Sensitivity Examination of the airEXODUS Aircraft Evacuation Simulation Model.
Sensitivity Examination of the airEXODUS Aircraft
Evacuation Simulation Model.
(1435 K)
Bukowski, R. W.; Peacock, R. D.; Jones, W. W.
International Aircraft Fire and Cabin Safety Research
Conference. Proceedings. November 16-20, 1998,
Atlantic City, NJ, 1-14 pp, 1998.
Keywords:
aircraft certification; behavior models; evacuating
(transportation); evacuation time; passenger vehicles;
toxic hazards
Abstract:
The Building and Fire Research Laboratory at the
National Institute of Standards and Technology (NIST)
has been a leader in the development, application and
evaluation of models for the simulation of fires and the
associated hazards and risk to people. NIST's expertise
is useful in the broader application of fire hazard
analysis to transportation systems in general. Such a
project related to passenger trains has been ongoing for
several years, sponsored by the Federal Railroad
Administration through the Volpe National Transportation
Systems Center. For aircraft, a fire hazard analysis
could determine not only the time required for
evacuation, but also the time available, based on the
fire performance of the total system. Such an
application requires both predictive models of the fire
environment and of passenger evacuation during fires.
The Federal Aviation Administration (FAA) is interested
in the use of computer simulation models for examining
the compliance of commercial aircraft with FAA
regulations for the evacuation of passengers and crew.
This interest derives from both a desire to reduce the
expense of testing every aircraft configuration and to
reduce dependence on evacuation tests involving people
and the attendant risk of injury in such tests. The
FAA's plan is to allow the use of modeling to qualify
derivative aircraft, those that exhibit variations in
interior configuration or stretch versions, for which
demonstration tests were performed on the basic version.
There is also interest in using models to assess
evacuation for advanced designs that may vary in
significant ways from current passenger aircraft. An
appropriately validated model could be used both to test
compliance of derivative designs and to evaluate
evacuation as part of the design of new aircraft prior
to actual compliance testing. In the longer term, the
coupling of an evacuation model with a fire model would
allow designers to evaluate evacuation under more
realistic fire conditions. One candidate model
specifically developed to simulate the evacuation of
commercial aircraft is airEXODUS developed at the
University of Greenwich in the United Kingdom. The
developers have been conducting extensive verification
of the predictive accuracy of the model using data from
the historical records of certification tests as a step
in developing the capability of the model to simulate
evacuation during a fire. The FAA asked the National
Institute of Standards and Technology to examine the
sensitivity of the airEXODUS model to reasonable
variation in user inputs to determine if the model
results might be unduly impacted without being obvious
to those reviewing the results. A copy of the latest
pre-release version (v 1.Ol) of the model was supplied
to NIST by the developers for this purpose. NIST has
studied the evacuation times for a simple geometry (a
passenger rail coach car with exits at both ends) and
found the times predicted by airEXODUS and two other
emergency evacuation models were nearly identical. The
following describes the results of an examination of
airEXODUS as applied to two aircraft designs along with
pertinent observations. However, any decision of the
suitability of the model for regulatory purposes can
only be judged by the FAA.
Building and Fire Research Laboratory
National Institute of Standards and Technology
Gaithersburg, MD 20899