On Apr 25, 4:17=A0pm, Murat Ozer <
murat.h.o...@gmail.com> wrote:
> [[Mod. note -- I have rewrapped the lines. =A0-- jt]]
>
> Hello,
> I would like to consult you to resolve an issue in a paper by
> L.I.Schiff (On Experimental Tests of the General Theory of Relativity,
> Am.J.Phys., 28(1960)340 ). Section II on p.340 reads ? Two identically
> constructed clocks are placed at rest, a distance h apart along the
> lines of force in a uniform or nearly uniform gravitational field
> of acceleration g, as in Fig. =A01(a) {In this figure, there is a
> round gravitating body =A0and above it are two clocks labeled A
> and B separated by a vertical distance =A0h. Clock A is at the top.}
>
> In accordance with the equivalence =A0principle, any comparison of
> the periods of these clocks can be made as well in a gravitation-free
> region, in which they are accelerated upward =A0with the acceleration
> g, as in Fig. 1(b).{In this figure, there are the two clocks A and
> B separated by the height h and there is a third clock C placed
> next to the clock A. The line of force of g is drawn upward under
> the clock B.} We accomplish this by comparing (continued on p.341)
> clocks A and B in turn with a third identically constructed clock
> C, which is permanently at rest, as they sweep by in near coincidence.
> We assume that the fact that A and B are undergoing acceleration
> as they pass C does not affect this comparison. Since C is at rest
> in a gravitation-free region, it is part of an inertial coordinate
> system, and makes a suitable standard for comparing A and B with
> each other. =A0Suppose that clock A has upward speed v_A when it
> passes C. Then if the period of C is T, the period of A that is
> seen by an observer on C is, according to special relativity,
>
> T_A =3DT(1 ? v_A^2/c^2)^(-1/2) ~ T(1 + v_A^2/2c^2), =A0 =A0 =A0(1)
>
> where the approximation assumes that the speed of light c is much
> greater than v_A. =A0Similarly, when clock B passes C with speed v_B
> a second comparison shows that the period of B observed by C is
>
> T_B =3D T(1 ? v_B^2/c^2) ~ T(1 + v_B^2/2c^2). =A0 =A0 =A0 =A0 =A0 =A0 =A0=
=A0 (2)
>
> On eliminating T between Eqs. (1) and (2), we find to the same
> approximation that
>
> T_B ~ T_A[1 + (v_B^2 ? v_A^2)/2c^2] =3D T_A(1 + gh/c^2), =A0 =A0 =A0(3)
>
> since v_B^2 =3D v_A^2 + 2gh. Thus the inertial observer on C can
> inform both A and B that the period of clock B exceeds that of clock
> A by the fractional amount (gh/c^2).?
>
> My question is this: Isn?t something wrong with the above discussion
> because it is known from the gravitational time dilation that between
> the two clocks, the period of the one that is at the higher
> gravitational potential exceeds the period of the other that is at
> the lower potential ? =A0Thus, it must be that the period of clock A
> exceeds that of clock B; not the other way around as stated by
> Schiff. If you agree with me, I will next point out what is wrong
> with Eqs. (1) and (2). Please state your opinions.
>
> Thanks.
As state (correctly) B exceeds A by gh/c*c for the gravitational case.
More elevated clocks appear to run more slowly than lower clocks just
as a planet at higher ellipsis appears to move more slowly than the
same body at closest approach.
The discussion as given is correct. If T is taken to be the small
interval dt. If it is taken to be a long interval such as a journey or
orbital period then integration is necessary and the formulas of the
discussion would not apply so well as higher orbits have longer
periods.
The relations between a malleable space time, velocity and momentum
are complex. For clock A of ABC the dv_A must be less than dV_b for
the equivalence to the gravitational case to be fully valid. Or, if
the distance h is split between A-C and B-C so h is constant and C is
in the middle then the differing potential represented by the greater
elevation of A would show up as a Doppler correction.
I discuss all these different cases because it is not easy to tell
where your understanding may have gone awry. Mr. Schiff's presentation
seems consistent with the principles and formulas of relativity,
including a good approximation of equivalence principle which is all
the formula's claim.
AAG