Preliminary Sensitivity Studies With CRASH 3D Bruce Fryxell CRASH Review October 20, 2009.

Slides:



Advertisements
Similar presentations
Page 1© Crown copyright 2007 Constraining the range of climate sensitivity through the diagnosis of cloud regimes Keith Williams 1 and George Tselioudis.
Advertisements

Active Appearance Models
Assessing Uncertainties in Radiative Shock Modeling James Paul Holloway University of Michegan Joslin Goh, Mike Grosskopf, Bruce Fryxell, Derek Bingham.
Sensitivity Analysis In deterministic analysis, single fixed values (typically, mean values) of representative samples or strength parameters or slope.
Progress Report on SPARTAN Chamber Dynamics Simulation Code Farrokh Najmabadi and Zoran Dragojlovic HAPL Meeting February 5-6, 2004 Georgia Institute of.
Effective Reinforcement Learning for Mobile Robots Smart, D.L and Kaelbing, L.P.
Collaborative Comparison of High-Energy-Density Physics Codes LA-UR Bruce Fryxell Center for Radiative Shock Hydrodynamics Dept. of Atmospheric,
Topic 6: Introduction to Hypothesis Testing
Momentum transport and flow shear suppression of turbulence in tokamaks Michael Barnes University of Oxford Culham Centre for Fusion Energy Michael Barnes.
Peyman Mostaghimi, Martin Blunt, Branko Bijeljic 11 th January 2010, Pore-scale project meeting Direct Numerical Simulation of Transport Phenomena on Pore-space.
Active Calibration of Cameras: Theory and Implementation Anup Basu Sung Huh CPSC 643 Individual Presentation II March 4 th,
Longitudinal Experiments Larry V. Hedges Northwestern University Prepared for the IES Summer Research Training Institute July 28, 2010.
Quantifying Uncertainties in Radiative Shock Experiments Carolyn C. Kuranz CRASH Annual Review Fall 2010.
1 Learning to Detect Objects in Images via a Sparse, Part-Based Representation S. Agarwal, A. Awan and D. Roth IEEE Transactions on Pattern Analysis and.
National Center for Earth-surface Dynamics Modeling physical and ecological dynamics of channel systems that shape Earth’s surface Moving boundary problems.
The Influence of Radiative Transfer on SPH Simulations of Star Formation Stuart C. Whitehouse and Matthew R. Bate
Diffusion Model Error Assessment Jim E. Morel Texas A&M University CRASH Annual Review October 29, 2010.
CRASH UQ Program: Overview & Results James Paul Holloway CRASH Annual Review Fall 2010.
Simulations of the Experiments Ken Powell CRASH Review October, 2010.
Assessment of Predictive Capability James Paul Holloway CRASH Review Meeting October
EFFECTS OF CHAMBER GEOMETRY AND GAS PROPERTIES ON HYDRODYNAMIC EVOLUTION OF IFE CHAMBERS Zoran Dragojlovic and Farrokh Najmabadi University of California.
Chamber Dynamic Response Modeling Zoran Dragojlovic.
Center for Radiative Shock Hydrodynamics Integration for Predictive Science R. Paul Drake.
Lecture 10 Comparison and Evaluation of Alternative System Designs.
Using ranking and DCE data to value health states on the QALY scale using conventional and Bayesian methods Theresa Cain.
A comparison of radiation transport and diffusion using PDT and the CRASH code Fall 2011 Review Eric S. Myra Wm. Daryl Hawkins.
Forecasting and Numerical Weather Prediction (NWP) NOWcasting Description of atmospheric models Specific Models Types of variables and how to determine.
AIAA th AIAA/ISSMO Symposium on MAO, 09/05/2002, Atlanta, GA 0 AIAA OBSERVATIONS ON CFD SIMULATION UNCERTAINITIES Serhat Hosder,
Code Comparison and Validation LA-UR Bruce Fryxell Center for Radiative Shock Hydrodynamics Fall 2011 Review.
Development of WRF-CMAQ Interface Processor (WCIP)
Segmental Hidden Markov Models with Random Effects for Waveform Modeling Author: Seyoung Kim & Padhraic Smyth Presentor: Lu Ren.
Cosmic Rays in the Heliosphere J. R. Jokipii University of Arizona I acknowledge helpful discussions with J. Kόta and J. GIacalone. Presented at the TeV.
Chapter 13 ANOVA The Design and Analysis of Single Factor Experiments - Part II Chapter 13B Class will begin in a few minutes. Reaching out to the internet.
