I will present current work on comparing GADGET+Sunrise hydrodynamic simulations with observations of IR-luminous galaxies at z~0.3-3. I will especially focus on the relative roles of stars and AGN in heating the dust, the role of merger stage on the emergent SED, and the most likely initial gas fractions of the merger progenitors. In all cases, I will compare the results of the simulations with earlier, more direct, measurements. I will address the successes and limitations of the current generation of hydrodynamic simulations.