A General Pressure Generation Model for Granular Propellant Fires
Dr. Frederick Paquet; General Dynamics – Ordnance and Tactical Systems, Canada, Valleyfield; Valleyfield,
Quebec, Canada
Dr. Hoi Dick Ng; Concordia University; Montreal, Quebec, Canada
Mario Paquet; General Dynamics – Ordnance and Tactical Systems, Canada, Valleyfield; Valleyfield, Quebec,
Canada
Keywords: fire safety, deflagration, explosion pressure, granular propellants, deflagration venting.
Abstract
Propellants fires can generate pressures that have disastrous consequences for the environment surrounding the event.
The main goal of this work is to provide a way to predict the pressure evolution inside a semi-vented enclosure during
a propellant fire. A general approach was taken in which the conditions of a semi-confined propellant fire are
transcribed into a set of differential equations. In order to verify the validity of the theoretical model and further adjust
some of the input parameters, tests were performed in vented enclosures. Several enclosure volumes varying from 60-
L to 1800-L were used to check the geometrical scaling effect. Granular propellant of various compositions and
geometries were tested. To quantify the unknown variables, an empirical approach was taken where the results of fire
tests were analyzed to yield the desired parameters. The general methodology applied involved comparing statistical
models derived from the experimental data and fitting the corresponding parameters with the theoretical model. To
solve such parameter estimation problems (also known as inverse problems), multivariate regression methods are
applied with a proper data regularization scheme. The results are simple model equations with parameters of known
values which account for changes in propellant configurations and event scales.
1. Introduction
Propellants are designed to provide mechanical energy through the action of the pressure generated by the
transformation of solid grains to a high temperature gas. For that same reason, an unwanted combustion will generate
pressures that can have disastrous consequences for the environment surrounding the fire. Propellants at very low
densities (small mass in a large volume) can easily generate pressures that far exceed what standard walls can resist.
When there is some confinement of the propellant, pressures in excess of 70 kPa can easily be generated at 12 m from
the event (Racette, Brousseau and Valliere 2004)
There are very few publications of pressure measurements made with propellant fires in open areas. Test results related
to storage areas and naval vessel compartments have been published, but often do not include a theoretical analysis,
or cover only certain limited cases (White et al. 2000) (Joachim 1991). A paper by Polcyn and Mullin examined the
pressure generation during airbag propellant fires in a 5.3 m
3
vented enclosure (Polcyn and Mullin 1998). An electric
detonator located at the bottom of the samples was used to ignite the propellants. Such a configuration can be
problematic, as it would generate a large amount of projections and thus yield potentially erratic results. An attempt
at modelling explosion pressure venting was made by Graham in cases involving the slow burning of high explosives
(such as RDX and Composition B) (Graham 1986). The model involved building a pressure-time derivative equation
by taking the difference between a pressure rise and pressure decay term. Each term was determined from basic
thermodynamics and gas flow dynamics principles. From the resulting equation, a critical vent area ratio was defined
as the solution yielding a pressure-time derivative of zero.
