Thats kind of what I said as assuming simply supported will be to lenient, and in some cases fully fixed will never be meet. the other issue is whether you assume a central point load or two point loads at the edge of the carriage and use the reduced deflection at these points as the actual deflection (ignoring max deflection in the middle). Its all a case of how far do you go?

I'm pretty sure that there wont be much difference in the deflection answers and the overall performance has more to do with how they are used. As such I don't believe they should be used as the structural part of the axis as any flex in the side supports will be transfered to the rails and exaggerate the defection, add to this the vibration at resonance and the bar will soon resemble a guitar string:naughty:

There are plenty of deflection calculators already out there but unless you know what all the parts mean and can rearrange the units to be all the same the answers can vary. So I'm going to model some of the supports and typical axis lengths and put them in a FEA program to test my answers.