(I think my question will be somewhat related to this one: Deriving gravitational potential energy using vectors .)

I know the change in the potential energy associated with a conservative force, $\vec{F}$, moving along any path between A and B is given by: $$\Delta U = - \int_{\vec{r_A}} ^{\vec{r_B}} \vec{F} \cdot d\vec{r}$$

So when we're thinking about the force of gravity, $$\vec{F_G} = - \frac{GMm}{r^2}\hat{r}$$ where I've chosen $\vec{F_G}$ to be the graviational force exerted on the mass, $m$ by $M$ and $\vec{r}$ to be the position of the COM of $m$ relative to $M$'s COM.

So I get: $$\Delta U_G = - \int_{\vec{r_A}} ^{\vec{r_B}} \left(- \frac{GMm}{r^2}\hat{r}\right) \cdot d\vec{r} = \int_{\vec{r_A}} ^{\vec{r_B}} \left(\frac{GMm}{r^2}\hat{r}\right) \cdot d\vec{r}$$

Now, how would I go about evaluating this line integral to get: $$\Delta U_G = \left(-\frac{GMm}{r_B}\right) - \left(-\frac{GMm}{r_A}\right)$$

  • $\begingroup$ Just want to point out that your expression for gravitational force only works if $M$ is a point mass or a spherical distribution. $\endgroup$ Commented May 16 at 13:50

1 Answer 1


We have the dot product: $$\hat r\cdot d\vec r=|\hat r|\cdot|d\vec r|=dr$$ Because the unit vector $\hat r$ and differential displacement vector $d\vec r$ are in the same direction. Note that the angle between $\hat r$ and $d\vec r$ is not necessarily zero, especially if the path is curved.

Change in the gravitational potential energy: $$\begin{align*}\Delta U_G=\int_{\vec{r_A}} ^{\vec{r_B}} \left(\frac{GMm}{r^2}\hat{r}\right) \cdot d\vec{r} & = \int_{\vec{r_A}} ^{\vec{r_B}} \frac{GMm}{r^2}\hat{r} \cdot d\vec{r}\\&=\int_{\vec{r_A}} ^{\vec{r_B}} \frac{GMm}{r^2}\ dr=-\dfrac{GMm}{r}\bigg|_{\vec{r_A}} ^{\vec{r_B}}\\&=-\dfrac{GMm}{|\vec r|}\bigg|_{\vec{r_A}} ^{\vec{r_B}}\\&=\left(-\frac{GMm}{|\vec {r_B}|}\right) - \left(-\frac{GMm}{|\vec {r_A}|}\right)\\\Delta U_G&=\left(-\frac{GMm}{r_B}\right) - \left(-\frac{GMm}{r_A}\right)=\left(\frac{GMm}{r_A}-\frac{GMm}{r_B}\right) \end{align*}$$ We can also use the concept of potential function and gradient theorem as follows: $$\vec{F_G} = - \frac{GMm}{r^2}\hat{r}=-\nabla\Phi(\vec r)=-\dfrac{\partial\Phi(\vec r)}{\partial r}\hat r\implies\Phi(\vec r)=-\dfrac{GMm}{r}$$ $$\Delta U_G = \int_{\vec{r_A}} ^{\vec{r_B}} \nabla\Phi(\vec r)\cdot d\vec{r} = \Phi(\vec{r_B})-\Phi(\vec{r_A})=\left(\frac{GMm}{r_A}-\frac{GMm}{r_B}\right)$$

I hope it helps!

  • 1
    $\begingroup$ Thanks. A little confused by how $\hat{r}$ and $d\vec{r}$ are in the same direction, but the angle between them isn't necessarily zero? $\endgroup$ Commented May 16 at 13:10
  • $\begingroup$ Yes, in general. The angle is not necessarily zero for a curved path. However, the infinitesimal change in the position vector is in the same direction when deriving gravitational potential energy. $\endgroup$ Commented May 16 at 13:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.