Search All of the Math Forum:

Views expressed in these public forums are not endorsed by NCTM or The Math Forum.

Notice: We are no longer accepting new posts, but the forums will continue to be readable.

Topic: Misner, Thorne and Wheeler, Exercise 8.5 (c)
Replies: 38   Last Post: Apr 13, 2013 11:57 PM

 Messages: [ Previous | Next ]
 Lord Androcles, Zeroth Earl of Medway Posts: 19 Registered: 4/13/13
Re: Misner, Thorne and Wheeler, Exercise 8.5 (c)
Posted: Apr 8, 2013 12:17 PM

"Dirk Van de moortel" wrote in message
news:kjugkr\$ns1\$1@speranza.aioe.org...

Hetware <hattons@speakyeasy.net> wrote:
> This is the geodesic equation under discussion:
>
> d^2(r)/dt^2 = r(dp/dt)^2
>
> d^2(p)/dt^2 = -(2/r)(dp/dt)(dr/dt).
>
> r is radius in polar coordinates, p is the angle, and t is a path
> parameter.
> The authors ask me to "[S]olve the geodesic equation for r(t) and
> p(t), and show that the solution is a uniformly parametrized straight
> line(x===r cos(p) = at+p for some a and b; y===r sin(p) = jt+k for
> some j and k).

Normally we'd write dotted variables, but with quotes it's easier.
So write
r' = dr/dt
r'' = d^2(r)/dt^2
p = dp/dt
p'' = d^2(p)/dt^2
then you have
r'' = r (p')^2 [1]
p'' = -2/r p' r' [2]

Deriving [1] gives
r''' = r' (p')^2 + 2 r p' p''
which with [2] gives
r''' = -3 r' (p')^2
which again with [1] gives
r r''' + 3 r'' r' = 0

So if you find some r(t) that has r''(t) = r'''(t) = 0 for all t,
you're done. Easy. Take
r(t) = A t + B [3]
then
r'(t) = A
r''(t) = 0
r''"(t) = 0
so indeed
r r''' + 3 r'' r' = 0

Now, from [1] and [3] you get
0 = ( A t + B ) (p')^2
so
p' = 0
so
p(t) = C [4]

So you get
r(t) = A t + B
p(t) = C
Check it out with [1] and [2]. Trivial.

So
x(t) = (A t + B ) cos(C)
y(t) = (A t + B ) sin(C)

There's probably another solution, but seems to be the one they're
after.

Dirk Vdm

======================================
Pulled sin and cos out of your arse, Dork?

For A = 0, B = 1, C =(0 to 2pi) you've drawn a circle.
Is that the "probably another solution" you were
looking for?

exp(i.t) = cos(t) + i.sin(t) -- Euler.

> Indeed, writing LT and inverse:
> x' = g ( x - v t ) [1]
> t' = g ( t - v x ) [2]
> x = g ( x' + v t' ) [3]
> t = g ( t' + v x' ) [4] -- A fucking idiot called Dork.

-- This message is brought to you from the keyboard of
Lord Androcles, Zeroth Earl of Medway.
When the fools chicken farmer Wilson and Van de faggot present an argument I
cannot laugh at I'll retire from usenet.