Enhanced ML Database to Develop Low Flammability Polymer Matrix Composites

Polaron Technologies Inc., DBA Polaron Analytics incorporated in September 2017 to serve government and private organizations in data analytics and computational modeling & simulations areas. Our core competency is using data analytics to develop novel materials and processes and our relevant experience includes building a polymer descriptor database and machine learning (ML) methods to design novel polymers with tailored Cp.Ā On this SBIR, Polaron will leverage our unique understanding of the NavyÆs mission need and polymer material characteristics, combined with our proven ML methods and expanded literature data to optimize numerous variables and their relationship with composite flammability properties to predict the flammability index from the chemical structure of neat resin. We will incorporate the entire process, including the ML descriptor database, into our commercial data analytics platform MatVerse, specifically into a new module called POLYCOMPRED, for discovering polymer composite materials by predicting ASTM E1354 cone calorimetry results for maximum heat release rate, time to ignition, and smoke density, and ultimately identify composite materials with low flammability that meet the NavyÆs performance needs for modulus, strength, thermal stability.