Ben the words 'can' and 'worms' instantly come to mind when I read your request regarding the Hornby Class 50 coupling problem.
There has been so much written in forums and readers letters in the various mags I reckon a book could be produced now. The early Class 50 was a nightmare with regard to the performance of its coupling mechanism and to make matters even worse, Hornby elected to fit the same mechanism to it's Class 31's and 60's. In fairness to Hornby, the later models are not so susceptible to derailing whatever is unlucky enough to be on the drawbar as they have fitted a weaker spring to the close coupling mechanism but believe me, it still happens occasionally. I don't want to get drawn into all the technicalities but suffice to say that the problem orginates from the fact the the couplings have been fitted to the body and not the bogies. This means that when the loco is on a curve the coupling is not in line with the centre line of the track but overhangs the track either into the cess or the sixfoot. The theory being than the 'load' of the train would pull the coupling to the desired position and into acceptible alignment with the coupling of the following vehicle. On the early Hornby Class 50 this did not happen and as a result it did have a disposition for derailing stock when going into a curve. This was because original spring in the coupling mechanism was too strong and thus prevented the coupling from being pulled sideways. As I've said, the weaker spring now fitted has improved this 95% but the fact that the coupling is fitted to the body and not the bogie has produced other problems. You will be aware that the body extremities of any rail vehicle will over hang the track when that vehicle is on a curve. The amount of overhang will depend on all sorts of things not least the curvature of the track, the length of the body and the the distance from the a bogie king post to the extremity of the body. When the coupling is fitted to the bogie (as the majority of manufacturers do) this body overhang is not really an issue because the centre line of the bogie, being of relatively short wheelbase, will always more or less follow the centre line of the track, so it follows that if the coupling is mounted to the bogie, it too will always more or less follow the centre line of the track. If the coupling is fitted to the body however the coupling position relative to the track will be determined by the body overhang. This means that when a Hornby loco (50, 31 or 60) is sat on a curve and not coupled to a train the coupling will not be anywhere near the centre line of the track and it will be impossible to couple any of them to a train. In fact, it's difficult to couple ony of these locos to a train unless at least the loco and first vehicle is effectively sat on straight track! (without the helping finger from the sky that is).