And are you sure that the picture from their runs is an instantaneous contour plots or rather an average one?
That density corresponds to which temperature? Is that the same temperature you used at inlet and outlets? I hope 304 K.
What about now you mesh resolution: you require a good near wall mesh with at least 10 cells in the boundary layers. Yplus of one is a must here. I will then use SST model as turbulence model.