The mere concept of a line integral is defined for a vector field, and I thus thought the following was a rigorous and general definition of potential energy:
Definition: Given a conservative force field $f:U\to\mathbb{R}^n$, where $U$ is an open subset of $\mathbb{R}^n$, we define the potential energy associated to $f$ as the function $V:U\to\mathbb{R}$ for which $$\int_Cf = V(x_A)-V(x_B)$$ for any piecewise curve $C\subseteq U$ that starts at $x_A$ and ends at $x_B$. Alternatively, our map $V$ is that for which $$f = -\nabla V.$$ Observe that $V$ is defined up to adding a constant.
In the above definition, potential energy is a property of a force field. However, I don't think this always needs to be the case. In the case of two point-masses $m$ and $M$, for example, one may talk about the work done in moving the mass $m$ through a certain curve even when there does not seem to be a single force field involved (if anything, at any time $t$ there is a force field $f_t$ created by the point-mass $M$) and one may even show -although I'm confused by the proof- that
$$W=\frac{GMm}{r(t_2)}-\frac{GMm}{r(t_1)}$$
is the work done on the point-mass $m$ due to gravity as $m$ moves from $\vec{r}(t_1)$ to $\vec{r}(t_2)$, and where $r(t_i)$ is the distance between $m$ and $M$ at time $t_i$.
- What is a general, rigorous definition of potential energy? If we define it by means of the formula
$$W=\int_C\vec{F}\cdot d\vec{r}=U(x_A)-U(x_B)$$
where $C$ is a curve that starts at $x_A$ and ends at $x_B$, then what is the vector field involved in the integral above? How does such vector field relate with the function $\vec{F}$?
- In the case of two (or more) point-masses, what is the vector field involved?