I was thinking about some generalizations of Fermat's Last Theorem, and I'm sure they have been studied before. The first one is looking for rational solutions of $$x^a + y^b = z^c$$ and trying to find a necessary and sufficient condition on the tuple $(a,b,c)$ such that solutions exist, whether finitely many or infinitely many, etc.
The other is the same question for the equation $$ax^n + by^n = cz^n$$ To what degree have these equations (and their natural synthesis) been studied? What progress has been made?
Edit: I have done some further research and learned that the $n=2$ case of the second problem I described is solved, it is Legendre's theorem on the ternary quadratic form. $ax^2 + by^2 =cz^2$ has nonntrivial solutions in the rationals iff $\left( \frac { - b c } { a } \right) = \left( \frac { - a c } { b } \right) = \left( \frac { a b } { c } \right) = 1$. More information here.
Look up the Beal Conjecture for the first equation. Solutions exist whereby x, y and z share a common factor. It is easy to generate one of an infinite number of solutions with this characteristic. Example:
From $2^3 + 3^3 = 35$
We can obtain $70^3 + 105^3 = 35^4$ by multiplying both sides by $35^3$
All terms share a common factor of $35$. The Beal Conjecture posits that no solutions exist without a common factor and to date no counter-example or proof of the conjecture exist.