000 Informatik, Wissen, Systeme
Refine
Document Type
- Conference Proceeding (147) (remove)
Institute
- In Zusammenarbeit mit der Bauhaus-Universität Weimar (82)
- Graduiertenkolleg 1462 (31)
- Institut für Strukturmechanik (ISM) (12)
- Professur Angewandte Mathematik (12)
- Professur Informatik im Bauwesen (5)
- Institut für Konstruktiven Ingenieurbau (IKI) (4)
- Professur Stochastik und Optimierung (3)
- Professur Computer Vision in Engineering (2)
- Professur Stahlbau (2)
- Institut für Bauinformatik, Mathematik und Bauphysik (IBMB) (1)
Keywords
- Angewandte Informatik (145)
- Angewandte Mathematik (145)
- Computerunterstütztes Verfahren (145)
- Architektur <Informatik> (74)
- Computer Science Models in Engineering; Multiscale and Multiphysical Models; Scientific Computing (74)
- Building Information Modeling (35)
- Data, information and knowledge modeling in civil engineering; Function theoretic methods and PDE in engineering sciences; Mathematical methods for (robotics and) computer vision; Numerical modeling in engineering; Optimization in engineering applications (34)
- BIM (1)
- Bauinformatik (1)
- Data, information and knowledge modeling in civil engineering (1)
Hochschulwege 2015
(2017)
Die in diesem Tagungsband zusammengeführten Beiträge beschäftigen sich mit dem Spannungsfeld, das sich zwischen externen Förderprogrammen, Veränderungsprojekten und den Zielen, Strukturen und Bedingungen der jeweiligen Hochschule ergibt. In diesem Spannungsfeld kommt es unweigerlich zu Reibungen, da vorhandene Strukturen und Ziele in Konflikt mit neuen Vorhaben und Ideen geraten. Ein Teil der Projekte stellt allein durch ihr finanzielles Volumen und die daraus resultierende Wirkkraft die tradierten Verhältnisse zwischen Lehre, Forschung und den wissenschaftsstützenden Bereichen in Frage und teils auf den Kopf. Die leitenden Fragen der Tagung und der hier versammelten Beiträge waren daher: Wie bringen Hochschulen ihre individuellen Ziele mit denen der bundesweiten Programme oder länderspezfifischer Fördermaßnahmen überein? Wie gehen Hochschulen mit ihren Projekten um? Wie vollzieht sich Veränderung an den Hochschulen? Und schließlich: Was bleibt von den Impulsen, die Projekte setzen? Die in diesem Tagungsband versammelten Beiträge geben darauf erste, auf dem bisherigen Erfahrungswissen basierende Antworten. Sie setzen sich intensiv mit den Faktoren auseinander, die den Erfolg von Veränderungsprozessen und Projekten befördern oder behindern können und leiten daraus Empfehlungen für Gestaltungsprozesse an Hochschulen ab.
30. Forum Bauinformatik
(2018)
Die Bauhaus-Universität Weimar ist seit langer Zeit mit dem Forum Bauinformatik eng verbunden. So wurde die Veranstaltung 1989 hier durch den Arbeitskreis Bauinformatik ins Leben gerufen und auch das 10. und 18. Forum Bauinformatik (1998 bzw. 2006) fand in Weimar statt. In diesem Jahr freuen wir uns daher besonders, das 30. Jubiläum an der Bauhaus-Universität Weimar ausrichten zu dürfen und viele interessierte Wissenschaftler und Wissenschaftlerinnen aus dem Bereich der Bauinformatik in Weimar willkommen zu heißen.
Das Forum Bauinformatik hat sich längst zu einem festen Bestandteil der Bauinformatik im deutschsprachigen Raum entwickelt. Dabei steht es traditionsgemäß unter dem Motto „von jungen Forschenden für junge Forschende“, wodurch insbesondere Nachwuchswissenschaftlerinnen und ‑wissenschaftlern die Möglichkeit geboten wird, ihre Forschungsarbeiten zu präsentieren, Problemstellungen fachspezifisch zu diskutieren und sich über den neuesten Stand der Forschung zu informieren. Zudem wird eine ausgezeichnete Gelegenheit geboten, in die wissenschaftliche Gemeinschaft im Bereich der Bauinformatik einzusteigen und Kontakte mit anderen Forschenden zu knüpfen.
