Voici les éléments 1 - 10 sur 15
  • Publication
    Accès libre
    An Improved Parallel Multiple-Point Algorithm Using a List Approach
    Among the techniques used to simulate categorical variables, multiple-point statistics is becoming very popular because it allows the user to provide an explicit conceptual model via a training image. In classic implementations, the multiple-point statistics are inferred from the training image by storing all the observed patterns of a certain size in a tree structure. This type of algorithm has the advantage of being fast to apply, but it presents some critical limitations. In particular, a tree is extremely RAM demanding. For three-dimensional problems with numerous facies, large templates cannot be used. Complex structures are then difficult to simulate. In this paper, we propose to replace the tree by a list. This structure requires much less RAM. It has three main advantages. First, it allows for the use of larger templates. Second, the list structure being parsimonious, it can be extended to include additional information. Here, we show how this can be used to develop a new approach for dealing with non-stationary training images. Finally, an interesting aspect of the list is that it allows one to parallelize the part of the algorithm in which the conditional probability density function is computed. This is especially important for large problems that can be solved on clusters of PCs with distributed memory or on multicore machines with shared memory.
  • Publication
    Accès libre
    Optimisation issues in 3D multiple-point statistics simulation
    (: Julián M. Ortiz and Xavier Emery, Mining Engineering Department, University of Chile., 2008-12) ;
    Walgenwitz, Alexandre
    ;
    Froidevaux, Roland
    ;
    ;
    Multiple-point statistics simulation has gained wide acceptance in recent years and is routinely used for simulating geological heterogeneity in hydrocarbon reservoirs and aquifers. In classical implementations, the multiple-point statistics inferred from the reference training image are stored in a dynamic data structure called search tree. The size of this search tree depends on the search template used to scan the training image and the number of facies to be simulated. In 3D applications this size can become prohibitive. One promissing avenue for drastically reducing the RAM requirements consists of using dynamically allocated lists instead of search trees to store and retrieve the multiple–point statistics. Each element of this list contains the identification of the data event together with occurence counters for each facies. First results show that implementing this list based approach results in reductions of RAM requirement by a factor 10 and more. The paper discusses in detail this novel list based approach, presents RAM and CPU performance comparisons with the (classical) tree based approach.
  • Publication
    Accès libre
    Hele-Shaw approximation for resin transfer molding
    A Hele-Shaw flow is studied as the limit of a Stokes flow. The velocity fields associated to each fluid solve a fluid dynamic problem with time dependent domain. The moving boundary formulation is based on a pseudo-concentration equation. Some existence of solution of this set of equations is given using the classical Leary-Schauder Theorem and the use of space-time integrated least square method for the transport equation.
  • Publication
    Accès libre
    Le cadre physique [Le lac de Neuchâtel]
    (2004)
    Persoz, Francis
    ;
    Schaer, Jean-Paul
    ;
    ;
    Robert-Charrue, Charles
  • Publication
    Accès libre
    Space-time integrated least squares: a time-marching approach
    (2004) ;
    de Montmollin, Gautier
    A time-marching formulation is derived from the space-time integrated least squares (STILS) method for solving a pure hyperbolic convection equation and is numerically compared to various known methods. Copyright (C) 2004 John Wiley Sons, Ltd.
  • Publication
    Accès libre
    PATUMOD: a compartment model of vegetation dynamics in wooded pastures
    A system of wooded pasture can be described by seven biological state variables (trees, shrubs, underwood grasslands, fallows, eutrophic meadows, oligrotrophic lawns and cattle) linked by a network of dynamic interactions, which are controlled by altitude and human activities. PATUMOD is a spatially implicit compartment model designed to simulate vegetation dynamics in such silvopastoral ecosystems at community level and according to an equilibrium paradigm. Computer simulations show that the state variables generally end up on a steady-state (one-point attractor), independent on their initial values but strongly dependent on cattle load. At a given altitude, to each value of the stock density is corresponding a stable equilibrium characterised by a given relative cover of each vegetation component. If the initial values are very far of the attractor, a long succession of intermediate stages is required before leading to the steady-state. A remarkable exception to this rule can occur at low altitude, with a repellor between switching trajectories towards two attractors, corresponding to a threshold between scarcely and densely wooded pastures. PATUMOD can be applied to simulate different management scenarios, which include changing global stock density and cutting trees or shrubs. (C) 2002 Elsevier Science B.V. All rights reserved.
  • Publication
    Métadonnées seulement
    Band preconditioners: Application to preconditioned conjugate gradient methods on parallel computers
    (: Nova Sciences Publisher, 2001) ;
    Dongarra, Jack
    ;
    Kontoghiorghes, Erricos John
  • Publication
    Accès libre
    A model of displacement and strain for arc-shaped mountain belts applied to the Jura arc
    (2000-09-01)
    Hindle, David
    ;
    ;
    Burkhard, Martin
    A plan view geometric model for simple, parallel, differential displacements is presented. As an analogue for models of arcuate mountain belt formation we use the model to predict strain patterns produced by parallel displacement in front of a rigid versus deformable indenter. A rigid indenter is simulated by an irregular quadrilateral displaced a constant amount along its hinterland boundary. A deformable indenter is simulated by an irregular quadrilateral displaced along a hinterland boundary that is allowed to rotate. Some simple test cases show that the deformable indenter model leads to a pattern of strain very similar to that encountered in arcuate mountain belts. Short axes orientations are most deviated from the transport direction along the lateral edges of the model with minimum displacement and shortening and vary across the central domain, mirroring strain features with variable orientations from mountain belts such as fold trends and minor deformation features such as horizontal stylolite peaks. The rigid indenter model also generates short axes deviated from the transport direction but showing far less variation. Displacement–strain relationships from the Jura mountains (Switzerland and France) are quantified based upon a simplified version of the finite displacement field for the Jura fold–thrust belt of Philippe, Y. (1995) ["Rampes latérales et zones de transfer dans le chaines plissées". (Unpublished PhD thesis, Université de Savoie)]. We find that the model short axis orientation pattern is very similar over at least the eastern and central Jura to the stylolite patterns from the region whilst the long axes closely match the fold axes trends. The model suggests that the Jura mountains could have formed as a result of a progressive deformation with uniform transport in a general northwest direction.