THE EFFECTS OF NUMERICAL METHODS ON THE STATISTICAL COMPARISON BETWEEN EXPERIMENTS AND SIMULATIONS OF SHOCKED GAS CYLINDERS.

Conference · Mon Jan 01 04:00:00 EST 2001

OSTI ID:975905

Rider, William; Kamm, J R ^[1]; Tomkins, C D ^[2]; Prestridge, K P ^[3]; Rightley, P M ^[4]; Benjamin, R F ^[5]; Zoldi, C A ^[6]

---

1. James R.
2. Chris D.
3. Katherine P.
4. Paul M.
5. Robert F.
6. Cindy A.

Validation of numerical simulations, Le., the quantitative comparison of calculated results with experimental data, is an essential practice in computational fluid dynamics. These comparisons are particularly difficult in the field of shock-accelerated fluid mixing, which can be dominated by irregular structures induced by flow instabilities. Such flows exhibit non-deterministic behavior, which eliminates my direct way to establish correspondence between experimental data and numerical simulation. We examine the detailed structures of mixing experiments and their simulation for Richtmyer-Meshkov (RM) experiments of Prestridge et al., Tomkins et al., and Jacobs. Numerical simulations of these experiments will be performed with several different high-resolution shock capturing schemes, including a variety of finite volume Godunov methods. We compare the experimental data for cOnfigurations of one and two diffuse cylinders of SF6 in air with numerical results using several multiscale metrics: fractal analysis, continuous wavelet transforms, and generalized correlations; these measures complement traditional metrics such as PDFs, mix fractions, and integral mixing widths.

• Research Organization: Los Alamos National Laboratory
• Sponsoring Organization: DOE
• OSTI ID: 975905
• Report Number(s): LA-UR-01-6762
• Country of Publication: United States
• Language: English

Related Subjects

97 MATHEMATICS AND COMPUTING
AIR
COMPUTERIZED SIMULATION
FLUID MECHANICS
FRACTALS
GAS CYLINDERS
METRICS
SIMULATION
VALIDATION