In diesem Jahr erhielten wir 49 interessante und qualitativ hochwertige Beiträge vor allem in den Themenbereichen Simulation, Modellierung, Informationsverwaltung, Geoinformatik, Structural Health Monitoring, Visualisierung, Verkehrssimulation und Optimierung. Dafür möchten wir uns ganz besonders bei allen Autoren, Co-Autoren und Reviewern bedanken, die durch ihr Engagement das diesjährige Forum Bauinformatik erst möglich gemacht haben. Wir danken zudem Professor Große und Professor Díaz für die Unterstützung bei der Auswahl der Beiträge für die Best Paper Awards.
Ein herzliches Dankeschön geht an die Kollegen an der Professur Informatik im Bauwesen der Bauhaus-Universität Weimar für die organisatorische, technische und beratende Unterstützung während der Planung der Veranstaltung.
We present recent developments of adaptive wavelet solvers for elliptic eigenvalue problems. We describe the underlying abstract iteration scheme of the preconditioned perturbed iteration. We apply the iteration to a simple model problem in order to identify the main ideas which a numerical realization of the abstract scheme is based upon. This indicates how these concepts carry over to wavelet discretizations. Finally we present numerical results for the Poisson eigenvalue problem on an L-shaped domain.
The application of partly decoupled approach by means of continuum mechanics facilitates the calculation of structural responses due to welding. The numerical results demonstrate the ability of a qualitative prediction of welded connections. As it is intended to integrate the local effects of a joint in structural analysis of steel constructions, it is necessary to meet higher approaches towards quality. The wide array of material parameters are presented, which are affecting the thermal, metallurgical and mechanical behavior, and which have to be identified. For that purpose further investigations are necessary to analyze the sensitivity of the models towards different material properties. The experimental determination of every material parameter is not possible due to the extraordinary laborious efforts needed. Besides that, experimentally identified parameters can be applied only for the tested steel quality for measured temperature-time regimes. For that reason alternative approaches for identification of material parameters, such as optimization strategies, have to be applied. After a definition of material parameters a quantitative prediction of welded connections will also be possible. Numerical results show the effect of phase transformation, activated by welding process, on residual stress state. As these phenomena occur in local areas in the range of crystal and grain sizes, the description of microscopic phenomena and their propagation on a macroscopic level due to approaches of homogenization might be expedient. Nevertheless, one should bear in mind, the increasing number of material parameters as well as the complexity of their experimental determination. Thus the microscopic approach should always be investigated under the scope of ability and efficiency of a required prediction. Under certain circumstances a step backwards, adopting a phenomenological approach, also can be beneficial.
Nodal integration of finite elements has been investigated recently. Compared with full integration it shows better convergence when applied to incompressible media, allows easier remeshing and highly reduces the number of material evaluation points thus improving efficiency. Furthermore, understanding it may help to create new integration schemes in meshless methods as well. The new integration technique requires a nodally averaged deformation gradient. For the tetrahedral element it is possible to formulate a nodal strain which passes the patch test. On the downside, it introduces non-physical low energy modes. Most of these "spurious modes" are local deformation maps of neighbouring elements. Present stabilization schemes rely on adding a stabilizing potential to the strain energy. The stabilization is discussed within this article. Its drawbacks are easily identified within numerical experiments: Nonlinear material laws are not well represented. Plastic strains may often be underestimated. Geometrically nonlinear stabilization greatly reduces computational efficiency. The article reinterpretes nodal integration in terms of imposing a nonconforming C0-continuous strain field on the structure. By doing so, the origins of the spurious modes are discussed and two methods are presented that solve this problem. First, a geometric constraint is formulated and solved using a mixed formulation of Hu-Washizu type. This assumption leads to a consistent representation of the strain energy while eliminating spurious modes. The solution is exact, but only of theoretical interest since it produces global support. Second, an integration scheme is presented that approximates the stabilization criterion. The latter leads to a highly efficient scheme. It can even be extended to other finite element types such as hexahedrals. Numerical efficiency, convergence behaviour and stability of the new method is validated using linear tetrahedral and hexahedral elements.
