A Finite Element Framework for Geotechnical Applications based on Object-Oriented Programming
Published on 21. January 2002
Book
Paperback/Softback
112 pages
978-3-7281-2828-7 (ISBN)
Description
This publication describes an alternative methodology for finite element programming based on object-oriented techniques. The concepts of object-oriented programming are outlined and it is explained, how this new paradigm facilitates design, implementation and support of large programming systems. Because the knowledge of object-oriented programs is not dispersed in the actual code, but rather localized in structures, the causal knowledge, meta knowledge and constraints may be integrated in a uniform manner. Control structures separate the external level ('what') from the internal ('how') by encapsulating the actual implementation. As programming language C++ is used.
Up till now several object-oriented finite element frameworks have been presented which are partially extendable. However, the extendibility is limited to a few specific directions, e.g. the introduction of new element types or solving strategies. Much less support is available for task control, creation of new material models, configurable field variable types or extensions of the analysis model. No framework is available which is especially designed to cover the problems encountered when dealing with geotechnical engineering. IMAGINE tries to close these gaps.
The nucleus of the Finite Element framework presented relies on abstractions, which include common concepts accepted in mathematics, mechanics, engineering and interactive visualization techniques, and serve as the fundamental object-oriented framework of classes for finite element applications and task management. The aim of the framework was not to include as many features as possible (e.g. a variety of types of finite elements or material laws), but to provide a sophisticated and robust foundation which may be used in the future due to its inherent capabilities of simple maintenance, adaptability to new resources and extendibility, especially directed towards applications in geotechnical engineering.
Based on this framework an experimental finite element application for geomechanics is presented. Here everything is regarded as an object: loads, load groups, computational tasks etc. Thanks to the graphical user interface under MS-Windows and the rule of non-anticipation on which the objects rely, the handling of the program is straightforward.
Some test examples illustrate its usage.
Content:
Abstract
1 Introduction
2 Motivation
2.1 Introduction
2.2 Procedural Programming Paradigm
2.3 Other Programming Paradigms
2.4 Object-Oriented Programming Paradigm
2.5 Requirements for Large Programming Systems, or how Object-Oriented Programming Meets the Goals
2.6 Summary
3 Object-Oriented Analysis: Aim and Funtion of the FE Framework IMAGINE
3.1 Introduction
3.2 Aim of the Framework
3.3 Documentation
3.4 Object Database
3.5 Development Environment
3.6 FE Kernel
3.7 Geometric Modeling
3.8 Postprocessing
3.9 Project Management
3.10 User Interface
3.11 State of the Art
4 Object-Oriented Design and Implementation of the FE Framework Imagine
4.1 Guidelines
4.2 Architecture of IMAGINE
4.3 Project and Ressource Management Subsystem
4.4 FE Modeling Subsystem
5 Design and Implementation of the User Interface
5.1 Introduction
5.2 Workbench Approach
5.3 FEM Modeling Workbench
5.4 FEM Task Workbench
6 Example
6.1 Introduction
6.2 Sector Without Lining
6.3 Solid Model versus Superelements
7 Conclusions
Appendix 1: From Procedural to Object-Oriented Programming
Appendix 2: Software Life Cycle
Appendix 3: The Document-View Concept of the MFC
Appendix 4: Style Guide
References
Up till now several object-oriented finite element frameworks have been presented which are partially extendable. However, the extendibility is limited to a few specific directions, e.g. the introduction of new element types or solving strategies. Much less support is available for task control, creation of new material models, configurable field variable types or extensions of the analysis model. No framework is available which is especially designed to cover the problems encountered when dealing with geotechnical engineering. IMAGINE tries to close these gaps.
The nucleus of the Finite Element framework presented relies on abstractions, which include common concepts accepted in mathematics, mechanics, engineering and interactive visualization techniques, and serve as the fundamental object-oriented framework of classes for finite element applications and task management. The aim of the framework was not to include as many features as possible (e.g. a variety of types of finite elements or material laws), but to provide a sophisticated and robust foundation which may be used in the future due to its inherent capabilities of simple maintenance, adaptability to new resources and extendibility, especially directed towards applications in geotechnical engineering.
Based on this framework an experimental finite element application for geomechanics is presented. Here everything is regarded as an object: loads, load groups, computational tasks etc. Thanks to the graphical user interface under MS-Windows and the rule of non-anticipation on which the objects rely, the handling of the program is straightforward.
Some test examples illustrate its usage.
Content:
Abstract
1 Introduction
2 Motivation
2.1 Introduction
2.2 Procedural Programming Paradigm
2.3 Other Programming Paradigms
2.4 Object-Oriented Programming Paradigm
2.5 Requirements for Large Programming Systems, or how Object-Oriented Programming Meets the Goals
2.6 Summary
3 Object-Oriented Analysis: Aim and Funtion of the FE Framework IMAGINE
3.1 Introduction
3.2 Aim of the Framework
3.3 Documentation
3.4 Object Database
3.5 Development Environment
3.6 FE Kernel
3.7 Geometric Modeling
3.8 Postprocessing
3.9 Project Management
3.10 User Interface
3.11 State of the Art
4 Object-Oriented Design and Implementation of the FE Framework Imagine
4.1 Guidelines
4.2 Architecture of IMAGINE
4.3 Project and Ressource Management Subsystem
4.4 FE Modeling Subsystem
5 Design and Implementation of the User Interface
5.1 Introduction
5.2 Workbench Approach
5.3 FEM Modeling Workbench
5.4 FEM Task Workbench
6 Example
6.1 Introduction
6.2 Sector Without Lining
6.3 Solid Model versus Superelements
7 Conclusions
Appendix 1: From Procedural to Object-Oriented Programming
Appendix 2: Software Life Cycle
Appendix 3: The Document-View Concept of the MFC
Appendix 4: Style Guide
References