1
An Improved Method to Calculate the Gas Pressure History from Partially Confined Detonations
Charles Oswald Ph.D, P.E.; Protection Engineering Consultants, San Antonio, TX, USA
Abstract
This paper discusses an improved fast-running method to calculate the gas pressure history from an internal
detonation in an explosion room with venting. It’s development is an intermediate step towards a a final, well
validated, gas pressure prediction model for DDESB that is significantly more accurate compared to test data than
the current method. As a first step in this research project, gas pressure data from over 100 confined and partially
confined blast tests were gathered and put into a gas pressures database. The existing FRANG and BlastX codes
were then used to model each of the tests and the calculated peak gas pressures and impulses were compared to
measured values. Also, an initial improved methodology, which combined parts of FRANG and BlastX, was
compared to the test data. These comparisons showed that all three gas pressure prediction methods were generally
very conservative compared to the test data. Based on trends noted in these comparisons, the current improved
methodology was developed to include a calculated rise time history for the gas pressure and to consider the effects
of mass loss, room volume change due to vent panel movement, and energy loss when venting occurs during the rise
time. This improved version, which has been programmed into an Excel spreadsheet, compares much better to
measured gas pressure data than the existing methods. It is currently under review by the DoD. DDESB is also
sponsoring additional gas pressure testing to address a lack of gas pressure test data for explosion rooms with a low
loading density and large covered vent areas. This test data will be used to develop the final version of the improved
methodology.
Keywords: gas pressure, internal detonations, venting
Introduction
A large database of internal detonation tests have been performed to measure the quasistatic, or gas pressure, in the
explosion room from an internal detonation, including many tests where the measured pressure history is published.
These tests have been gathered into a electronic database for the DDESB (Oswald, 2017). Most of these tests were
performed in the 1970s and 1980s by the U.S. Navy, and were used to develop the empirical equations for the
FRANG computer code that is currently used by the DoD Explosive Safety Board (DDESB) to calculate the gas
pressure in the explosion room for explosive safety designs. FRANG calculates the peak gas pressure in the
explosion room and venting of the gas pressure through openings in the room to the atmosphere, including openings
that are initially covered with a panel (Wager and Connett, 1989). For the case of a covered opening, the internal
shock and gas pressure are assumed to cause the panel to move away from the opening as a rigid body with
negligible attachments to the wall or roof of the explosion room. The calculated vent area at each time step is equal
to the panel perimeter multiplied by the distance the panel has deflected away from the explosion room, up to the
full area of the initially covered opening (Tancreto and Helseth, 1984). Modeling venting through vent areas that are
originally covered by panels is a critical capability because many DoD operation bays have large, lightweight
exterior walls and/or roofs that are intended to fail quickly in the event of an accidental internal detonation to allow
venting as they are blown away from the bay.
Other agencies in the DoD use the BlastX computer code to calculate the shock and gas blast pressures in the
explosion room, and surrounding rooms, from internal detonations. BlastX uses a thermochemical approach to
calculate the peak gas pressure, which is based on the chemical reaction equation for the explosive and the heats of
formation for the chemical products and reactants. It uses a first principles approach to calculate venting through
uncovered vent areas by modeling this as an adiabatic process for isentropic flow of an ideal gas through a nozzle
(Bessette and Emmanuelli, 2014). BlastX calculates venting through an opening that is initially covered by
