Ni-Mn-Ga magnetic shape memory alloys (MSMA) are well-known smart materials for actuation applications, due to their large magnetic field-induced shape change of up to 10%. The production of larger amounts of single-crystalline material from these alloys with reproducible and homogeneous properties is demanding and calls for optimization of the corresponding crystal growth process. In order to support this optimization, sensitive process parameters are varied in simulations and their effects are studied. Here, we report on thermal field simulations in a Bridgmasn-Stockbarger furnace. The lab furnace is equipped with liquid metal cooling (LMC) to achieve high and homogeneous thermal gradients at the crystallization front during crystal growth of cylindrical Ni-Mn-Ga-rods. The calibration of the thermal simulation model requires (i) the knowledge/measurement of the relevant thermophysical properties of the Ni-Mn-Ga alloy as functions of temperature and (ii) thermal data from a reference benchmark experiment in the lab furnace using the same alloy. The calibrated simulation model is used for the simulation of a specific virtual Bridgman-experiment and for the determination of the temperature distributions. Moreover, the influence of the type of liquid metal coolant on the simulation results is investigated. (C) 2011 Elsevier Ltd. All rights reserved.