There is no question that
For years it has been hypothesized that Tyrannosaurus and its close relatives (like Daspletosaurus and Albertosaurus) actively crushed and ingested bones as part of their regular diet. Compared to other large theropods, like Allosaurus and Giganotosaurus, the tyrannosaurids had very robust skulls and teeth that appeared to be adapted to crunching bones and not just shearing flesh. Strangely, though, traces of this kind of feeding behavior are rare. Throughout the dinosaur fossil record the few bones that have been found with theropod toothmarks on them contain scrapes and punctures that suggest that such contact with bone was accidental. Direct evidence of large predatory dinosaurs actively biting bone in order to consume it, like traces readily seen in the later mammal fossil record, is all but absent.
That large theropods ingested some bones is a certainty, though. Coprolites (or fossilized dino dung) from large theropods often contain scraps of bone, and these dinosaur probably ingested fragments of ribs, vertebrae, and other relatively small bones while feeding. This was not exploitation of bone as a food resource by itself, as seen among modern spotted hyenas, but a by-product of other feeding habits. This would make even more sense if, as Hone and Rauhut suggest, large theropods preferentially fed upon juvenile dinosaurs.
A documentary scene involving a pack of Allosaurus attacking an adult Diplodocus makes for a compelling restoration, but Hone and Rauhut argue such events were probably rare. It would be difficult, and very dangerous, for even large theropods to take down such a large animal. Instead large theropods probably fed upon sick, old, and young individuals, just like large carnivores today. If this is correct it could explain why juvenile dinosaurs are rare in the fossil record and why they are often found in groups.
There is no doubt that large theropods at least sometimes attacked adult prey animals, but juveniles would have probably made for much easier prey. Likewise, juvenile animals would be small enough that large theropods would not have been able to avoid ingesting at least a few bones while feeding on the smaller animals. Thus the presence of bone in coprolites and the lack of bones bearing evidence of theropod consumption is reconciled.
As Hone and Rauhut note, however, hypotheses about how theropods hunted and consumed prey will by tested by further evidence. It may be that fossils that might help us understand the habits of large theropods were not recorded or destroyed during excavations, and it would be helpful if paleontologists could keep these kinds of questions in mind while in the field or studying old specimens. At present it does not appear that large theropods regularly crushed large bones for consumption, but it would be fantastic if evidence that they did could be found!
For more on this research see Dave Hone's Archosaur Musings blog.