Simplified Combustion Modeling of Double Base Propellant: Gas Phase Chain Reaction Vs. Thermal Decomposition

Abstract

Simplified combustion modeling of nitrocellulose (NC), nitroglycerin (NG) double base propellant is considered. Two models with simple but rational chemistry are compared: the classical thermal decomposition, high gas activation energy (Eg/RT> > 1) Denison-Baum-Williams (DBW) model, and a new chain reaction, low gas activation energy (Eg/RT < < 1) model recently proposed by Ward, Son, and Brewster (WSB). Both models make the same simplifying assumptions of constant properties, Lewis number unity, single-step, second order gas phase reaction, and single-step, zero order, high activation energy condensed phase decomposition. The only difference is in the gas reaction activation energy Eg which is asymptotically large for DBW and vanishingly small for WSB. The results show that within the same set of assumptions (those listed above plus constant heat release and radiative heat feedback) the WSB model more accurately predicts the steady-state gas phase temperature profile, burning rate sensitivity parameters and oscillatory combustion response (pressure- and radiation-driven) for NC/NG propellant than the DBW model. The oscillatory results support our earlier finding that for accurate unsteady predictions it is necessary to use the full AEA decomposition expression of Lengelle and Ibiricu/Williams, with its inherent negative Jacobian parameters (ns,q < 0), rather than the usual constant-prefactor Arrhenius pyrolysis relation (ns,q = 0). All thermophysi-cal, thermochemical, and chemical kinetic parameters are consistent with what is known about detailed chemistry of NC/NG combustion. Initiation in the condensed phase, thought to occur by CO-NO2 bond homolysis, is represented by Eg = 40 kcal/mol. The primary flame zone, thought to be dominated by NO2/aldehyde reactions with an effective activation energy of Eg ∼ 5 kcal/mol, is better represented by a vanishingly small value of Eg (WSB) than an asymptotically large value (DBW). The main implication of this finding is that the important (regression rate determining) gas reaction zone near the surface has more the character of chain reaction than thermal decomposition. An important consequence is that the burning rate may be more sensitive to condensed phase decomposition kinetics than previous simplified models have allowed. Also, the BDP monopropellant model is shown to be essentially equivalent to the gas kinetically controlled DBW model and a quantitative BDP-DBW link is demonstrated.