I have a question regarding Misner, Charles W.; Thorne, Kip S.; Wheeler, John Archibald (1973), Gravitation ISBN 978-0-7167-0344-0. It is a book about Einstein's theory of gravitation.

In page 166 of chapter 6.2 about Hyperbolic Motion, the authors present a person feeling constant acceleration $g$ along the direction $x^1$. The authors get the following equations:

$$a^0 = \frac{du^0}{d \tau} = gu^1$$ $$a^1 = \frac{du^1}{d \tau} = gu^0$$

And the system is solved by the authors to get:

$$t=g^{-1} \sinh{g\tau}$$ $$x=g^{-1} \cosh{g\tau}$$

So obviously $u^0 = t$ and $u^1 = x$, but I am not sure about the change of variable between $t$ and $\tau$ to make the appearance of the Lorentz factor.

How do the authors solve the differential equation?

  • 1
    $\begingroup$ $u$ is the four velocity. $u^0 = dx^0/d\tau$, i.e. $dt/d\tau$, and $u^1 = dx^1/d\tau$ $\endgroup$ Jun 26 '19 at 16:13
  • $\begingroup$ Take another derivative to get uncoupled second-order ODEs for $u^0$ and $u^1$. They look like harmonic oscillators but with opposite sign. Solve these, then integrate to get $x^0$ and $x^1$. $\endgroup$
    – G. Smith
    Jun 26 '19 at 16:19
  • $\begingroup$ equation 1 ${\frac {d}{d\tau}}t \left( \tau \right) =gx \left( \tau \right) $ equation 2 ${\frac {d}{d\tau}}x \left( \tau \right) -gt \left( \tau \right)$ take $\frac{d}{dt}$ of equation (1) with equation (2) you get: ${\frac {d^{2}}{d{\tau}^{2}}}t \left( \tau \right) -{g}^{2}t \left( \tau \right) $ The solution is $t(\tau)=\frac{1}{g}\,\sinh(g\,\tau)$ $\endgroup$
    – Eli
    Jun 26 '19 at 17:31
  • $\begingroup$ @Eli No, see John Rennie’s comment. Also, your solution doesn’t satisfy your equation. $\endgroup$
    – G. Smith
    Jun 26 '19 at 19:33
  • $\begingroup$ @G. Smith I see: equation (1) is correct but equation (2) must be equal to zero. The differential equation is : ${\frac {d^{2}}{d{\tau}^{2}}}t \left( \tau \right) -{g}^{2}t \left( \tau \right) =0$ and the solution is what i wrote bevor? check it please. $\endgroup$
    – Eli
    Jun 26 '19 at 20:20


$${\frac {d}{d\tau}}u_{{0}} \left( \tau \right) -gu_{{1}} \left( \tau \right) =0\tag 1 $$

$${\frac {d}{d\tau}}u_{{1}} \left( \tau \right) -gu_{{0}} \left( \tau \right) =0\tag 2 $$

and the constraint condition that

$$dsq=\left( {\frac {d}{d\tau}}u_{{0}} \left( \tau \right) \right) ^{2}- \left( {\frac {d}{d\tau}}u_{{1}} \left( \tau \right) \right) ^{2}= \epsilon\tag 3 $$

where $\epsilon=0$ or $1$

with $\frac{d}{d\tau}eq(1)$ and equation (2) we obtain:

$${\frac {d^{2}}{d{\tau}^{2}}}u_{{0}} \left( \tau \right) -{g}^{2}u_{{0} } \left( \tau \right) =0 \tag 4$$ $\Rightarrow$ $$u_0(\tau)=1/2\,{\frac { \left( gA+B \right) {{\rm e}^{g\tau}}}{g}}+1/2\,{\frac { \left( -B+gA \right) {{\rm e}^{-g\tau}}}{g}}\tag 5 $$ where $A=u_0(0)$ and $B=D(u_0)(0)$ are arbitrary initial conditions

with equation (3) and (2) we get for dsq

$$dsq(\tau)=\left( {\frac {d}{d\tau}}u_{{0}} \left( \tau \right) \right) ^{2}- (g\,u_0(\tau))^2= \epsilon $$

thus: for $dsq(0)=\epsilon$ we can obtain the initial condition $B=B(A)$ and get:

$$B=\sqrt {{g}^{2}{A}^{2}+\epsilon}$$

with $A=0$ and $\epsilon=1$ we get:

$$u_0(\tau)=1/2\,{\frac {{{\rm e}^{g\tau}}}{g}}-1/2\,{\frac {{{\rm e}^{-g\tau}}}{g }}=\frac{1}{g}\sinh(g\tau) $$

$$u_1(\tau)={\frac {1/2\,{{\rm e}^{g\tau}}+1/2\,{{\rm e}^{-g\tau}}}{g}}=\frac{1}{g}\cosh(g\tau) $$ and for $A=1 \,,\epsilon=0$ we get

$$u_0(\tau)=e^{g\,\tau}$$ $$u_1(\tau)=e^{g\,\tau}$$

  • $\begingroup$ How did you get the constraint condition? And why should we pick A=0 and epsilon = 1? $\endgroup$ Jun 30 '19 at 16:04
  • $\begingroup$ @PackSciences This is because the 4 velocities $u_\mu\,u^\mu=c^2=1=\epsilon$ , and for example with $A=0$ you can fulfil this requirement. $\endgroup$
    – Eli
    Jul 1 '19 at 6:58
  • $\begingroup$ Why is epsilon = 0 possible? Why the condition A=0 imposed by the question? What in the question tells A=0 or epsilon=1? $\endgroup$ Jul 1 '19 at 16:13
  • $\begingroup$ @PackSciences for particle with mass $(\frac{ds}{d\tau})^2=c^2=\epsilon=1$ for a massless particle (light) $\epsilon=0$ so you get the light cone this mean that $u_0(\tau)=u_1(\tau)$ or if you want $t=x$. $\endgroup$
    – Eli
    Jul 1 '19 at 16:30

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.