Intelligent Light Breakfast Briefing
Wednesday Jan 9, 7:00am - 8:00am, in the Seaport G Ballroom at the Grand Hyatt San Diego
Start the day by hearing the latest from Intelligent Light, catching up with the FieldView community and having breakfast on us!
RSVP required as space is limited. If you would like to join us, please send email to firstname.lastname@example.org with your name, company and the number of people in your party.
We look forward to seeing you in San Diego!
Application of a CFD Uncertainty Quantification Framework for Industrial-Scale Aerodynamic Analysis
John Schaefer and Andrew Cary of The Boeing Company and Earl Duque and Seth Lawrence of Intelligent Light
In a joint effort between the Applied Research Group (Intelligent Light) and Boeing Research and Technology (The Boeing Company), the application of the latest developments in Uncertainty Quantification (UQ) are demonstrated as case studies for industrial scale CFD problems. Starting with a 2D NACA 0012 airfoil at zero lift, then moving to a 3D full-scale aircraft in high lift configuration, the latest UQ frameworks are implemented to determine numerical and model input uncertainty. This effort provides a step towards standardizing the use of UQ as a design and engineering tool to build confidence and reliability in scientific computational engineering and design.
Assessment of Model Validation and Calibration Approaches in the Presence of Uncertainty
Thu, Jan 10, 4:00pm - 4:30pm, Harbor H
Nolan Whiting and Christopher Roy of Virginia Polytechnic Institute and State University, and Earl Duque and Seth Lawrence of Intelligent Light
This paper discusses the investigation and implementation of various methods for quantifying Model Form Uncertainty (MFU) in scientific computing. Metrics were obtained through rigorous implementation of several MFU methods to produce comparison studies that quantify the conservativeness of each method. The findings include which methods perform best under different circumstances and provide a guide to best practices when choosing a method for quantifying the model form uncertainty in a computational UQ study.
Summary of 2017 SciTech Computational Environments Special Session Toward the CFD Vision 2030
Fri, Jan 11, 9:30am - 10:00am, Vista B
Andrew Lofthouse, CREATE AV Team, Earl Duque, Intelligent Light, Roger Davis, University of California‚ Davis
Computational Environments can be defined as the hardware and software required to efficiently use high performance computing resources and to easily extract knowledge from that data. As such, they underlie nearly all large-scale simulations. The CFD Vision 2030 report published recommendations for advancements in all areas of computational fluid dynamics that are needed in order to take advantage of Exascale hardware that is expected to be in operation by the year 2030. This paper summarized a special session held at SciTech 2017 sponsored by the Meshing, Visualization and Computational Environments technical committee that looked in more detail at the current status of computational environments and where additional research is required in order to efficiently use Exascale machines.
The session consisted of technical presentations and a question and answer period with a panel of invited experts. Computational environments are vital for efficient use of heterogeneous hardware architectures, large-scale data management, interfaces between multidisciplinary domains, uncertainty quantification and error analysis, as well as automation of complex simulations.
Spectre: A Computational Environment for Managing Total Uncertainty Quantification of CFD Studies
Fri, Jan 11, 10:00am - 10:30am, Vista B
Earl Duque and Seth Lawrence of Intelligent Light
The future of uncertainty quantification as a tool for computational engineering and design hinges on the management, execution and visualization of complex data-driven workflows. Whether the data is historic experiments or modern computational simulations, UQ workflows require onerous dedication to the management and synchronization of all available data. To address the needs of complex UQ workflows, the Applied Research Group at Intelligent Light has developed Spectre: A Computational Environment for Managing Total Uncertainty Quantification of CFD Studies. By encompassing a database, server and user-interface into a RESTful web deployed platform, Spectre provides seamless integration of data management, solver execution and visualization.
MVCE-05: Visualization and Computational Environments
along with Co-Chair: Andrew Lofthouse , US Air Force Academy
Fri, Jan 11, 9:30am - 12:30pm, Vista B