CORROSION RATE PREDICTION OF HIGH ENTROPY ALLOY ALCOCRFENI IN NACL SOLUTION USING MACHINE LEARNING

• Main Author: Yulianto, Dwi
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/66035
• Institution: Institut Teknologi Bandung

High entropy alloy (HEA) as a new alloy with four core effects has the potential to be applied in corrosive environments such as salt media. The development of HEA related to its corrosion resistance properties continues to be developed, but experimental studies through corrosion testing require a large amount of time and money. Methods to accelerate the development of HEA which has good corrosion resistance as an initial reference before conducting experiments need to be carried out. For this reason, machine learning is used in this research. The algorithm models applied are Random Forest (RF), Decision Tree (DT), and Gradient Boosting (GB). The results obtained indicate that the random forest algorithm as the most optimal prediction model based on the accuracy value and the results of prediction validation. This study also obtained the results of the prediction of the corrosion rate of HEA AlCoCrFeNiMn, AlCrFeNiMnTi, AlCo0,1CrFeNi, AlCo0,5CrFeNi, AlCoCr0,1FeNi, and AlCoCr0,5FeNi with the three algorithm models.