Computer Generated Forces (CGF) simulations use physical models as the basic components of simulated entities, which contain mathematical representation of combat systems and their interactions with the environment and other simulated entities, therefore it is essential that they should be built using authoritative data to provide realistic physical dynamics. However, the development of such validated models is resource intensive because of the complex data representation and mathematical implementation. With military’s increasing LVC integrated simulation training requirements, the concept of building validated physical models once and allowing different types of simulation systems to use/reuse these precalculated model data as needed is more attractive than ever. 

Validated physical models can be from different domains and disciplines with different format and standards. The mathematical formulas are designed to use the ground-truth data to produce meaningful data values that a simulation system can use to simulate entities. A proof of concept development of Rapid Simulation Model Development (RSMD) Toolkit implemented a physical modeling process by using United States Army Materiel Systems Analysis Activity (AMSAA) validated physical models stored in Physical Knowledge Acquisition Documents (PKADs). The RSMD’s framework entrusted Modular Open Systems Approach (MOSA) compliant technologies by plugging in the PKADs as an authoritative data source, constructed several physical models for the machine gun M-16,  and produced a set of validated physical data which was consumed by the M-16 entity simulated in a well know CGF system named VR-Forces. 

The physical models implemented were Direct Fire Weapon Accuracy, Rate of Fire, and Direct Fire Weapon Characteristics. The VR-Forces M-16 scenario that used the validated data showed a statistically significant enhancement in the gun’s hit and miss calculation while the architecture ensures multiple data sources and simulation engines can be plugged into or removed from the architecture as needed. This paper provides the statistical improvements found in our study and the software architecture design methodologies.