For a while now, I have been thinking about designing things such as small bedside tables, game/dvd/bluray racks for 3d printing. I've always thought that making them modular would be a good way to go about doing this as well.
Modular design would help to create an end result that is vastly larger than the print volume of my 3d printer. I might even be able to recycle models for use in other projects. However, I'm not sure of what I need to think about if I decide to go ahead with these ideas I have floating around in my head.
I'm assuming that certain joints (dovetail, etc), tolerances for different types of plastic due to shrinkage, and print settings (% infill, in particular) would be important to have thought about and evaluated to some extent, but I'm not sure about what else I might be missing.
So my question is to anyone who has designed anything to be modularly printed. Have you really had to think carefully about the engineering side of the print? Or am I simply overthinking this? Should I just design what I want and give it reasonable infill, walls and whatnot, and just go for a trial and error approach? I'm sure there is a method to this madness, but is a concrete understanding of this type of engineering absolutely paramount when it comes to this sort of stuff?
EDIT: Although I've marked darth pixel's answer as accepted, I'm still going to follow JKEngineer's advise and check out that book as well since I feel as though proper engineering techniques alongside a good mentality towards how I would tackle the problem (as outlined in darth pixel's answer) would prove to yield better results in the long run.