Accurate thermophysical properties of high temperature metallic liquids are important for both industrial applications and scientific research. For the former, as predictive numerical simulations play an increasingly important role in pivotal industries, such as casting, welding and sintering, the lack of precise thermophysical properties, especially at high temperatures, hamper their further applications. On the other hand, from the stand point of basic metals physics, being able to measure calorimetric properties of liquid metallic alloys allows the calculation of their thermodynamic functions, which facilitates the study of solidification kinetics of quasicrystals. This research focuses mainly on measuring properties that governs heat and fluid flow, namely specific heat and thermal conductivity. Conventional techniques generally require test materials, whether solid or liquid, in direct contact with a fastener or a container, which limits the application of these methods at high temperatures due to elevated thermal reactivity and also intensified heat and mass transfer. To overcome these drawbacks, a non-contact modulation calorimetry has been developed in the 1990s. Specifically, a spherical sample, usually smaller than 10mm in diameter, is levitated using electromagnetic levitation (EML) in a vacuum or gas cooling atmosphere. Thermophysical properties are extracted by analyzing the temperature response on the surface of the sample induced by modulated heating input. This method has been proved relevant for the measurements of specific heat but is limited in thermal conductivity measurements for liquid samples because of the influence of forced convection. In this study, a two-dimensional axisymmetical model is developed in COMSOL Multiphysics. Using magnetohydrodynamic simulation, firstly, the relevance of the operational principles of this method based on coupled heat flow model developed by Wunderlich, et al., is examined at both reported and unknown parameter range for specific heat measurements. Secondly, systematic error of thermal conductivity measurements using empirical equations in EML based modulation calorimetry is evaluated. Lastly, the possibility of combining numerical simulations and experiments to measure thermal conductivity is explored.