Research done on Smilodon, particularly S. fatalis, is widespread. This is a very good thing for our inquisitive minds. Topics of research have varied from forelimb strength to animal growth based on the teeth to behaviors implied from skeletal positioning or even computer simulation. Unlike with dinosaurs, mammals tend to be connected a lot more in the human psyche to the world we see around us, given that mammals are very much in charge of the world at this point overall, this makes a lot of sense. Being mammals we tend to see mammals as the highest class of animals around us. However, all this does, in the end, is help us to add more and more research to the pile that can be, and many times is, influenced, for better or worse, by the existence of living cousins or descendants of the animal in question. Perhaps that is why there is so much research available on Smilodon; because so many people love cats after all.
The research done on the forelimb strength was conducted by Julie Meachen and Blaire Van Valkenburgh of the University of California Los Angeles, colloquially known as UCLA of course (you can see Dr. Van Valkenburgh's other projects and ongoing research on his website). The paper states that through radiography of humeral bone it was possible to determine that the Smilodon was capable of grappling and subduing large prey due to superior forelimb strength. It goes on to comment on the need to do so in order to protect those fragile saber length canines as previously discussed here.
The research done on the canines themselves by Robert Feranec of UCBerkeley (people in California love extinct mammals- probably something to do with La Brea started the craze!) discusses a large number of things including diet, growth, and the development of the canines. Isotopic evidence was used to look inside the canine and determine how it grew and developed. Total animal growth was also determined, or inferred at least, by looking at the growth of the canines; the animal does have to grow to hold those teeth after all and any inferences made from there are most likely fairly reliable. The diet question is answered through the same analysis but is not involved in naming specific animals, only making reference to animals with specific isotopic values in their systems which would require a second set of data to name the animals and extensive testing of other extinct animals which were contemporary with Smilodon.
The final paper for today was authored by an Australian group headed by Colin McHenry of the University of Newcastle's departments of engineering and life sciences. The paper uses S. fatalis, like the other two papers and the researchers constructed a three dimensional computer simulation around the markers of behavior so far found in the cat. From my experience raising Smilodon in zoos (see Zoo Tycoon 2) I know, granted my simulation was probably not as rigorous as the Australian study by a great margin, that Smilodon are quite social animals with some odd habits (such as loving being washed). All joking aside, however, the simulation which was constructed was very thorough and used all sorts of different points of knowledge available to the gentlemen and ladies conducting the simulation to model, as accurately as was possible, the behaviors of Smilodon. Personally, I think they have done a fantastic job and have probably come quite close to the actual lifestyle of Smilodon, but only time travel could tell, so we may well never know.