Post-processing and Visualization of large-scale DEM Simulation Data with the open-source VELaSSCo Platform
Regardless of its origin, in the near future the challenge will not be how to generate data, but rather how to manage big and highly distributed data to make it more easily handled and more accessible by users on their personal devices. VELaSSCo (Visualization for Extremely Large-Scale Scientific Computing) is a platform developed to provide new visual analysis methods for large-scale simulations serving the petabyte era. The platform adopts Big Data tools/architectures to enable in-situ processing for analytics of engineering and scientific data and hardware-accelerated interactive visualization. In large-scale simulations, the domain is partitioned across several thousand nodes, and the data (mesh and results) are stored on those nodes in a distributed manner. The VELaSSCo platform accesses this distributed information, processes the raw data, and returns the results to the users for local visualization by their specific visualization clients and tools. The global goal of VELaSSCo is to provide Big Data tools for the engineering and scientific community, in order to better manipulate simulations with billions of distributed records. The ability to easily handle large amounts of data will also enable larger, higher resolution simulations, which will allow the scientific and engineering communities to garner new knowledge from simulations previously considered too large to handle. This paper shows, by means of selected Discrete Element Method (DEM) simulation use cases, that the VELaSSCo platform facilitates distributed post-processing and visualization of large engineering datasets.
CSG Ray Tracing Revisited: Interactive Rendering of Massive Models Made of Non-planar Higher Order Primitives
GRAPP 2017. Proceedings
International Conference on Computer Graphics Theory and Applications (GRAPP) <12, 2017, Porto, Portugal>
In many scientific and engineering areas, CAD models are constructed by combining simple primitives using Boolean set operations. Rendering such a dataset usually requires a preprocess, where the surface of the CAD model is approximated by an often highly complex triangle mesh. Real-time ray tracing provides an alternative to triangle rasterization as it allows for the direct visualization of (higher-order) solid and planar primitives without having to triangulate them. Additionally, Boolean compositing operations can be performed implicitly per ray, primitives have low storage requirements, and curved surfaces appear pixel-accurate. In this paper we demonstrate these properties using massive real-world CAD models.
Real-Time Parallel Streamsurface Computation
München, TU, Master Thesis, 2016
Streamsurfaces are one of the powerful visualization tools, which are used to gain insight into characteristics and features of flow fields. In practice, streamsurfaces are approximated by triangulating adjacent pairs of integral curves, originating from a seeding line. The generation of integral curves bears quite some similarities to ray tracing algorithms used in physically based renderers. Although, the techniques used in ray tracing may not have good performance in the streamline computation context due to their different computational nature, they can be optimized for streamline computation by introducing some modifications. In this master thesis, I present my work on accurate streamsurface computation and rendering in real-time, by exploiting the scalability and portability features of parallel architectures in heterogeneous computing, and utilizing concepts from physically based rendering. To improve the efficiency, I use a scheduler to divide the streamsurface computation and rendering tasks on different devices proportional to their computation powers. Additionally, I apply acceleration structures and the concepts of caching to improve the efficiency and utilization of streamsurface generation on modern GPUs and CPUs to achieve real-time results. Furthermore, the possible impact of applying ray-packing and ray-sorting to the streamline computation is investigated.