Re: Circularly polarized plane wave produces mechanical stresses, which cannot be explained by the Maxwell electrodynamics. So the electrodynamics is not complete.

On Sat, 11 Oct 2008, khrapko_ri@xxxxxxxxxxx wrote:

On Oct 8, 12:42 am, "Timo A. Nieminen" <t...@xxxxxxxxxxxxxxxxx> wrote:

What experimental refutation do you think is good enough? IMO, our Bishop
2004 PRL was sufficient, with just the proof-of-principle experiment
therein. Parkin 2007 PRE has the mature, practical for quantitative
application, version. But as far as double angular momentum goes, this is
just a refinement of the earlier one. What can be wrong with the result?
In Parkin 2007, the effect of heating on the viscous drag is measured and
accounted for and the measurement of the amount of circular polarisation
is unambiguous. Our number for the beam power at the focus is not out by a
factor of 2; the power needed to obtain the observed rotation rate,
assuming single angular momentum, agrees to within a few percent with
measurement by transmission through an identical objective/"condenser"
pair and that required for the measured spring constant. I can add that
computational modelling of scattering by vaterites (Loke 2007 JQSRT)
agrees.

You don't think Bishop 2004 PRL is sufficient - do you think the
refinement of it is good enough? I'd also say that Friese 1996 PRA and
Simpson 1997 OL (refs 22 and 17 in your JMO paper) are decisive.

Dear Timo:
I proposed a decisive experiment in 2002 (in English see Measurement
Techniques, 46, No. 4, 317, 2003).
However, now the work of Simpson et al. OL 22, 52 (1997) rather
confirms our result, torque=2 Power/Omega, if one take into account a
transformation of LG beam into HG modes as I show in "Inevitability of
the Electrodynamics' Spin Tensor" (www.sciprin.org
Inevitability_spin_khrapko_mar07.pdf).

If the particles are not rotationally symmetric, or are off-centre from the beam axis, there will be conversion of orbital AM (as well as absorption of orbital AM). Off-centre positioning will result in visible orbital motion, so can be ruled out. Conversion to (you say HG modes, but I'd say to) other LG modes of differing orbital AM will depend on how not-rotationally-symmetric the particle is. Some will occur, but will it be significant? To assume a particular number in order to support your T=2P/omega only shows T=2P/omega is possible (i.e., not definitely eliminated as a possibility), not that it is likely, or even plausible.

[moved]
Why must I believe Friese 1996 PRA?

Because it agrees well with Simpson 1997 OL. It's the same effect - absorption of spin and orbital AM from a beam carrying both, with the spin being varied. That Simpson used LG01, and Friese used LG03, to get the orbital AM doesn't make any real difference.

In this my paper, I show also Bishop at al. PRL 92, 198104 (2004)
yields rather torque=2.4 Power/Omega, although the work of Bishop PRL
2004 does not present enough data.

You're quite right that there isn't enough data in the paper. Figure 2 looks like it's for a 4-5 micron diameter vaterite; this isn't the same one that the viscosity figures in the text are for.

For enough data, see Parkin, Physical Review E 76(4), 041507 (2007). Figures 2 and 3 are for the same vaterite. The particle diameter, laser power at the focus, measured change in circular polarisation, and rotation rate are given. The largest errors would be the vaterite diameter, which could be up to 2-3% (1-2% is usual, but this depends on how close to the same size other vaterite can be found in the sample to do the size measurement), and the beam power at the focus, which should be within 5%, or better.

In my paper, I show also that Parkin et al. Optics Express 14 6963
(2006) and Garces-Chavez et al. PRL 91, 093602 (2003) yield rather
torque=2 Power/Omega.
In my paper, I pay attention that Loudon's PRA68 013806 (2003) also
gives torque=2 Power/Omega.

This my paper was submitted to JOSAA, JOSAB, OL, PRA, PRL, AO, OE, OC,
PLA, AOP, JMO, EJP, EPL, JOPA, JOPB, NJP, JPA, JMP, but I received
only one referee report, which I can send to anyone (see also "
http://groups.google.com/group/sci.physics.electromag/browse_thread/thread/6cb0bb7e6f7bb4ba?hl=en#
)
I sent this paper to Allen, Barnett, Bishop, Dholakia, Garces-Chavez,
Loudon, Nieminen, Padgett, but I had no answer.
In JMO 55 1487 (2008), I show the reasoning of Allen and Padgett in
AJP 70 568 (2002) about absorbing of a circularly polarized beam is
wrong, but I have no reply.
Why must I doubt that I am right?

