Natural Science Forum / Physics / Relativity / September 2005

So you say.  But according to SR this is *not* true.
Are we simply to assume what you say, or will you give
some arguments in favor of your opinion on this matter?

Don't be a jerk.  Roberts gave you a lot of his time
and does not deserve these ungracious comments, even
granting that you disagree with him.

The whole problem is your contrary-to-fact assumption
that I called attention to above.

Or after.  LET doesn't tell you which, because you don't
know what "real time" is.  The frame of the ether is
unknown -- it *could* be moving faster than the train!

All you really know is that according to the embankment
frame the rear weight is released before the front
weight.  And of course that is the prediction of SR as
well.  We can all sleep easy.

The rotation would be real, if however unaccountable

Why would he assume that?  He doesn't, because the release
is not simultaneous in his frame.

However, one point of clarification:  the balls will *not*
rotate past vertical in this experiment as predicted by
LET or SR.  The angle of the rod is entirely consistent
with its being parallel to the car in the train frame.

When the assembly goes into a spin and the

Ah, I see I anticipated you, above.  The answer to this
is it never happens.

It would seem that he would be required to

Your error was in the second sentence of your post.

Perry, you always have been talking from first principles
through your bottom and you will always continue talking
from first principles through the same bottom:
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/TempForce.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/SimplyPut.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/Counter.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/AccelPerry.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/HowdyDoo.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/LorentzPerry.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/SRValid.html
  http://users.pandora.be/vdmoortel/dirk/Physics/Fumbles/AAB.html

Don't you ever get tired of making a fool of yourself?

Dirk Vdm

Ok up to this point.

Note that the horizontal distance between the masses remain the same,
only their elevation differs, so they can never get vertical. But ok,
since you said "virtually", that is, "Almost but not quite".

This is faulty since the angle of rotation is always under 90 degrees.
Therefore the order does not change. Also, this rotation is not like
the needle of a compass, the rod gets distored.

All observers agree what the elevation of any given part of the rod is,
according to the time the clocks on the platform show. So this is not a
contradiction, just another apparency.

Perhaps redundant to repeat, but there is no spinning like crazy or
such reversal :)

I suggest you do the experiment before further embarrassing yourselves
with this tripe.

When the first mass is released the cord will result in a swing toward
the other mass, which hasn't been released. Propagation delays in the
cord won't change this. It's a simply Newtonian experiment a.shole Dirk.
If you allow the delay to be excessive enough you can provide for the
first mass to actually swing back up and strike the platform on the
other side of the leading mass. Providing the platform extends that far.
What a bunch of morons we have around here. I'd wager than the Newtonian
principles that I invoked can be demonstrated adequately in any 7th
grade science lab.

FOAD.

Richard Perry

Your whole description is overly complicated, and your analysis is
woefully inadequate, in part because of the complexity of your
description. Gravitation is COMPLICATED, so let's eliminate it;
rods/cords/strings are also complicated, so let's omit them, too: Let me
propose a different gedanken which exhibits the essential aspects of
yours, but is MUCH easier to analyze:

This is purely SR: In Minkowski spacetime consider inertial frame A and
inertial frame B moving rapidly along a common X/X' axis relative to A;
the point X'=0 is at rest in B and moves to increasing values of X in
frame A. Their Y and Z axes are parallel (everything happens in the
plane Z=0).

Initially at rest in frame A are two identical rockets, R1 and R2,
separated along X with R1 at X1 and R2 at X2 in frame A, X2>X1; both are
located at Y=Z=0 in frame A, and both carry standard clocks that are
initially synchronized with the (standard) coordinate clocks of frame A.
At time T=0 in A both R1 and R2 blast off along the Y axis with
identical and constant proper acceleration. At any time t>0 in frame A,
both rockets have identical values of Y, and their clocks both display
the same time (which is progressively less than the coordinate time of
frame A at their current location). Both rockets asymptotically approach
speed c relative to frame A, and their direction is always along the Y axis.

