Closed bomb testing is a prominent means of characterizing the combustion behavior of solid gun propellants. This sub-scale test allows the propellant to burn in a constant volume environment, where the resulting pressure-time trace can be collected via a pressure transducer. Historically, numerical procedures have been developed to determine the burn rates of the gun propellants from these pressure-time traces; however, no standardized procedure exists to determine the burn rates of grains with variable surface thermochemistry and ignition criteria. To address this capability gap, a non-linearly constrained, multivariate optimization algorithm has been developed to decouple propellant grain surfaces and determine surface-specific burn rates [1]. In this work, the optimization algorithm as well as the legacy Excel-based Closed Bomb (XLCB) program [2] were used to determine the burn rates of homogeneous, deterred, and layered propellants from experimental data. Closed bomb simulations using these burn rates were then conducted with the two-phase, multidimensional, interior ballistics solver, iBallistix [3]. The maximum mean error between the simulated and experimental pressure-time curves was 6.8 % for the optimization algorithm and 23.8 % for XLCB, showing a marked improvement with our new approach. Furthermore, the approach discussed herein improves burn rate predictions of complex solid gun propellants when compared with legacy closed bomb data reduction analysis programs.