Generating Rich Data Sets for Fire Model Validation
Spatially and temporally rich
data sets are helping Sandia
researchers understand and
corral fire’s unpredictable ways.
Fire has always been one of humankind’s
greatest threats and one of its most useful
tools. To modern-day science, it is an exquisitely
complex chemistry problem that results
in a self-lofting, turbulent plume, which is
engulfed in a sea of mostly unseen infrared
photons. Now, high-tech laser-based diagnostics
combined with new world-class fire
laboratories in Sandia’s new Thermal Test
Complex and state-of-the-art multiphysics fire
simulation tools, optimized to take advantage
of large-scale, high-performance computers,
are giving Sandia researchers a look into the
world of large-scale, complex fires.
Thermal Test Complex allow for
controlled experiments that permit the
development of better fire models.
These techniques, in turn, are returning
spatially and temporally rich data that are
helping to better predict how fire reacts and
how a weapons system might respond in a
fire. Modeling and simulation is the modern
application of theory in which the numerical
simulation tools are the codification of
our theoretical understanding, and the
application simulations are scientific
hypotheses. Experiments test the hypothesis
and the quantitative comparisons validate
our current level of understanding.
Knowledge gained is then codified. The goal
is to quantify our uncertainties to establish
weapon system safety in fire environments.
An example of a fire simulation of a validation
experiment, done on a complex calorimeter
in a well-controlled fire, can be seen in the
illustration (a) on this page.
The codified knowledge permits fire
scientists to design experiments to challenge
systems realistically. Fire simulations like the
one above represent 385 million unknowns,
calculated with tens of thousands of time
steps. The image here depicts a fire in a crosswind
at a snapshot in time.
High-fidelity experiments often result
in new knowledge. An example (b) is a
sequence of images showing bubble and
spike structures, mixing at the edge of a
simple helium plume, as the plume becomes
unstable. An edge instability results in a
circular vortex that grows, engulfing the
bubble and spike structures. This vortex
draws in heavy air over the light helium and
the cycle repeats itself. The same dynamics
occur in a fire.
The codified knowledge permits fire
scientists to design experiments to challenge
systems realistically. Fire simulations like the
one above represent 385 million unknowns,
calculated with tens of thousands of time
steps. The image here depicts a fire in a crosswind
at a snapshot in time.
For more information:
Sheldon R. Tieszen, Ph.D., 505-844-6526,
srtiesz@sandia.gov
