luinthoron said:
I am looking for simple examples showing that the equivalence principle implies or at least suggests spacetime curvature
This doesn't make sense, since the point of the equivalence principle is that, whether a spacetime is globally curved or not, locally, i.e., in a small enough patch of spacetime, you can treat it as flat and ignore the effects of curvature.
luinthoron said:
Every book that I found talks about light going up a gravity well, the Pound-Rebka experiment.
This experiment was not testing the equivalence principle. It was testing gravitational time dilation. There is a relationship between the two, but I don't think it's what you think it is.
Einstein's original argument for why there should be gravitational time dilation was indeed based on the equivalence principle: his argument was basically that we should expect time dilation between the top and bottom of an accelerating "elevator" (nowadays we would probably say "rocket ship") in flat spacetime, and therefore, by the equivalence principle, we should also expect time dilation between the top and bottom of a similar "elevator" (or rocket ship) sitting at rest in a gravitational field. But that argument also means that the presence of that time dilation, in and of itself,
does not tell you that spacetime is curved. It can't, because that time dilation is also present in flat spacetime. This is an example of why you
can't use the equivalence principle to detect the presence of spacetime curvature.
luinthoron said:
I would like a simple dynamic example that would show that equivalence principle implies that the spacetime interval (which my students know from Special Relativity) now contains nonvanishing nonconstant coefficients like on a curved surface.
If you are trying to teach your students about curved spacetime, you will
have to teach them about the difference between "the spacetime interval" (an invariant) and "the components of the metric tensor" (which depend on your choice of coordinates). The two are not the same, but you are treating them as the same. I don't think that approach will work.
In the example discussed above, about time dilation in an accelerating rocket ship in flat spacetime, you can set up coordinates in which the rocket ship is at rest, and an observer at rest inside the ship feels acceleration. (The technical term for this is "proper acceleration".) These coordinates are called "Rindler coordinates", and the metric tensor in these coordinates has a "nonvanishing nonconstant coefficent", the metric coefficient ##g_{00}##, which now depends on "height" (the spatial coordinate in the direction of the acceleration). But
these coordinates are still describing flat spacetime. The geometry of spacetime, which includes all "spacetime intervals", is the same. So this is an example of why you cannot use the presence of "nonvanishing nonconstant coefficients" in the metric tensor to tell that spacetime is curved: you can have the same thing in flat spacetime.
The correct way to think of spacetime curvature is as
tidal gravity. (Note that this is how the Insights article that
@Dale wrote and referred you to talks about it.) The equivalence principle, by definition, cannot tell you anything about tidal gravity, since by definition it only applies to a patch of spacetime that is small enough that no tidal gravity is observable.
For "dynamic examples" of tidal gravity in action, which should be accessible to your students if they understand how Newtonian gravity works, consider these two scenarios:
(1) Two apples, both momentarily at rest relative to each other and Earth, along the same radial line but at different altitudes. Both apples will start falling towards Earth, but the lower apple will have a greater acceleration towards Earth, so the two apples will gradually diverge.
(2) Two apples, both momentarily at rest relative to each other Earth, at the same altitude but separated tangentially. Both apples will have the same magnitude of acceleration towards Earth, but in different directions (since both are accelerating towards the center of the Earth), so the two apples will gradually converge.
Note that neither of these scenarios can be analyzed using the equivalence principle, because any region of spacetime large enough to cover both apples and include their divergence or convergence will be too large for the equivalence principle to apply.