If you didn't think you were right, you wouldn't think you were right. "Doubt" would be the wrong word. However, especially when it's a controversial issue, one should consider the possibility that one is wrong.

Your spin tensor yields the correct result for monochromatic beams. So far, so good. Whether or not Soper et al. are wrong is another matter. (Doesn't Soper's spin, Humblet's spin, also give the correct result?) The validity of various theoretical procedures or derivations when they give the same observable result is always a question that is difficult to resolve.

P/omega vs 2P/omega, on the other hand, should be amenable to experimental check. Experiments in space (to be free from friction, drag, etc) are not so easy to arrange (and a drop tower might not give enough time). So, for the moment, we are stuck with terrestrial laboratory experiments. I think Friese 1996 PRA (and Simpson 1997 OL) is sufficient, but see what you think about Parkin 1997 PRA above.

There was also a nice older experiment, F. Chute, The reaction torque on an axial multipole radiator, IEEE Transactions on Antennas and Propagation, 15(4), 585-587 (1967). A multipole antenna was suspended, and the reaction torque due its radiation measured. It radiated in an m=1 mode, so it's essentially a crossed-dipole antenna, theoretically radiating P/omega angular momentum (theoretically, though with some disagreement on the issue, this should be half orbital, half spin). Measured torque was 36.5 dyn.cm, predicted was 46.2 dyn.cm.

There's one important theoretical point that agree with P/omega, and not with 2P/omega. If, for example, driving a 1/2-wave plate with a circularly polarised beam, conventional theory gives a torque of 2P/omega, yours 4P/omega. If the waveplate is rotating at terminal speed in a viscous fluid, the beam is doing work on the waveplate, at a power equal to torque.Omega (Omega = rotation angular speed). So, the power of the beam must be reduced accordingly.

Conventional theory gives rotational frequency shift as the source of the power. There were a bunch of papers on this in 1998, but there are many earlier ones too. The earliest thorough treatment of this I know of is
R. d'E. Atkinson, Energy and angular momentum in certain optical problems, Phys Rev 47, 623-627 (1935).

Doing this in the paraxial limit, with the conventional separation between spin and orbital AM gives P/omega as the AM flux for circular polarisation. For spherical multipole fields, mP/omega.

Given that the derivation of rotational frequency shift is independent of any choice of expression for AM density or flux, I think it's a strong check on AM results.

--
Timo Nieminen - Home page: http://www.physics.uq.edu.au/people/nieminen/
E-prints: http://eprint.uq.edu.au/view/person/Nieminen,_Timo_A..html
Shrine to Spirits: http://www.users.bigpond.com/timo_nieminen/spirits.html

Relevant Pages

  • Re: Bell and non-Bell spookiness
    ... indistinguishable given variations in the global environment. ... An assumption being made is that the measurement is random *because* ... Note that a spin zero pair of particles is spin zero before ... There are these beam ...
    (sci.physics)
  • Re: Space Exploitation
    ... > Let me offer a different spin on solar power satellites. ... And here's another spin on SPS: ... great depths of rock. ... A gigawatt-scale high energy particle beam weapon ...
    (sci.space.policy)
  • Re: Where is the mathematical beauty of relativity !!!
    ... observe is that the raw charge carries a magnetic moment inherently. ... stationary charge (though it is tough to make an electron stand still) ... Physics circa Y2K simply tacks on the spin ... needed is an inhomogeneous magnetic field to strip out half the beam. ...
    (sci.physics.relativity)
  • Re: Where is the mathematical beauty of relativity !!!
    ... Then, gradually, the deposit became visible, showing a beam separation ... not a discrete spin. ... structured spacetime which relativity theory takes only part way. ... allow electron spin to be relegated as a generic feature. ...
    (sci.physics.relativity)
  • Experimental determination of classical spin
    ... ---A circularly polarized light beam carries an angular momentum ... central part uniformly and bring the (spin) angular momentum \hbar per ...
    (sci.physics.electromag)