THE IMPLEMENTATION OF PANDAS PROFILING AS A TOOL FOR ANALYZING MECHANICAL PROPERTIES DATA OF NICKEL-BASED SUPERALLOYS BASED ON ALLOY CHEMICAL COMPOSITION

Abstract

The purpose of this study is to evaluate the mechanical properties of nickel-based superalloys with variations in alloy chemical compositions using the Exploratory Data Analysis (EDA) method with the assistance of the pandas profiling library on Google Colab. In this study, data from 312 tensile tests of nickel-based superalloys were used as research samples, with alloy chemical compositions including carbon (C), manganese (Mn), silicon (Si), chromium (Cr), nickel (Ni), molybdenum (Mo), vanadium (V), nitrogen (N), niobium (Nb), cobalt (Co), tungsten (W), aluminum (Al), and titanium (Ti), as well as mechanical properties such as yield strength (YS), tensile strength (TS), and elongation (EL). The methodology used in this study was the EDA method with the assistance of the pandas profiling library on Google Colab, which enables the automatic creation of a dataset report, presenting information on various aspects such as data structure, descriptive statistics, correlation, distribution, and missing values. The results show that yield strength has a fairly high correlation with titanium (0.51), medium correlations with nickel (0.25), vanadium (0.2), and cobalt (0.2). Tensile strength in nickel-based superalloys has a fairly high correlation with yield strength (0.88), carbon (0.49), and cobalt (0.55), and medium correlations with titanium (0.25) and vanadium (0.25). Elongation in nickel-based superalloys has a negative and fairly high correlation with tensile strength (-0.62) and yield strength (-0.58). Some warnings for missing data and zero values in some variables were identified. These results indicate that the pandas profiling library can be used as a tool to analyze the data of mechanical properties of nickel-based superalloys quickly and easily, and provide clear information on data patterns, data structure, and correlation among data.