Governing equations: Navier-Stokes equations, Two-dimensional shallow-water equations, Saint-Venant equations, compressible water hammer flow equations.
Center for Radiative Shock Hydrodynamics Fall 2011 Review Assessment of predictive capability Derek Bingham 1.
Objectives 2.1Scatterplots  Scatterplots  Explanatory and response variables  Interpreting scatterplots  Outliers Adapted from authors’ slides © 2012.
December 9, 2014Computer Vision Lecture 23: Motion Analysis 1 Now we will talk about… Motion Analysis.
CFD Refinement By: Brian Cowley. Overview 1.Background on CFD 2.How it works 3.CFD research group on campus for which problem exists o Our current techniques.
Georgia Institute of Technology Initial Application of the Adaptive Grid Air Quality Model Dr. M. Talat Odman, Maudood N. Khan Georgia Institute of Technology.
Center for Radiative Shock Hydrodynamics Fall 2011 Review PDT and radiation transport Marvin L. Adams.
Using the model and algorithm shown to the right, we obtain the theoretical images above. These images, with A=4.2*10 7 cm 3 /s (defect pixel) and A=8.2*10.
Order of Magnitude Scaling of Complex Engineering Problems Patricio F. Mendez Thomas W. Eagar May 14 th, 1999.
Jeff J. Orchard, M. Stella Atkins School of Computing Science, Simon Fraser University Freire et al. (1) pointed out that least squares based registration.
Introduction to LDA Jinyang Gao. Outline Bayesian Analysis Dirichlet Distribution Evolution of Topic Model Gibbs Sampling Intuition Analysis of Parameter.
TEMPLATE DESIGN © A high-order accurate and monotonic advection scheme is used as a local interpolator to redistribute.
Electromagnetic Waves
Variational data assimilation: examination of results obtained by different combinations of numerical algorithms and splitting procedures Zahari Zlatev.
A COMPUTATIONAL STUDY OF THE EFFECT OF EXIT GEOMETRY ON SHARKSKIN INSTABILITY IN THE EXTRUSION OF A NON-NEWTONIAN VISCOELASTIC POLYMER Amanda Pascoe
MCMC reconstruction of the 2 HE cascade events Dmitry Chirkin, UW Madison.
Geant4 Tracking Test (D. Lunesu)1 Daniela Lunesu, Stefano Magni Dario Menasce INFN Milano GEANT4 TRACING TESTs.
Jets and α S in DIS Maxime GOUZEVITCH Laboratoire Leprince-Ringuet Ecole Polytechnique – CNRS/IN2P3, France On behalf of the collaboration On behalf of.
1 Tom Edgar’s Contribution to Model Reduction as an introduction to Global Sensitivity Analysis Procedure Accounting for Effect of Available Experimental.
Clustering Approaches Ka-Lok Ng Department of Bioinformatics Asia University.
Beam-Beam simulation and experiments in RHIC By Vahid Ranjbar and Tanaji Sen FNAL.
Computer Graphics Lecture 30 Mathematics of Lighting and Shading - IV Taqdees A. Siddiqi
1 A latent information function to extend domain attributes to improve the accuracy of small-data-set forecasting Reporter : Zhao-Wei Luo Che-Jung Chang,Der-Chiang.
Shadow Detection in Remotely Sensed Images Based on Self-Adaptive Feature Selection Jiahang Liu, Tao Fang, and Deren Li IEEE TRANSACTIONS ON GEOSCIENCE.
A V&V Overview of the 31st Symposium on Naval Hydrodynamics
Chamber Dynamic Response Modeling
Analysis on Performance Controllability under Process Variability: A Step Towards Grid-Based Analog Circuit Optimizers Seobin Jung, Sangho Youn, Jaeha.
7/21/2018 Analysis and quantification of modelling errors introduced in the deterministic calculational path applied to a mini-core problem SAIP 2015 conference.
Parameterization for ATLAS EMEC
D. Odstrcil1,2, V.J. Pizzo2, C.N. Arge3, B.V.Jackson4, P.P. Hick4
Lecture Slides Elementary Statistics Twelfth Edition
The application of an atmospheric boundary layer to evaluate truck aerodynamics in CFD “A solution for a real-world engineering problem” Ir. Niek van.
Plants from Images CS 658.
Experience-Dependent Asymmetric Shape of Hippocampal Receptive Fields
Michael E Wall, James B Clarage, George N Phillips  Structure 
Emulator of Cosmological Simulation for Initial Parameters Study
Survey Networks Theory, Design and Testing
Presentation transcript:

Preliminary Sensitivity Studies With CRASH 3D Bruce Fryxell CRASH Review October 20, 2009

Page 2 Description of simulations Uniform 3D Cartesian grid Grid size – 1200 x 240 x 240 –A second set of simulations was performed with a grid size of 600 x 120 x 120 to test grid convergence Simplified Physics –Gamma-law equation of state for each material –Gray flux-limited diffusion All simulations initialized from Hyades 2D output at 1.3 ns using nominal values of input parameters

Page 3 Description of simulations 64 Three-dimensional simulations performed varying four input parameters –Be gamma (1.4 – ) –Xe gamma (1.1 – 1.4) –Be opacity scale factor ( ) –Xe opacity scale factor (0.7 – 1.3) Parameter combinations determined from Latin hypercube design All other parameters were fixed at their nominal values Each simulation used ~ 5 hours on 1024 cores of hera Entire set was completed in approximately 2 weeks

Page 4 Input parameter correlations (experiment design) Be Gamma Xe Gamma Be Opacity Xe Opacity Good coverage of input parameter space with relatively few points

Page 5 Reference solution Log Rho X Y Slice through z=0 plane Morphology of main shock differs significantly from experiments Wall shock appears similar to experiments

Page 6 Axis feature is robust Reference Solution Asymmetric Initial Conditions Rusanov Solver log rho Also investigated: Changing interface treatment Varying plastic opacity Multigroup diffusion 2D cylindrical grid For all tests, axis feature persists We are now investigating the possibility that the feature results from the Hyades initial conditions Inadequate grid resolution Inconsistent physics

Page 7 Sensitivity study of observable features Main shock location Wall shock angle Distance between wall and triple point 10 additional quantities were studied, but will not be discussed here, since they can not be measured in the experiments Output quantities of interest

Page 8 Main shock location is not very sensitive to variations of input parameters Range of values from experiment Be Opacity Scale Factor Xe Opacity Scale Factor Be Gamma Xe Gamma This study does not address sensitivity of shock location to these parameters during the first 1.3 ns, prior to the initialization of CRASH

Page 9 Shock location is not converged at low resolution Shock location is not converged at low resolution, but error is still less than variation in experiment High resolution uses the same grid size as the nearly converged 1D solution Multidimensional simulations may require higher resolution that 1D for convergence Be Opacity Scale Factor Xe Opacity Scale Factor Be Gamma Xe Gamma Red – Low resolution Blue – High resolution

Page 10 Wall shock angle is correlated only with Xe opacity Value from experiment is ~ 10° ± 2° Value from simulations is ~ 6° - 8° These values were determined in different ways and may not be inconsistent For future studies, all features will be extracted from radiographs in consistent manner Correlation between the Xe opacity and the wall shock angle makes physical sense Be Opacity Scale Factor Xe Opacity Scale Factor Be Gamma Xe Gamma

Page 11 Low resolution produces larger wall shock angles Be Opacity Scale Factor Xe Opacity Scale Factor Be Gamma Xe Gamma Red – Low resolution Blue – High resolution

Page 12 Triple point location shows weak correlation with Xe Gamma Value from experiment is ~ 6.7 ± 1 x m Value from simulations is ~ 7 – 10 x m Be Opacity Scale Factor Xe Opacity Scale Factor Be Gamma Xe Gamma Triple point location is not very sensitive to grid resolution

Page 13 Relative importance – Main shock location MARS MART The main source of variation in the shock location is the Xe opacity scale factor. However, the size of the variation is very small. The shock location may be more sensitive to these parameters during the first 1.3 ns, before the initialization of CRASH.

Page 14 Relative importance – Wall shock angle MARS MART Variations in the Xe opacity scale factor produce almost the entire variation in the wall shock angle.

Page 15 Relative importance –Triple point location MARS MART Variations in the Xe gamma produce most of the variation in the wall triple point location.

Page 16 Conclusions Location of main shock, the angle of the wall shock, and the location of the triple point show surprisingly good agreement with the experiments However, the morphology of the experimental results differs significantly from the simulations These differences appear to be robust for the given set of initial conditions Differences are likely due to either insufficient numerical resolution in the 2D Hyades initialization calculation, or inconsistent physics between Hyades and CRASH, which creates a strong transient in the solution

Page 17 Conclusions Location of main shock varies little with changes in Be and Xe equation of state and opacity Angle of wall shock depends only on the Xe opacity scale factor Location of triple point depends primarily on the Xe equation of state All conclusions must be regarded as preliminary until higher-fidelity simulations have been completed

Page 18 We expect to be cycle limited Current study required 2 weeks on hera A study with twice the resolution on a uniform grid would require approximately 1 year The use of AMR will reduce this significantly Future studies may require hundreds rather than tens of simulations Use of more sophisticated radiation transport algorithms will increase computer time dramatically We intend to make judicious use of two-dimensional and reduced physics simulations Two-dimensional simulations will not be possible for simulations of year five experiments Completing the project given the computer resources currently available will be a challenge

Page 19 Sample flow morphologies