DATE:

Friday, May 10th, 2019

TIME:

3:00 PM

LOCATION:

Donald P. Shiley Bioscience Center

SPEAKER:

M. Giselle Fernandez, Ph.D, Postdoctoral Researcher, X-Computational Physics-Verification and Analysis, Los Alamos National Laboratory.

ABSTRACT:

Dense layers of solid particles surrounding a high-energy explosive generate instabilities after detonation.
Conjectures as to the cause of these instabilities include imperfections in the casing, inhomogeneities in
the initial distribution of particles, characteristics of the particles, and others. In particular, a
multiphase detonation with a cylindrical configuration where the initial distribution of particles is
initially highly axisymmetric is studied. The main physical mechanism responsible for the instabilities
observed in the experiments is assumed to be the initial distribution of particles. Therefore, the particle
volume is perturbed in the simulations using azimuthal sinusoidal waves and the final distribution of the
particles is observed quantifying the amplification of the departure from axisymmetry.

An instability mechanism, that is possibly partially the result of non-classical Raleigh-Taylor and/or
Richmyer-Meshkov instabilities, is observed. We have called it channeling instability and it has two main
effects (i) accelerate particles located in low particle volume radial sectors and (ii) push particles
from low particle volume sectors to high particle volume sectors.

To quantify the particles departure from axisymmetry, a metric based on energy is constructed, which
depends on the parameters of a trimodal sinusoidal perturbation (amplitudes, wavelengths, and relative phases).
The metric dependence on the relative phases was found to be negligible and unimodal perturbations were found
to amplify the metric the most.

The simulations considered are expensive and surrogate models can achieve accurate predictions of the metric
dependence on the amplitudes and the wavelengths at a low computational cost. Linear regression is used to
construct low-fidelity, high-fidelity, and multi-fidelity surrogate models. It was found that for this
problem with this particular surrogate and basis functions, multi-fidelity models have better performance
compared with single-fidelity models.


Options for reducing the number of simulation used to construct surrogates while maintaining accuracy by
taking advantage of parametric symmetries were also explored. The inherent parametric symmetries of the
model were imposed while building the multi-fidelity surrogate and the performance compared with the
multi-fidelity surrogate without imposing symmetries. For a small number of high-fidelity points it was
found that the performance of the surrogate using symmetries is much better, however, for more than 100 HF
data points their performance is indistinguishable.

HOST:

Satchi Venkataraman

DOWNLOAD:

PDF File