I was looking into the formal proof of Kepler's laws of planetary motion and I happened to come upon this equation for the eccentricity of the orbit:


For the proof of Kepler's first law I was studying, it seems necessary to prove that e is always $\ge0$ in order to solve this integral: $$\theta=-\int_{ }^{ }\frac{1}{\sqrt{\frac{e^{2}}{r_{0}^{2}}-\left(u-\frac{1}{r_{0}}\right)^{2}}}du $$ Where $u=1/r$ and $r_0 = \frac{L^2}{GMm^2}$.

Since $e$ is always $\ge0$, $\frac{2EL^{2}}{G^{2}M^{2}m^{3}}+1$ must also be $\ge0$. I have tried to prove this by substituting $$E=\frac{1}{2}mv^{2\ }-\frac{GMm}{r}$$ as well as $$E=\frac{1}{2}m\dot{r}^2 + \frac{1}{2}mr^2\dot{\theta}^2-\frac{GMm}{r}$$ but so far I've been unable to prove it. Could anybody help me or at least point me in the general direction?

  • 4
    $\begingroup$ What would an ellipse with negative eccentricity look like? $\endgroup$
    – J...
    May 31 '21 at 10:09
  • $\begingroup$ When I plugged it into the equation in Desmos, it looked pretty much the exact same with $-1<e<0$ being an ellipse, $e=-1$ being a parabola and $e<-1$ showing a hyperbola. $\endgroup$ May 31 '21 at 10:56

By definition, $e$ is a ratio of two lengths. Since these are $\ge0$, so is $e$. Meanwhile, $$1+\frac{2EL^2}{G^2M^2m^2}=1+\frac{2Er_0}{GM}$$ is $\ge0 \iff E\ge-\frac{GM}{2r_0}$, which is true for closed orbits by the virial theorem.

  • 1
    $\begingroup$ Is true also for open orbits as $E>0$ $\endgroup$ May 30 '21 at 21:47

You start with the energy

$$E=\frac m2 \vec v\cdot \vec v-\frac{G\,m\,M}{|\vec R|} $$


$$\vec R=r(\varphi)\,\begin{bmatrix} \cos(\varphi) \\ \sin(\varphi) \\ \end{bmatrix} ~,\quad r(\varphi)=\frac{p}{1+e\cos(\varphi)}~,p=\frac{h^2}{G\,M}$$

$$\vec v=\frac{\partial\vec R}{\partial\varphi}\,\dot\varphi~, \quad \dot{\varphi}=\frac{h}{R^2}~, \quad h=\frac{L}{m}$$

Thus the energy is:

$$E=\frac 12\,{\frac { \left( {e}^{2}-1 \right) {M}^{2}{m}^{3}{G}^{2}}{{L}^{2}} } $$

solving this equation for $e^2$ you obtain


How you obtain $~r(\varphi)$ \begin{align*} &\text{start again with the energy }\\ &E=\frac{m}{2} \left( {{\dot{r}}}^{2}+{r}^{2}{\dot\varphi }^{2} \right) +U \left( r \right)\\ &\text{with}\\ &\dot{\varphi}=\frac{h}{r^2}\qquad,\frac{dr}{dt}=\frac{dr}{d\varphi}\dot{\varphi}\\ &E=\,m{h}^{2} \left( {\frac {{{\left( \frac{dr}{d\varphi}\right)}}^{2}}{{r}^{4}}}+\frac{1}{r^2} \right) +U \left( r \right)\\ &\text{solving this equation for $~\frac{d\varphi}{dr}~$ and with } ~ h^2=\frac{L^2}{m^2} \\ &\frac{d\varphi}{dr}=\frac{1}{r^2}{\frac {1}{\sqrt {2\,{\frac {m \left( E-U \left( r \right) \right) }{{L}^{2}}}-\frac{1}{r^2}}}}\\ &\text{the potential energy $~U(r)=-\frac{G\,m\,M}{r}~$ and for eliptical path $~E=0~$ you obtain}\\ &\frac{d\varphi}{dr}=\frac{1}{r^2}{\frac {1}{\sqrt {2\,{\frac {G\,M{m}^{2}}{r{L}^{2}}}-\frac{1}{r^2}}}}\\ &\Rightarrow\\ &\varphi=2\,\arctan \left( {\frac {\sqrt {2\,G\,M{m}^{2}r-{L}^{2}}}{L}} \right)\\ &\text{or}\\ &r={\frac {{L}^{2}}{G\,M{m}^{2} \left( \cos \left( \varphi \right) +1 \right) }}\\ &\boxed{r(\varphi)=\frac{p}{1+\cos(\varphi)}} \end{align*}

  • 1
    $\begingroup$ Very nice explanation. I have voted up all :-) $\endgroup$
    – Sebastiano
    May 30 '21 at 21:44
  • $\begingroup$ Doesn't this proof require you to already know r as a function of angle? I wanted to prove that $e\ge0$ so that I could then solve the integral and find r in terms of angle. $\endgroup$ May 31 '21 at 1:30
  • $\begingroup$ @ShudheshVelusamy if e > 1 then the energy E > 0 you obtain hyperbolic path, e < 1 elliptic path e=1 parabolic path $\endgroup$
    – Eli
    May 31 '21 at 6:00
  • $\begingroup$ @ShudheshVelusamy I put more information how to obtain elliptical path $\endgroup$
    – Eli
    May 31 '21 at 15:50
  • 1
    $\begingroup$ @Eli Thank you! $\endgroup$ Jun 1 '21 at 8:46

A more trivial proof is by using the eccentricity vector (which is a vector with the same direction as the semi-major axis and whose modulus equals the eccentricity of the conic)

$$\mathbf{e}\equiv\frac{\mathbf{A}}{mk}=\frac{1}{mk}(\mathbf{p}\times \mathbf{L})-\hat{\mathbf{r}}$$ $$|\mathbf{e}|\equiv\frac{|\mathbf{A}|}{mk}\geq0$$



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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