Fidelity in visualizing large-scale simulations

Computer simulations are powerful tools frequently used today in many important applications, for example to build safer buildings, to crash-test an automobile before it is built, to stabilize the Pisa tower, to design artificial joints that are comfortable and durable, or to investigate what-if scenarios to avoid and best recover from natural or man-made disasters. The simulation codes have reached a very high-level of sophistication and, by running on powerful computing machinery, can accurately track with infinitesimal time steps dozens of physical properties of millions of interacting elements under extreme conditions. In order to take fully advantage of the bounty of information concealed in the data produced, visualization is a uniquely powerful tool since it caters to the sense that provides our highest bandwidth connection to the surrounding world. Unfortunately, simulation results are usually examined with graphics and visualization tools that are one or several steps behind the state-of-the-art. We describe our efforts of producing highfidelity visualizations of the results of large-scale simulations using the latest commercial rendering and animation systems. To this effect we built a scalable and reusable link between the software worlds of animation and simulation. Our system also offers a set of tools that allow integrating the results of the simulation in the surrounding scene, of great importance when the intended audience extends beyond the researchers that designed the simulation. We built our system as part of the efforts of a larger, interdisciplinary team to produce a high-quality, physically accurate visualization of the September 11 attack on the Pentagon. Introduction and Motivation Physical simulations have become an important tool to understand and analyze critical events, such as an airplane crash or an earth quake, and, routinely, to analyze the performance of devices. Using finite element analysis (FEA), several commercial programs offer sophisticated tools for evaluating and simulating physical events. By measuring specific Figure 1 Simulation data visualization using our system. quantities such as stress and strain, or simply by scrutinizing the unfolding of an event in small time steps, important insights can be had that illuminate specific strengths and weaknesses of structures and mechanical systems, as well as provide clues on how to improve the performance of engineering designs. Much work has been done to simplify and automate formulating an FEA problem from a given shape design, particularly in the context of discrete manufacturing. Furthermore, considering only the geometric aspect of this problem, geometric mesh generation addresses the key problem of converting a boundary-based shape representation, familiar from computer-aided design (CAD), into an equivalent pointand element-based representation suitable for describing the shapes to be analyzed by the FEA, as well as representing ambient space for flow problems. This work has migrated into many commercial software packages, such as Ansys, Fluent, Elfini, to name a few. Once the FEA has completed, the results of the analysis are then visualized using a variety of tools geared to the FEA representation. These standard visualization tools are adequate to show salient features of the analysis in a manner that is familiar to engineers and specialists, and are well-suited to discrete manufacturing. For communicating the results of a FEA to a public audience, however, the traditional way of visualizing is not optimal. When evaluating an abnormal event such as a plane crash, public and policy interests are also in play, and properly communicating the event to the wider audience suggests that the visualization of the results should be ready to be grasped by people who are unfamiliar with, e.g., false color schemes, and other traditional tools for accentuating engineering quantities. Moreover, a realistic and engaging visualization of the simulation results is useful as well for integration into virtual reality training and documentary presentation. In this paper, we address how to visualize with high fidelity FEA results for a forensic simulation of an event by combining the physical quantities and events computed by the FEA with realworld, natural environments, or even synthetic environments familiar form computer games. In particular, we discuss how to convert the FEA representation to ones suitable for commercial animation systems, how to add special effects, and how to situate the results in the surrounding, natural environment. Some aspects of integrating visualization with simulation have been discussed in [2]. and have focused on the overall problems of managing simulation complexity and federating commercial FEA systems with commercial visualization and animation systems. We discuss our subject in the context of a case study, the September 11 attack on the Pentagon. One of our collaborators at Purdue University had been invited to participate in the forensic study of the damage the Pentagon building sustained in that attack. To better understand the performance of the building, eventually reported in [1], several simulation studies were done in 2002 using the commercial FEA code LS-Dyna [7, 8]. The simulations culminated in two runs on an IBM Regatta supercomputer, and the data so obtained was first analyzed using the standard post-processor of LS-Dyna. See [3], Phase I, for the visualizations done using traditional postprocessing. It was clear at the time that the visualizations so obtained were interesting, but were inferior to visualizations common in computer graphics. This motivated us to investigate how to achieve high fidelity in visualizing the simulation using 3DS Max [21], a commercial rendering and animation software package.