We consider a structural truss problem where all of the physical model parameters are uncertain: not just the material values and applied loads, but also the positions of the nodes are assumed to be inexact but bounded and are represented by intervals. Such uncertainty may typically arise from imprecision during the process of manufacturing or construction, or round-off errors. In this case the application of the finite element method results in a system of linear equations with numerous interval parameters which cannot be solved conventionally. Applying a suitable variable substitution, an iteration method for the solution of a parametric system of linear equations is firstly employed to obtain initial bounds on the node displacements. Thereafter, an interval tightening (pruning) technique is applied, firstly on the element forces and secondly on the node displacements, in order to obtain tight guaranteed enclosures for the interval solutions for the forces and displacements.
Steel structural design is an integral part of the building construction process. So far, various methods of design have been applied in practice to satisfy the design requirements. This paper attempts to acquire the Differential Evolution Algorithms in automatization of specific synthesis and rationalization of design process. The capacity of the Differential Evolution Algorithms to deal with continuous and/or discrete optimization of steel structures is also demonstrated. The goal of this study is to propose an optimal design of steel frame structures using built-up I-sections and/or a combination of standard hot-rolled profiles. All optimized steel frame structures in this paper generated optimization solutions better than the original solution designed by the manufacturer. Taking the criteria regarding the quality and efficiency of the practical design into consideration, the produced optimal design with the Differential Evolution Algorithms can completely replace conventional design because of its excellent performance.
A practical framework for generating cross correlated fields with a specified marginal distribution function, an autocorrelation function and cross correlation coefficients is presented in the paper. The contribution promotes a recent journal paper [1]. The approach relies on well known series expansion methods for simulation of a Gaussian random field. The proposed method requires all cross correlated fields over the domain to share an identical autocorrelation function and the cross correlation structure between each pair of simulated fields to be simply defined by a cross correlation coefficient. Such relations result in specific properties of eigenvectors of covariance matrices of discretized field over the domain. These properties are used to decompose the eigenproblem which must normally be solved in computing the series expansion into two smaller eigenproblems. Such decomposition represents a significant reduction of computational effort. Non-Gaussian components of a multivariate random field are proposed to be simulated via memoryless transformation of underlying Gaussian random fields for which the Nataf model is employed to modify the correlation structure. In this method, the autocorrelation structure of each field is fulfilled exactly while the cross correlation is only approximated. The associated errors can be computed before performing simulations and it is shown that the errors happen especially in the cross correlation between distant points and that they are negligibly small in practical situations.
PARAMETER IDENTIFICATION OF MESOSCALE MODELS FROM MACROSCOPIC TESTS USING BAYESIAN NEURAL NETWORKS
(2010)
In this paper, a parameter identification procedure using Bayesian neural networks is proposed. Based on a training set of numerical simulations, where the material parameters are simulated in a predefined range using Latin Hypercube sampling, a Bayesian neural network, which has been extended to describe the noise of multiple outputs using a full covariance matrix, is trained to approximate the inverse relation from the experiment (displacements, forces etc.) to the material parameters. The method offers not only the possibility to determine the parameters itself, but also the accuracy of the estimate and the correlation between these parameters. As a result, a set of experiments can be designed to calibrate a numerical model.
Since the 90-ties the Pascal matrix, its generalizations and applications have been in the focus of a great amount of publications. As it is well known, the Pascal matrix, the symmetric Pascal matrix and other special matrices of Pascal type play an important role in many scientific areas, among them Numerical Analysis, Combinatorics, Number Theory, Probability, Image processing, Sinal processing, Electrical engineering, etc. We present a unified approach to matrix representations of special polynomials in several hypercomplex variables (new Bernoulli, Euler etc. polynomials), extending results of H. Malonek, G.Tomaz: Bernoulli polynomials and Pascal matrices in the context of Clifford Analysis, Discrete Appl. Math. 157(4)(2009) 838-847. The hypercomplex version of a new Pascal matrix with block structure, which resembles the ordinary one for polynomials of one variable will be discussed in detail.