Organic phase change materials have been attracting great attentions for their promising potential in thermal energy storage applications. Due to their poor thermal conductivity and thermal diffusivity, thermally conductive fillers are often added to form composites to enhance the thermal performance. To achieve the optimized performance without sacrificing the thermal capacity, a novel numerical methodology has been developed to model the thermal behavior of phase change material composites, which has been validated by the experimental results for pure n-octadecane and n-octadecane/expanded graphite composites. Effects of different filler concentration distributions have been analyzed and compared. It is found that the phase change time is significantly affected by the filler distribution. An optimal polynomial filler distribution can reduce the phase change time by more than 50% with the same filling content, compared with the uniform distribution. Entropy analysis indicates that a shorter phase change time is correlated with a lower entropy generation rate. (c) 2014 Elsevier Ltd. All rights reserved.