Scientific Machine Learning at Sandia: A Spotlight on Operator Learning
Explore the source record for details and available documents.
Engineering topics
Publications and source records attributed to Stewart, James R..
Explore the source record for details and available documents.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
Abstract not provided.
The USACM Thematic Workshop on Uncertainty Quantification and Data-Driven Modeling was held on March 23-24, 2017, in Austin, TX. The organizers of the technical program were James R. Stewart of Sandia National Laboratories and Krishna Garikipati of University of Michigan. The administrative organizer was Ruth Hengst, who serves as Program Coordinator for the USACM. The organization of this workshop was coordinated through the USACM Technical Thrust Area on Uncertainty Quantification and Probabilistic Analysis. The workshop website (http://uqpm2017.usacm.org) includes the presentation agenda as well as links to several of the presentation slides (permission to access the presentations was granted by each of those speakers, respectively). Herein, this final report contains the complete workshop program that includes the presentation agenda, the presentation abstracts, and the list of posters.
Abstract not provided.
Caterpillar and Sandia continued its partnership to improve Sandia’s Dakota software product and effectively deploy it to Caterpillar engineering design and analysis problems. Dakota use by Caterpillar analysts identified several challenges, which led to Sandia enhancements and bug fixes to Dakota. These benefit both Caterpillar and the Sandia Dakota user community.
Explore the source record for details and available documents.
Abstract not provided.
Abstract not provided.
This paper presents the results of a computational study on the shock interference problems on a cylindrical body typical of an engine inlet cowl leading edge at a nominal Mach number of 16. The two-dimensional Navier-Stokes equations are solved assuming the flow to be in chemical and thermal equilibrium and using a finite element method. The algorithm employs a cell-centered fully implicit upwind scheme. Adaptively generated unstructured meshes of triangles and quadrilaterals are employed. Under certain conditions the finite element code resulted in oscillatory solutions for shock interference at Mach 16. Some of the causes of the unsteady behavior are identified, and to the extent possible, such situations are avoided in the present application. Two shock interference conditions involving a Type IV (supersonic jet) and a Type III (attaching shear layer) are solved. The results are compared with available experimental data and reasonable agreement is observed. A semi-empirical method is also used to estimate the maximum surface heat flux and static pressure.