[Date Prev][Date Next][Thread Prev][Thread Next]   [Date Index] [Thread Index] [Author Index

Re: Rosennean challenge? Brun's



Tim,
 
Thank you for posting that little paraphrase (Brun's Theorem via Whittaker) because it's the kind of thing I don't tend to seek out and in this case it's very interesting. It's interesting to me because having the Rosennean perspective but very limited formal education into "Analytical Dynamics of Particles and Rigid Bodies" (etc)... there are several aspects to this whole situation which leap out at me. The main ones have to do with the exactness of the numbers of "integrals" supposedly needed for a complete determination of the three coordinates and three velocity components of the three-body-problem. "there are no more than the currently-known 10 algebraic
integrals (the "classical integrals" - six of motion of center of gravity,
three of angular momentum, and one of energy). This falls short of the 18 integrals needed"
 [snip...] I question such exact numbers, in general, and I suggest that the missing pieces of the puzzle (which would be hard to quantify, to say the least) are the relational pieces.
 
Another aspect of this which intrigues me is that Whittaker remarks on "excess degrees of freedom"-- which means, I take it, that he can't account for as many as there are, using his calculations.  This situation is rife with ironies, but this isn't really the place or the time, I suppose... In any case, once again, my intuition is that the extra degrees of freedom have to do with relational issues of context which have been dispensed with in the models he's using for his calculations.
 
The third issue here is also about those "excess degrees of freedom" that the system exhibits but which the calculations are mute about. This goes back to the discussion we had on non-holonomic constraints and how a complex system as a "maximally non-holonomically constrained system" has infinitely more degrees of freedom than a totally unconstrained system or even a minimally constrained system. The peculiar thing about non-holonomic constraints (meaning "context-dependent" constraints) is that such constraints end up changing themselves, in a roundabout way. This is what concrete numbers can't represent and they likewise cannot approach the flexibility that the constant interactions (of system with context with system; ad infinitum) create.
 
So, the solution to the three-body-problem will be a relational one. Are you going to give it a shot?
 
Judith

 
----- Original Message -----
From: Tim Gwinn
To: ***
Sent: Tuesday, January 04, 2005 11:47 AM
Subject: Re: [ROSEN] Rosennean challenge? Brun's

Dan,

Paraphrasing Brun's Theorem from ch. XIV, "The Theorems of Bruns and
Poincare" from E.T. Whittaker's "A Treatise on the Analytical Dynamics of
Particles and Rigid Bodies" (Cambridge Univ. Press, 1988), it essentially
states that there are no more than the currently-known 10 algebraic
integrals (the "classical integrals" - six of motion of center of gravity,
three of angular momentum, and one of energy). This falls short of the 18
integrals needed for a complete detemination of the 3 coordinate components
and 3 velocity components of the 3 bodies. This is why the 3-body system is
sometimes said to have "excess degrees-of-freedom", which I find a curious
inversion of which is the system and which is the analysis. Whittaker gives
the proof,  which lasts for 21 grueling pages. He cites Bruns original
publication of it:
Berichte der Kgl. Sachs. Ges. der Wiss, 1887, pp 1-55.

Sommerfeld's "Mechanics", Vol 1, p. 79-80, (Academic Press, 1943)  has a
similar discussion but without the proof.

Doh!! I could have saved myself some typing - Wolframs site has the same
information at
:http://scienceworld.wolfram.com/physics/BrunsTheorem.html

Regards,
Tim

> -----Original Message-----
> From: ROSEN Forum [mailto:*** Behalf Of Dan
> Fiscus
> Sent: Tuesday, January 04, 2005 10:06 AM
> To: ***
> Subject: Re: Rosennean challenge? Brun's
>
>
> Tim,
>
> Do you have a cite for the publication in which the
> Brun's Theorem is proved? Also, can you paraphrase
> or tell us what the proof shows or how it is done -
> what logic or key, etc. in general?
>
> Thanks for any info,
>
> Dan