CIMNE - Miguel Adolfo Pasenau de Riera

CIMNE Miguel A. Pasenau This web is the personal user page of Miguel A. Pasenau de Riera
Researcher and developer of GiD the personal pre and postprocessor.
GiD, TclTk and other Manual pages

Eurographics Association member.

Papers and Projects

2014 July
Detail-preserving mesh simplification for scientific visualization presented in the 11th. World Congress on Computational Mechanics (WCCM XI)
Abstract: Dealing with large simulation data is a growing challenge. Current computational capabilities allow engineers to tackle large and complex simulations. The visualization of these large-scale problems is becoming an issue as current solutions rely on powerful hardware resources very close to the data being generated. Scientists located at remote sites need to download the simulation data to its own computer for local visualization, resulting in large download times, or visualize the simulations results remotely, resulting is poor interaction. One way to accelerate the visualization is to provide a simplified view while the user interacts with the model. Given a polygonal mesh, the goal is to generate another mesh which approximates the underlying shape but includes fewer primitives. Early methods focused only on preserving the overall shape of the geometric model, whereas current methods also handle meshes with per-vertex attributes, like colours and texture coordinates, so that both the mesh shape and the mesh appearance are preserved. This work describes several detail-preserving techniques based on vertex clustering and quadric error functions to simplify efficiently large simulation meshes. These techniques have also been incorporated into GiD, a pre- and post-processor application. Authors: Miguel A. Pasenau and Carlos Andújar
PDF abstract, slides PDF ( 13 MB),
2013 September
Detail-preserving mesh simplification
Abstract: Mesh simplification is an important problem in computer graphics. Given a polygonal mesh, the goal is to generate another mesh which approximates the underlying shape but includes less polygons, edges and vertices. Early methods focused only on preserving the overall shape of the geometric model, whereas current methods also handle meshes with attributes (normal vectors, colors, texture coordinates) so that both the mesh shape and the mesh appearance are preserved. The goal of this master thesis is to develop, implement and test a mesh simplification algorithm able to simplify large models in-core using a vertex clustering algorithm. Several detail-preserving techniques will be examined and implemented and a new filter is proposed, taking into account geometry features and nodal defined attributes. We also review recent advances in spatial hash tables to achieve a more compact storage, and we analyze and evaluate their impact in the simplification process. Advisor: Carlos Andújar
PDF report ( 19 MB), slides PDF ( 21 MB), Linux 64 binaries ( 1.6 MB), MS Windows 64 binaries ( 1.2 MB), Source code ( 9.5 MB).
2012 May
Mesh Simplification for Visualization presented in the 6th GiD Convention 2012
Abstract: As size and complexity of simulation grows, the visualization of these results on modest computers is becoming a challenge. The amount of data needs to be reduced, preserving details in the regions of interest and simplifying other uninteresting regions, so that they can be handled at reasonable speed. Previous simplification algorithm has also extended to support attributes and preserve thin features, packed as library and included in GiD.
PDF slides ( 12,8 MB)
2011 June
Simplificación de Mallas de Triángulos ( Triangle Mesh Decimation in Spanish)
Abstract: An algorithm has been developed to simplify triangles meshes using the uniform vertex clustering scheme coupled with an cuckoo hybrid hash table. The approach proposed by Christopher DeCoro has been used but instead of using his probabilistic octree to store the occupied cells of the 3D grid, the cuckoo hybrid hash table proposed by Dan A. Alcantara has been implemented. This hybrid hash table combines a classical sparse perfect hashing and the newly developed cuckoo hashing and is used to accumulate the quadric error functions and the vertices of the cell. A entirely multi-core cpu algorithm has been developed and tested with more than thirty models in three different platforms. The simplification algorithm has been also enhanced by forcing the simplified vertex to remain in its cell, correcting he normals of flipped triangles, getting the collapsed triangles as lines and filtering out the repeated triangles and lines. The hybrid tree has been enhanced to support with 64 bit keys. Some experiments with the hybrid hash has also been presented.
PDF report ( 23 MB), slides PDF ( 6,1 MB), slides PPTX ( 11 MB), Source code ( 1,8 MB).
2011 May
Scalable System for Large Unstructured Mesh Simulation presented in Parallel CFD 2011 congress
Abstract: Dealing with large simulation is a growing challenge. Ideally for the wellparallelized software prepared for high performance, the problem solving capability depends on the available hardware resources. But in practice there are several technical details which reduce the scalability of the system and prevent the effective use of such a software for large problems. In this work we describe solutions implemented in order to obtain a scalable system to solve and visualize large scale problems. The present work is based on Kratos MutliPhysics framework in combination with GiD pre and post processor. The applied techniques are verified by CFD simulation and visualization of a wind tunnel problem with more than 100 millions of elements in our in-hose cluster in CIMNE.
PDF paper ( 1,4 MB), PDF slides ( 3,7 MB), Animation ( 10 MB).
2002 February
Textures in Simulation presented in the 1st GiD Conference 2002
Abstract: Textures, a well known and widespread used technique in games, can also be useful in simulation of manufacturing processes, not only as a tool to improve the appearance of the results, but also to validate the simulation code and reduce manufacturing costs. Textures, basically, are images glued to polygons. These textured images can be used to make the simulation's results understandable at a glance, to improve quality and manufacturing efficiency, for instance in the can industry, and to contrast the simulation with experimental results.
PDF paper ( 1,4 MB),

Videos in you tube:

Flash videos created with GiD