A powerful suite of tools optimized for each other, Cascade’s products provide a simulation workflow that is simple to use yet powerfully scalable. These tools will help you leverage all the compute resources you have access to.
Predictive high-fidelity simulations require meshes of suitable quality. By leveraging clipped Voronoi-diagrams, Cascade’s mesh generator is fast, scalable, and robust enough to process complex geometries. Users can easily introduce resolution where needed, run coarse simulations of complex geometries, and simulate moving geometries. Two tools comprise our mesh generator—Surfer and Stitch.
Discretized geometry can be difficult to produce cleanly (without flaws) and robustly. Surfer is a geometry preprocessor that manipulates discretized geometry in various ways including: identification and repair of surface quality issues, combining multiple geometries via Boolean operations, and transforming existing geometries. Users can ensure the surfaces they pass into Stitch are suitable for meshing without the need for design modification in CAD/CAE software.
Starting from a discrete representation of the domain, Stitch uses clipped Voronoi-diagrams to enable unique meshing capabilities: arbitrary mesh coarsening or refinement near complex geometry, design and visualization of large meshes with minimal compute resources, and highly-parallel execution. Leverage the same compute resources as the flow solver and quickly build your meshes without having to move any data.
Realizing the predictive benefits of LES requires much more than just turning on time-dependence and changing the turbulence model. Cascade’s flagship solver, CharLES, combines advanced numerical methods and models, and extreme scalability on CPU-based and GPU-based high-performance computing environments. Our solver discretization is a finite volume approach based on a generalization of the discrete entropy framework for unstructured meshes to treat a variety of flow regimes, including low speed flows, high-speed flows with shocks, and reacting flows. This approach leads to a stable, homogenous flux discretization without complex sensors, upwinding hybridization, or tuning of coefficients for stability. This unique solver discretization is combined with the latest sub-grid scale modeling and wall modeling to deliver a LES capability that shows remarkable robustness to grid resolution.
Dimensional reduction of the large amount of data produced by high-fidelity simulation is a supercomputing problem of its own. Cascade’s software comes with multiple parallel and serial tools to take the friction out of this process and get you to answers faster.
CharLES enables a robust set of post-processing capabilities designed specifically to aid in analyzing high-fidelity, time-dependent data. Data reduction is possible through a variety of probing commands which allows extraction of point data, fluxes, and probability distributions to name a few. A powerful expression parsing capability gives users the flexibility of manipulating raw simulation variables to quantities of interest. Post-processing is fully scalable to the size of the underlying computation. In addition, data export to third-party formats such as Tecplot, Ensight Gold, or VTK (Paraview) is supported.
A unique and powerful feature of CharLES is the ability to produce quantitative PNG images from a simulation. Fast, parallel rendering of planar and volume data is supported as well as iso- and boundary surface data. While viewable in any standard image reader, Cascade images are enriched with custom metadata that maps every pixel to its location in simulation space and any associated data value. Images naturally sample the complex, 3D domain at a scale of interest to the engineer. Sequences of quantitative images can be used to rapidly compute statistics, modes, and even highlight differences between cases.
Prediction of far-field radiated noise from LES requires a hybrid approach. First, the important scales of turbulence within the noise-producing region are resolved in the simulation. Second, the propagation of the small amplitude acoustic fluctuations from the near-field source region to the far-field is computed analytically as a post-processing step from the high-fidelity transient flow data recorded during the simulation. The Ffowcs Williams – Hawkings (FW-H) equation is one of the most commonly used hybrid methods, and an efficient, massively-parallel implementation of the FW-H formulation is available in Cascade’s suite of tools.
DMD and POD modal decomposition capabilities are available for both image and full field data.
The Cascade app manages every point in the simulation workflow—from surface preparation and meshing to simulation and post-processing. A lightweight, graphical user interface designed to run on a laptop or desktop, the app interacts remotely through secure file transfers. The app enables scalable 3D interrogation of complex engineering domains by leveraging the quantitative imaging capabilities of Cascade’s solvers. In addition to visualization, the app supports runtime modifications to simulation settings, a command glossary with auto-completion, and line plots of time varying quantities of interest.