A typical composite solid propellant is a mixture of a crystalline oxidizer and a polymeric fuel binder. Conventional crystalline oxidizers include ammonium perchlorate (AP), ammonium nitrate (AN) and potassium perchlorate, whereas among the fuels there are hydrocarbon-structured binders, e.g., polybutadiene or polyurethane. Modern propellants consist of several more chemical ingredients, such as metal particles added as fuel, burning rate modifiers, processing aids, anti-aging agent, combustion instability suppressant and curatives. The selection of the ingredients is based on the desired combustion characteristics and mechanical properties. The combustion of such a heterogeneous mixture is understandably complex since the solid components are usually blended in three or four discrete sizes to aid in processing and control burning rate. In developing an analytical model for composite propellant combustion, these complexities are further limited by the incomplete understanding of the various possible reactions present in composite propellants. There is need for mathematical tractability in representing these reactions as well as those that are more understood. This study presents a computerized code, based on the BDP model and the Cohen at el. analytical model of composite propellant combustion. In the first stage, the model results of the burning rate were compared with the Miller and Foster experimental results. The goal was to examine the model accuracy and correct the thermochemical constants to match materials used in the Israeli industries. Once the values for these constant parameters were established, they were not changed during the calculations for burning rate and burning rate exponent. After the determination of the model constants, the model burning rate results were compared with experiments for several propellant compositions produced in Rafael. The model showed good prediction for low coarse/fine particle size ratio compositions but had a large deviation from the experiments for high ratio compositions. Trying to overcome this problem, the model was modified and the improvement was implemented in the basic model. This improvement includes a new approach of calculating the O/F ratio with respect to the AP size fractions in the propellant. This new approach of calculating the O/F ratio affects the characteristic surface dimension that is used in determining the diffusion height and also the  parameter of the BDP model. After the implementation, the prediction for high coarse/fine particle size ratio compositions was improved significantly and the predictions for "fast" (low C/F ratio) propellants were still very good, all within15-20% deviation from the experimental results.

It may be concluded that the current work improved the BDP model with a new approach for various calculations, resulting in a very good burning rate prediction of all reduced-smoke; bi-modal; AB-10; IPDI propellant compositions. Most probably, it can also predict well the burning rate for tri-modal propellants of the same kind; however, this was not tested methodically.