On 2/26/2012 7:07 AM, clipka wrote:
> Am 25.02.2012 12:46, schrieb Warp:
>>
>> The rod will start falling.
>
> Why should it? Downward motion won't put the endless rod into a state of
> lower potential energy, so it will not happen.

What you say is of course true, but there are so many issues raised with 
portals and potential energy (or, in general, the interaction of portals 
with fields) that I'm inclined to throw up my hands and just assume that 
things somehow work like you'd expect in the "Newtonian physics" sense 
and just go from there.


> To the contrary, every coaxial acceleration of the rod would impose an
> ever so slight length contraction due to relativistic effects, leading
> to buildup of internal stress (and hence internal energy - not sure how
> the expert would call this type), while deceleration would reduce the
> internal stress, so deceleration is likely to happen spontaneously while
> acceleration would require external energy input (and I mean energy
> input, not just some force). So even if the rod was falling in the first
> place, given sufficient time its motion will actually /stop/.
>
> (You /could/ force it into motion by heating it up though: As the
> material would try to expand in all directions, again internal stress
> would be induced, and as any increase in speed - whether up- or
> downwards - would reduce this stress due to length contraction, a tiny
> push /would/ eventually get it up to relativistic speed - provided the
> rod doesn't melt long before due to air friction. Or at least that's how
> the thought experiment goes - maybe there's a flaw in it as well.)

There's one bit of your reasoning that I don't follow.  If you put 
yourself in the rod's reference frame, then the velocity of the rod does 
nothing to change the internal stresses, so whence the acceleration?  It 
seems like you're calculating the elastic energy of the rod in purely in 
rest coordinates without modeling how elastic energy behaves in 
relativistic settings.  I think this is probably most clearly shown by 
that fact that your argument doesn't obey conservation of momentum. 
Note, however, that I'm pretty weak on relativistic physics and I 
haven't calculated any stress-energy tensors or anything for this, so I 
might be wrong.

In addition, the argument you've stated would seem to apply equally well 
to inducing a rotation in a ring.  It's possible there's some subtle 
reason why a ring is different than a linear rod here, but I suspect 
it's an indication that the original analysis doesn't really work.

Post a reply to this message