishing power throughput. Although machines can reach peak efficiencies above 95%, the actual efficiency during operation, which includes part-load situations, is much lower. The invention of digital fluid power displacement units has been able to address this problem. The main idea of the digital fluid power displacement technology is to disable individual chambers, by use of electrical actuated valves. A displacement chamber is disabled by keeping the valve, between the chamber and the low pressure manifold, open throughout the shaft revolution. As no power output is produced, the design of sliding and sealing surfaces becomes very important. The study of such surfaces is the study tribology, the science and technology of friction, lubrication and wear. In consequence, useful tribological design methods and tools are important to the development of digital fluid power machines.

The work presented in this dissertation is part of a research program focusing on the development of digital fluid power MW-motors for use in hydraulic drive train in wind turbines. As part of this development, the design, analysis and optimization of efficiency and reliability of tribological interfaces in the motor are essential. Consequently, this dissertation concerns development of a suitable tool and methodology in order to reach optimality in the design. The main part is on development of tribodynamic modeling.

A fundamental issue with tribodynamic systems modeling is the computational effort. The ambition of solving the partial differential equations, which governs the physics in tribodynamics, is closely followed by careful use of efficient computational code. The overall aim of tribodynamic modeling is to overcome the difficult task to seek the optimum design. An approach to optimum design of fluid power displacement machines is suggested in this thesis. In the first step of the design process a topological selection is performed, based on application framework, existing and new ideas, experience, engineering judgement and prospective methods. The basic idea is thereafter to utilize a model, which is practical for wide scale design analysis, and subsequently refine the design space, and then more accurate models can be utilized prior to prototype preparation. This approach is an acknowledgement of the advanced state of the art models being impractical in wide scale design analysis, due to high simulation durations.

In this dissertation a topological selection and a preliminary design of a digital fluid power displacement unit is presented. The main focus is the tribodynamic modeling of this design, with the aim to perform simulations, which are practical for wide scale design analysis. However, until now only very limited optimization have been performed with existing tribodynamic models, which is attributed to impractical simulation durations. The state of the art reveals that for very advanced models the use becomes limited to numerical experiments, rather than calculation of object functions in optimization algorithms. Consequently, the ambition within development of a tribodynamic model is the ability to perform simulations with sufficient accuracy in a sufficiently low amount of time. Three main areas of focus for such development is efficient coding, high processing capacity and analytical approximations. The main area of focus in this dissertation is the development of analytical approximations in order to achieve faster modeling tools. Three main contributions are made in this context. A multibody tribodynamic modeling framework, where the lubricant wall stress vector is analytically coupled with a Newton-Euler multibody dynamics approach, a piezoviscous short bearing solution applicable to oil hydraulic lubrication films, where a series of necessary conditions, in order to justify the use of this solution, are elaborated and an asymptotic approximation of laminar lubrication thermal fields at low reduced Peclet and Brinkman number. In addition, a dynamical analysis of the tribodynamics in the preliminary digital fluid power displacement motor design is performed, which provide insight to the challenge of performing more advanced tribodynamic simulation, while retaining practicality of the simulation model with regard to wide-scale design analysis and optimization purposes.
Close

Details

In fluid power engineering, efficiency and reliability optimization have become a major objective. The interest in using fluid power transmission in wind and wave energy applications are producing requirements concerning efficiency and reliability in order to compete with other transmission systems. In fluid power motoring and pumping units, a significant problem is that loss mechanisms do not scale down with diminishing power throughput. Although machines can reach peak efficiencies above 95%, the actual efficiency during operation, which includes part-load situations, is much lower. The invention of digital fluid power displacement units has been able to address this problem. The main idea of the digital fluid power displacement technology is to disable individual chambers, by use of electrical actuated valves. A displacement chamber is disabled by keeping the valve, between the chamber and the low pressure manifold, open throughout the shaft revolution. As no power output is produced, the design of sliding and sealing surfaces becomes very important. The study of such surfaces is the study tribology, the science and technology of friction, lubrication and wear. In consequence, useful tribological design methods and tools are important to the development of digital fluid power machines.

The work presented in this dissertation is part of a research program focusing on the development of digital fluid power MW-motors for use in hydraulic drive train in wind turbines. As part of this development, the design, analysis and optimization of efficiency and reliability of tribological interfaces in the motor are essential. Consequently, this dissertation concerns development of a suitable tool and methodology in order to reach optimality in the design. The main part is on development of tribodynamic modeling.

A fundamental issue with tribodynamic systems modeling is the computational effort. The ambition of solving the partial differential equations, which governs the physics in tribodynamics, is closely followed by careful use of efficient computational code. The overall aim of tribodynamic modeling is to overcome the difficult task to seek the optimum design. An approach to optimum design of fluid power displacement machines is suggested in this thesis. In the first step of the design process a topological selection is performed, based on application framework, existing and new ideas, experience, engineering judgement and prospective methods. The basic idea is thereafter to utilize a model, which is practical for wide scale design analysis, and subsequently refine the design space, and then more accurate models can be utilized prior to prototype preparation. This approach is an acknowledgement of the advanced state of the art models being impractical in wide scale design analysis, due to high simulation durations.

In this dissertation a topological selection and a preliminary design of a digital fluid power displacement unit is presented. The main focus is the tribodynamic modeling of this design, with the aim to perform simulations, which are practical for wide scale design analysis. However, until now only very limited optimization have been performed with existing tribodynamic models, which is attributed to impractical simulation durations. The state of the art reveals that for very advanced models the use becomes limited to numerical experiments, rather than calculation of object functions in optimization algorithms. Consequently, the ambition within development of a tribodynamic model is the ability to perform simulations with sufficient accuracy in a sufficiently low amount of time. Three main areas of focus for such development is efficient coding, high processing capacity and analytical approximations. The main area of focus in this dissertation is the development of analytical approximations in order to achieve faster modeling tools. Three main contributions are made in this context. A multibody tribodynamic modeling framework, where the lubricant wall stress vector is analytically coupled with a Newton-Euler multibody dynamics approach, a piezoviscous short bearing solution applicable to oil hydraulic lubrication films, where a series of necessary conditions, in order to justify the use of this solution, are elaborated and an asymptotic approximation of laminar lubrication thermal fields at low reduced Peclet and Brinkman number. In addition, a dynamical analysis of the tribodynamics in the preliminary digital fluid power displacement motor design is performed, which provide insight to the challenge of performing more advanced tribodynamic simulation, while retaining practicality of the simulation model with regard to wide-scale design analysis and optimization purposes.
Original languageEnglish
PublisherDepartment of Energy Technology, Aalborg University
Number of pages243
ISBN (Print)978-87-92846-49-5
StatePublished - Dec 2014
Publication categoryResearch

Download statistics

No data available
ID: 207212154