I know about T and S duality, infact I've developed a way of solving the constraints on orbifold fluxes in general for an orbifold with S and T duality, and I'm sure I read about the 'simple definition' of U duality in Polchinski but I can't seem to find it. I know it's a combination of S and T duality and if memory serves it's a way of enforcing them in different directions but I'm not sure.

Does anyone (ie Ben) knows the simple definition or can find it? I think the orbifold I'm working on is sufficiently simplistic that I've solved the U duality constraints too but I want to be sure before I mention it in a paper.

Ok, now that that's over.

All I know is that U duality is just T + S duality. I've never actually done any calculations with it.

But that doesn't seem to be different from doing just S and T together? I'm sure I read that it involves S duality in particular directions, which is normally an all or nothing thing, but I can't seem to find it. :(