As stated last post, below is a core part of the latest revision of an email 'flyer' being sent out to various folks. Red highlight emphasizes that which is evidently strictly taboo topic. I wonder why!
****************************************************************************************************************************************************************
This may seem a silly question, but have you ever bothered to check if those elegant GR solutions for GW's, apparently making good sense viewed on a far-field small patch basis, e.g.
https://en.wikipedia.org/wiki/Gravitational_wave#Effects_of_passing
, still make sense on a far-field global view?
I contend a severe consistency problem, obscured in the usual orbiting binary scenario, is acutely evident in the simplest possible axially symmetric system. That of a linear quadrupole radiator. One asks what kind of ostensibly purely transverse, far-field metric deformations can act on an imaginary spherical shell centred about the linear radiator. Here is a link to a bare bones illustration I used in a below linked to forum thread:
https://s26.postimg.org/axee7pdmh/GR_GW_paradox_2.png
At rock bottom, all one need ask is can there logically be an axially symmetric azimuthal 'strain' h_φφ (usual spherical coordinates r, φ, θ) - locally tangent to any given line of latitude e.g. equator line? Given that in far-field, GR GW's are exclusively transverse and pure shear in character?
To claim that a circular hoop can undergo uniform circumferential strain without an accompanying radial expansion/contraction is laughably illogical. Yet the parallel in the linear radiator GW case has somehow never been noticed. And it's easily seen that relaxing the strictly transverse character of GR GW's to allow radial 'breathing', cannot resolve things in favour of transverse strains declining in amplitude in a 1/r radiation field manner. Instead, transverse strains decline as 1/r^2, and one has overwhelmingly radial i.e. longitudinal far-field 1/r displacement amplitudes. Which I do not believe a credible way out, for a number of reasons, including the following.
The situation just becomes worse when polar strains h_θθ acting along meridian lines of that imaginary spherical shell is considered. As you are no doubt aware, at any given local patch and any given instant t, h_φφ, h_θθ strains have equal amplitude but opposite sign ('+' polarizations - applying to axially symmetric linear oscillator). Hence, while say at t = t_0, latitude lines notionally want to 'breathe' outwards, meridian lines notionally want to 'breathe' inwards. And vice versa other half cycle. This is not QM - an initially static test particle can't both move in and out nor be in two places at the same time!
It was always a puzzle as to why one never sees a global view of the h_φφ, h_θθ strains in any of the many online illustrations and animations. Typically just cartoon 'water wave' depictions, and then only for binary case. Strange indeed!
Back in 2012 I first discovered and aired this issue in a forum thread, but that was a bad move ending in a life ban. My particular approach was to think about it in terms of Feynman's famous sticky beads argument e.g.
https://en.wikipedia.org/wiki/Sticky_bead_argument#Description_of_the_thought_experiment
https://en.wikipedia.org/wiki/Sticky_bead_argument#Feynman.27s_argument
As soon as such is applied to the axially symmetric linear quadrupole oscillator case, it becomes evident Feynman got it very wrong. Laughably obvious when one extends it to a large, uniformly distributed circular array of sticks and beads, sticks joined end to end around a far-field r>>λ equator line! The beads will logically move azimuthally somehow?! Given the axial symmetry? He he he.
Moral - go global view - not local patch view!
****************************************************************************************************************************************************************