Re: Want to prove E=mc²? University labs should try this!

On Tue, 19 Nov 2024 9:41:32 +0000, ProkaryoticCaspaseHomolog wrote:

>
On Tue, 19 Nov 2024 0:13:44 +0000, rhertz wrote:
>
When I look at the screenshot on the following page, taken during an
actual run, https://tinyurl.com/2v3kssvv I see a circulating power of
358.8 KW versus an input power of 63.5 W. This implies an
amplification factor of 5650. This is not a freely adjustable number,
but one dictated by the physics of the Fabry-Perot mirrors. Look down
at the diagram captioned "Basic Michelson interferometer with Fabry
Perot cavities." The far mirrors will be 99.999% reflective, but the
mirrors near the beam splitter must necessarily be partially
transmissive, otherwise power cannot enter the Fabry Perot cavity.
It's an interesting optimization problem. The more transmissive the
near mirrors, the more light enters the FP cavity, but the fewer the
reflections. The more reflective the mirrors, the more the power
amplification factor, but less light enters the cavities, and the
system's sensitivity to high frequency gravitational waves is also
limited. Once they settled on an optimum transmissivity for the near
mirrors, they could not arbitrarily adjust this parameter.
>

And about the multilayered coating used in the mirrors, which
also evolve with the pace of time, can be perfectly applied to
the interior of the cavity that I proposed, without restrictions
except money.

>
Nope. Physics limits their application.
You can achieve 99.999% reflectivity only at one specific angle
(which is dependent on the mirror design). If the mirror reflects
99.999% of light normal to the surface, it won't reflect 99.999% of
the light at other angles.

Exactly.  I didn't realize how complex the LIGO optical train was,
nor the "power recycling" concept:
https://arxiv.org/pdf/1105.0305
I'm still quite certain, however, that when you throw 750 kW
into a 10 cm ball with walls that are 0.999999 reflective,
the losses will, as you say, cause serious problems.
For one, that's a loss of 0.75 W/bounce, and bounces will happen
c/0.1 = 3x10^9 times per second -- IF one could supply the power
to keep it operating.  In which case, the whole thing would make
a beautiful incendiary display.  With only 5 W input to drive
the system, however, it would heat up to about 300 C, according
to my radiation slide rule.
For another, most of the light bouncing around won't be coming in
at normal incidence, so losses will be greater.
I still think it's a loss cause.  My worst loss cause was I built
a system to spin up a 4-inch diameter aluminum disc, 1 inch thick,
to 27000 rpm (it was scary!).  Some engineer has patented  what he
called a "kinemassic effect."  I thought it might be a gravitational
drag effect from spinning nuclei, and an equation in a paper by
Robert Forward seemed to confirm it, and my experiment should have
revealed it.
It didn't.
Turns out, Forward left out a term in his equation that almost
canceled the L-T effect :-(