A Comparative Study on Two Different Direct Parallel Solution Strategies for Large-Scale Problems

2009-04-08
Bahcecioglu, T.
Ozmen, S.
Kurç, Özgür
This paper presents a comparative study on two different direct parallel solution strategies for the linear solution of large scale actual finite element models: global and domain-by-domain. The global solution strategy was examined by utilizing the parallel multi-frontal equation solver, MUMPS [1], together with a finite element program. In a similar manner a substructure based parallel solution framework [2] was utilized for investigating the domain-by-domain strategy. Various large-scale structural models were solved with both solution strategies in order to illustrate the efficiencies and weaknesses of each solution strategy. The test runs were performed on a homogeneous PC cluster composed of eight computers connected with an ordinary 1 GBit network switch.

Suggestions

A generative model for multi class object recognition and detection
Ulusoy, İlkay (2006-01-01)
In this study, a generative type probabilistic model is proposed for object recognition. This model is trained by weakly labelled images and performs classification and detection at the same time. When test on highly challenging data sets, the model performs good for both tasks (classification and detection).
A Unified approach for center-based clustering problems on networks
Eroğlu, Derya İpek; İyigün, Cem; Department of Industrial Engineering (2018)
In this thesis, Center-Based Clustering Problems on Networks are studied. Four different problems are considered differing in the assignment scheme of the data points and the objective function. Two different assignment schemes are considered, hard assignment and soft assignment. In hard assignment, data points (vertices) are strictly assigned to one cluster, while in soft assignment, vertices are assigned to the multiple clusters with a membership probability. Objective function of a clustering problem cou...
A Novel Parameter Identification Toolbox for the Selection of Hyperelastic Constitutive Models from Experimental Data
Dal, Hüsnü; Açıkgöz, Kemal (2017-10-13)
This paper presents a novel parameter identification toolbox based on various multi-objective optimization strategies for the selection of the best constitutive models from a given set of homogeneous experiments. The toolbox aims at providing an objective model selection procedure along with the material parameters for the rubber compound at hand. To this end, we utilize the multi-objective optimization using genetic algorithm of MATLAB. For the validation purposes, we use 10 constitutive laws.
A LINEAR MATHEMATICAL-MODEL FOR THE SEISMIC INPLANE BEHAVIOR OF BRICK MASONRY WALLS .2. DETERMINATION OF MODEL PARAMETERS THROUGH OPTIMIZATION USING EXPERIMENTAL-DATA
Sucuoğlu, Haluk; McNiven, Hugh (Wiley, 1984-01-01)
The parameters appearing in the mixture and effective modulus models proposed in Part 1 are determined through optimization by matching theoretical and experimental responses. The optimization analysis is performed in frequency space. The response quantities chosen to be matched are the complex frequency response functions (experimental and theoretical) relating the Fourier transforms of top and base accelerations of the wall. Computations in optimization analysis are carried out by introducing an object (e...
A Comparative Study on the Hyperelastic Constitutive Models for Rubber
Dal, Hüsnü (2018-10-25)
This paper presents a novel approach for comparison of constitutive models forrubber. Model parameters are identified by simultaneous fitting of a given data of a set ofhomogeneous experiments (uniaxial tension, equibiaxial tension, and pure shear, Figure 1) bymultiobjective optimization. Using these parameters, fitting error can be calculated. In thisstudy, however, quality of fit parameter, which represents the amount of error normalized bythe data, is calculated for unbiased comparison of models. 40 diff...
Citation Formats
T. Bahcecioglu, S. Ozmen, and Ö. Kurç, “A Comparative Study on Two Different Direct Parallel Solution Strategies for Large-Scale Problems,” 2009, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/54056.