In the casting of single-crystal turbine blades, the composition (c) and temperature (T) dependencies of the liquid-phase molar volume (V(c, T)) play a critical role in driving convective instabilities and the associated formation of solidification defects. To support an effort aimed at the development of validated mathematical criteria for predicting solidification defect formation in Ni-based superalloys, ab-initio molecular dynamics (AIMD) simulations have been performed for elemental, binary and ternary alloys of Ni with Al, W, and Re, to compute equations of state at temperatures of 1,830K and 1,750K. Where comparisons with measurements are available, AIMD-calculated volumes agree to within 0.5-2.5% of experiment. The results are used to test the accuracy of the predictions of a recently proposed parametrization for composition and temperature dependent molar volumes in liquid multicomponent superalloys. For Ni-Re the model is found to be highly accurate over a wide range of compositions while for Ni-W it gives rise to qualitatively incorrect predictions at higher W concentrations.