Refine
Document Type
- Article (17)
- Conference Proceeding (13)
- Doctoral Thesis (1)
Institute
Keywords
- CAD (5)
- Finite-Elemente-Methode (5)
- Bauwerk (3)
- Modellierung (3)
- Tragwerk (3)
- Architektur (2)
- Baustatik (2)
- Beulung (2)
- Elastoplastizität (2)
- Freie Schwingung (2)
Year of publication
- 1997 (31) (remove)
Ideally, multiple computational building evaluation routines (particularly simulation tools) should be coupled in real-time to the representational design model to provide timely performance feed-back to the system user. In this paper we demonstrate how this can be achieved effectively and conveniently via homology-based mapping. We consider two models as homologous if they entail isomorphic topological information. If the general design representation (i.e., a shared object model) is generated in a manner so as to include both the topological building information and pointers to the semantic information base, it can be used to directly derive the domain representations (>enriched< object models with detailed configurational information and filtered semantic data) needed for evaluation purposes. As a proof of concept, we demonstrate a computational design environment that dynamically links an object-oriented space-based design model, with structurally homologous object models of various simulation routines.
The general motivation of this research is to develop software to support the handling of the increased complexity of architectural design. In this paper we describe a system providing general support during the whole process. Instead of only developing design tools we are also addressing the problem of the operating environment of these tools. We conclude that design tools have to be integrated in an open, modular, distributed, user friendly and efficient environment. Two major fields have to be addressed - the development of design tools and the realisation of an integrated system as their operation environment. We will briefly focus on the latter by discussing known technologies in the field of information technology and other design disciplines that can be used to realise such an environment. Regarding the first subject we have to state the need of a detailed tool specification. As a solution we suggest a strategy where the tool functions are specified on the basis of a transformation, where a hierarchical process model is mapped into specifications of different design tools realising appropriate support for all sub-processes of architectural design. Using this strategy the main steps to develop such a support system are: implementation of a framework as basis for the integrated design system decision whether the tool specification are already implemented in available tools in this case these tools can be integrated using known methods for tool coupling otherwise new design tools have to be developed according to the framework
The purpose of this paper is to review model for finite element techniques for non-linear crack analysis of reinforced concrete beams and slabs. The non-linear behaviour of concrete and steel were described. Some calculations of >self-stress< for concrete and reinforced concrete beam was made. Current computational aspects are discussed. Several remarks for future studies are also given. The numerical model of the concrete and reinforced concrete was described. The paper shows the results of calculations on a reinforced concrete plane stress panel with cracks. The non-linear, numerical model of calculations of reinforced concrete was assumed. Using finite elements method some calculations were made. The results of calculations like displacements, stresses and cracking are shown on diagrams. They were compared with experimental results and other finding. Some conclusions about the described model and results of calculation are shown.
This is a paper about knowledge in design and how to elicit knowledge from design processes. The paper is a preparation for an empirical study of interaction in the design process. Reasonings of three authors - Schön, Broadbent and Lundequist - on design processes is presented. They all have a pragmatic perspective in common, and regard the process as an activity without a definite form. Design is seen as an activity of creating models of forms and shapes, by addressing expert knowledge in a dialogic way to problematic situations. Due to the pragmatic approach I find the pragmatist Dewey´s understanding of knowledge and elecitation of knowledge appropiate for studying design processes. According to him it is possible to build up objectified descriptions of experiences, also of such, which are based on experiences of emotional and intuitive nature. There need not be a definite border, which separates tacit knowledge from explicit knowledge - when it comes to the question of the possibility of verbal descriptions. Tacit knowledge is possible to articulate within pragmatic thinking. The conclusion is, that it is possible to study the tacit knowledge of design processes, and get some qualitative insights useful for theory building. A study of design processes can look at three different forms of knowledge. It appears as a precognitive understanding of the design situation, as integrated in the design activity - seeing the situation as something known - and in the process of creating something new.
A Product Model of a Road
(1997)
Many errors and delays frequently appear when data is exchanged between particular tasks in the lifecycle of the road. Inter-task connections are therefore of great importance for the quality of the final product. The article describes a product model of a road wich is the kernel of an integrated information system intended to support all important stages of the road lifecycle: design, evaluation (through different analysis procedures), construction, and maintainance. Since particular tasks are often executed at different places and in different companies, the interconnections are supported by a special metafile which contains all specific data of the product model. The concept of the integrated system is object and component oriented. Additionally, existing conventional program packages are included to support some common tasks (methods). A conventional relational database system as well as an open spatial database system with the relevant GIS functionality are included to support the data structures of the model.