Observed from frame B, R1 is the lead/front rocket, and R2 is the
lagging/rear rocket. R2 blasts off at a smaller value of t' than does
R2. So at any coordinate time in B after R2's blastoff R2 will have a
larger value of Y' than R1. This difference in Y' values will increase
over time in B, approaching an asymptotic value (which is finite, and
the value depends on lots of details). Both rockets asymptotically
approach speed c relative to frame B, but their direction changes
(initially along -X', asymptotically approaching Y') -- this is their
direction of motion, not the direction in which they are pointing (which
is always along Y').

Note, however, that the observer in frame B could observe the locations
of the rockets at a given value of their onboard clocks (rather than at
a common value of coordinate time t'); when this is done both rockets
have the same value of Y' and both have the same speed and direction of
motion relative to frame B.

You wanted to attach a rod/cord/string between the rockets. That has
many difficulties, so instead let's line up a series of rockets between
R1 and R2, and have them all blast off at t=0 in A, with the same proper
acceleration as R1 and R2. Think of these rockets as short pieces of
"string", but there's no "sagging" or tension, and the entire "string"
accelerates along with the rockets R1 and R2. Clearly in frame A, at a
given coordinate time t all the rockets (incl R1 and R2) have the same
value of Y, and their onboard clocks all read identical values.

Observed from frame B after R2 blastoff, this row of rockets is not a
straight line, but asymptotically approaches one. It is this complex
shape that makes a simple string extremely difficult to analyze.
Moreover, each of these rockets is moving differently relative to frame
B, in a complicated and changing direction (initally along -X', but
increasingly pointed toward Y', and asymptotically parallel to Y'; the
different rockets approach the asymptote differently as a function of
frame-B time t'). Note, however, that if observer B measures the
locations of all the rockets at a given value of their individual
onboard clocks, they all have the same value of Y' and all have the same
speed and direction relative to frame B.

You want to discuss the "tension" in the "string" between R1 and R2, and
made wildly-incorrect statements about it. This gedanken permits us to
analyze it. You seem to think that because the path length of the
"string" in frame B is clearly longer than a straight line, the observer
in frame B must conclude there is tension in the string. That is not so.
The deviation from a straight line is due to the difference in
coordinate clock synchronization, and that quite clearly does not affect
the string itself.

In any short region of a string, its molecules will arrange themselves
so in ther own rest frame the inter-molecular bonds are as relaxed as
possible. That's why I split the continuous string into a series of
discrete rockets. The relationships between adjacent rockets are only
simple at a common value of their proper times. Clearly this is frame
independent, and the above construction implies that the relationship
between adjacent rockets at common proper time (in their
instantaneously-comoving rest frame) is EXACTLY the same as it was
initially in frame A.

So by comparing this gedanken to yours, we can conclude there is no
tension in your rod/cord/string (assuming there was no tension
initially, which you did not specify but is clearly required). Clearly
the wild rotations you imagined will not occur.

Tom Roberts    tjroberts@lucent.com

This is problematic, because there's no gravity in SR. It's not clear to me
that the weak-field Newtonian approximation will work here, given that
relativistic velocities are involved. You could avoid this by using a
uniform electric field instead of gravity, but the analysis is still tricky
since magnetism will come into play. Still, I understand what you're getting
at, so perhaps these complications can be ignored.

He's right: sound waves are electromagnetic. The solid objects of everyday
experience are electromagnetically bound. Electromagnetism is what
propagates an applied force through the whole object, so that the whole
object appears to move as a unit. Changes in the electromagnetic field
propagate at c, so when you push or pull one end of any solid object, no
matter what its composition, the other end cannot move until at least a time
d/c later, where d is the distance to the other end. In practice the delay
is much longer than this in realistic materials, because they have a certain
amount of elasticity to them, i.e. a willingness to move independently of
the rest of the object, within certain limits. Even materials that we think
of as very rigid are very weakly coupled in the relativistic domain. When
you formulate relativistic thought experiments, you have to think of
everything as being made of Slinkies.

This is the problem with your setup (or rather with the way you're thinking
about it). Connecting two objects with a rigid rod will not ensure that
their mutual separation remains constant. What actually happens will depend
on the details of the experimental setup, but whatever it is, it won't
resemble a spinning baton.

-- Ben