Community Research and Development Information Service - CORDIS

3D visualisation of very large data sets

Newly developed techniques for vector field visualisation will equip scientists with the ability to analyse vast data sets generated by fluid flow simulations.
3D visualisation of very large data sets
As the size of numerical simulations continues to increase, efficient visualisation plays an increasingly key role in assisting scientists to obtain insight into the complex dynamics of fluid flows. However, 3D visualisation of large-scale vector fields is not feasible with current methods.

The EU-funded LARGEFLOWVIS (Large scale flow visualization and analysis) project focused on so-called integration-based visualisation methods that rely on the analysis of trajectories of massless particles. Such curves are ideal for the study of transport, mixing and other similar processes.

Researchers developed new algorithms for the computation of integral curves able to exploit a broad range of parallel computing architectures. These algorithms are well suited to clusters and supercomputers, enabling scientists to visualise and analyse very large data sets from simulations as well as measurements.

The new approach to data reduction adopted in LARGEFLOWVIS allows the visualisation of integral curves on commodity hardware. Scientists combined different techniques from applied mathematics, parallel computing and software design to address open problems in fluid flow dynamics.

The resulting capacity to analyse large data sets will facilitate fundamental scientific research with industrial applications ranging from aerodynamics to oceanography. This includes research related to open problems such as turbulence around wind turbines and multi-scale ocean flows.

Related information

Keywords

Data sets, vector field, fluid flow, numerical simulations, LARGEFLOWVIS
Record Number: 190939 / Last updated on: 2017-01-25
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top