In the pursuit of optimizing aluminium reductionAluminium reduction processes, the integration of advanced computational fluid dynamics (CFDComputational Fluid Dynamics (CFD)) models has become pivotal. One such model, employing magnetohydrodynamics (MHD), is crucial for enhancing efficiency and reducing energy consumptionEnergy consumption. However, the validation of these numerical simulationsNumerical simulation against real-world data remains a critical challenge. This paper presents a novel approach to validate MHD model by leveraging 3D scanning3D Scanning technology to analyse consumed anodeAnode butts and data mining tools to implicitly investigate metalMetals heave oscillations throughout anodeAnode life cycleCycle. Through rigorous analysis and comparison, this study aims to assess the accuracy and reliability of the MHD model in simulating the complex phenomena inherent in aluminium reductionAluminium reduction processes. By bridging the gap between computational simulationsSimulation and real-world observations (continuous cell voltage dropVoltage drop monitoringOnline monitoring and 3D scanning3D Scanning of the consumed anodeAnode butts), this study contributes to the continual evolution and innovation of aluminium reduction technologyAluminium reduction technology.