Review of Discrete Optimization Techniques for CAD Discrete optimization in the structure design Morphological method The alternative graph approach Convex discrete optimization without objective function Matroidal Decomposition in design Decomposition of layered matrices Discrete Optimization in Designing Packing problem Optimal arrangement of rectangles and shortest paths in L1-metrics Partition problems Discrete optimization in computational geometry and computer graphics Maxima of a point set on the plane Triangulation One of the main problems in computer graphics is removing hidden lines and surfaces
In the superelliptic shell joined to a circular cylinder bending stresses are absent when it is subjected to uniform pressure.Some geometrical characteristics have been found. Expressions for determining stresses in the shell crest(in the singular point of plane type) are suggested. The problem of a theoretical critical buckling load of an elongated shell supported by frames is studied. A critical buckling load for two shells with different specifications was found experimentally.
In this paper the results of the investigations of the free oscillations of the pre-stressed flexible structure elements are presented . Two cases of the central preliminary stress are investigated : without intermediate fastening of the tie to the flexible element and with the intermediate fastening in the middle of the element length. The given physical model can be applied to the flexible sloping shells and arches, membranes, large space antenna fields (besides flexible elements). The peculiarity of these systems is the possibility of the non-adjacent equilibrium form existence at the definite relations of the physical parameters . The transition from one stable equilibrium form to another, non-adjacent form, may be treated as jump. In this case they are called systems with buckling or the systems with two potential «gaps». These systems commenced the new section of the mathematical physics - the theory of chaos and strange attractors. The analysis of the solutions confirms the received for the first time by the author and given in effect of the oscillation period doubling of the system during the transition from the «small» oscillations relatively center to the >large< relatively all three equilibrium conditions. The character of the frequency (period) dependence on the free oscillation amplitudes of the non-linear system also confirms the received earlier result of the duality of the system behaviour : >small< oscillations possess the qualities of soft system; >large< oscillations possess the qualities of rigid system. The >small< oscillation natural frequency changing, depending on the oscillation amplitudes, is in the internal . Here the frequency takes zero value at the amplitude values Aa and Ad (or Aa and Ae ); the frequency takes maximum value at the amplitude value near point b .The >large< oscillation natural frequency changes in the interval . Here is also observed . The influence of the tie intermediate fastening doesn't introduce qualitative changes in the behaviour of the investigated system. It only increases ( four times ) the critical value of the preliminary tension force
The aim of researches conducted within gamsau about urban simulation, in particular Remus project, is to allow rapid modeling of large and regular urban zones, for purpose of interactive navigation (like VRML) or for realistic rendering (ray-tracing methods). One of problems to be solved in this context is the multiplicity of data formats : inputs come from different sources, and outputs are for heterogeneous systems of visualization. Typically CSG and boundary representation must be generated, treated and converted during building of models. Furthermore, the generated models can be more or less refined, depending on requests and type of use. This paper describes the general context of data models conversion, problems concerning levels of detail and implementation done in Remus, based on object oriented approach.
The frame of this paper is the development of methods and procedures for the description of the motion of an arbitrary shaped foundation. Since the infinite half-space cannot be properly described by a model of finite dimensions without violating the radiation condition, the basic problems are infinite dimensions of the half-space as well as its non-homogeneous nature. Consequently, an approach has been investigated to solve this problem indirectly by developing Green's function in which the non-homogeneity and the infiniteness of the half-space has been included. When the Green's function is known, the next step will be the evaluation of contact stresses acting between the foundation and the surface of the half-space through an integral equation. The equation should be solved in the area of the foundation using Green's function as the kernel. The derivation of three-dimensional Green's function for the homogeneous half-space (Kobayashi and Sasaki 1991) has been made using the potential method. Partial differential equations occurring in the problem have been made ordinary ones through the Hankel integral transform. The general idea for obtaining the three-dimensional Green's function for the layered half-space is similar. But in that case some additional phenomena may occur. One of them is the possibility of the appearance of Stonely surface waves propagating along the contact surfaces of layers. Their contribution to the final result is in most cases important enough that they should not be neglected. The main advantage of results presented in comparing to other obtained with numerical methods is their accuracy especially in the case of thin layers because all essential steps of Green's function evaluation except of the contour integration along the branch cut have been made analytically. On the other hand the disadvantage of this method is that the mathematical effort for obtaining the Green's function is increasing drastically with the increase of the number of layers. Future work will therefore be directed in simplifying of the above described process