Keywords: computational steering, interactive simulation, distributed computing, message-passing interface, high-performance computing, virtual reality, computational fluid dynamics, lattice Boltzmann method.

separated preprocessing, computation and postprocessing steps is fused into one simultaneous work flow. In this paper we describe a computational steering environment which enables its user to visualize simulation results on-the-fly and to interact with the running simulation. The main modules constituting the computational steering application, i.e. the simulation kernel, preprocessing, postprocessing and steering, as well as the communication layout between them are introduced. The main focus is laid on the optimization of the data flow between these components.

We address the central aspects, problems, and principle approach to realizing a computational steering application. In particular, the utilization of supercomputers for fluid simulations and high-end visualization techniques are discussed. In the application presented it is possible to load arbitrary geometries exported from CAD systems without any special preparation. Other approaches, so far, support only predefined parameterized objects. This powerful interaction possibility with regard to the geometrical layout is the main feature distinguishing our approach from all other current computational steering approaches. Besides interacting with the geometry, the user can also modify flow parameters, define new or change existing boundary conditions during runtime. For advanced users or benchmarking purposes even optimization options can be changed during the runtime.

The applicability of the computational steering application presented is demonstrated by simulating a real operating room. Comparison to simulations with a very fine grid resolution show that even for moderately-sized grids with grid points every 3.5cm, a good qualitative estimate of the flow can be made after only about 50 seconds. This enables an engineer to quickly test several setups and experiment with them interactively during the simulation within only a short amount of time. After examining the principle fluid behavior, a few carefully selected test cases can be run additionally in a more detailed offline simulation for quantitative analysis.

By using an integrated front-end for simultaneous visualization and steering augmented through a virtual-reality environment this kind of interactive simulation tool becomes a true intuitive means of exploration and allows to quickly gain insights even when studying complex flow phenomena.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £85 +P&P)