Ruggero Maria Santilli Foundations of Theoretical Mechanics II Birkhoffian Generalization of Hamiltonian Mechanics s Springer-Verlag New York Heidelberg Berlin‘Ruggero Maria Santili ‘The Institute for Basic Research 96 Prsoot Sree Cambridge, MA 02138 USA. Ears: Wolf Beiglbéck Elliott H. Lieb Institut fir Angewandte Mathematik Department of Pysies Universitit Heidelberg Joseph Henry Laboratories Im Neuenisimer Feld 8 Princeton University 6000 Heidelberg | P.O, Box 708. Federal Republi of Germany Princeton, NY 08540 USA, Walter Thircing Insts fr Theoredsche Physik er Universitit Wien Bolzmanngasse 5 10125 Torino ‘A-1080 Wien tay Austria Library of Congress Cataloging in Publication Data Sani, Ruggero Maia, 1935— Birkhoffan gevealztion of Hamiltonian mechanics. (Foundations of theoretiea! mechanics; 2) (Texts and monographs in physics) ‘ibliograpty p. Troludes index i Mestanies” 2. Inverse problems (Diflerevial equations) 3. Hamiltonian sys- tems. T Tile” 1 Series: Santi, Ruggero Marg, 1935~ Foundation of theoreti ceil mochanics; 2. TH. Series: Tests and monogmapbs in physics, {Qas0s:25a° 1973 vol? S815 [S31] 8219519 [ons0s) All righ reserved [No patt of this Book may be translated or reproduced in any form without writen permission from Springer-Verlg, 175 Filth Avenue, New York, New York 10010, USA {© 1983 by Springer-Verlag New York nc. ‘Typeset by Composition House Limited, Salisbury, Baglnd, Printed and bound by R.R. Donnelley & Sons, Harrisonburg, VA, Printed inthe United Sats of Ameria 987654321 ISBN 0-387-09482-2 Springer-Verlag New York Heidelberg Brin ISBN 3-540-09482-2 Springer-Verlag. Berlin Heidelberg New YorkLeste volame 6 dedicate at mivt gonétere Mila ed Barmanno Santille con davosione od efitsContents Preface xi Acknowledgments xix Introduction 1 4 Birkhof’s Equations 2 44 Statement ofthe problem 12 42 Birkbot's equations 30, 43 Birktoffan representations of Newtonian systems 42 44 Isotopic and genotopciansformations of fis-order stems 46 45 Direct universality of Birkhol's equations SH hors: 4.1 Lack ofalgebrae character of nonautonomous BiskhofT's equations 6 442 Algebraic signicance of isotopic and genotopic transtormations 70 43 Havas's theorem of universality ofthe inverse problem for systems of arbicary order and dimensionality 75 44 Rudiments of diferent geometry 77 45 Gtobat treatment of Hamiton's equations 81 46 Global weatment of Biekhof's equations 85 47 Lieadmissblesymplectc-admissible generalization of Bickhofl's equa tions for aonlcal nonpotenil systems 90 Examples 98 Problems 108Contents Transformation Theory of Birkhott’s Equations 10 5.1 Statement ofthe problem 110 5.2 Transformation theory of Hamion's equations 114 5.3 Transformation theory of Bikol’ equations 126 hors: ‘Need to generalize the contemporary formulation of Li’ theory 148 52 Isotope generalization of the universal enveloping associative algebra 154 53 Irotopc generalization of Li's frst, second, and third theorems 163 54 Isotopic generalizations of enveloping algebras, Lie algebras, and Lie sroupe in classical and quantum mechanics 173 55 Darbou's theorem of the symplectic sd contac geometries 184 5.6 Some defsiton of eanoncal transformations 187 57 Isotopic and genotopi transformations of variational principles 188 Examples 194 Problems 197 Generalization of Galilei’s Relativity 199 6:1 Generalization of Harton Jacob theory 199 62 Inder universality of Hamilton’ equations 217 {63 Gonerlizaton of Gal's relativity 225 Charts: 1 Applications to hadron physics 253, 62 Applications to statsical mechanics 261 63 Applications to space mechanics 26 64 Aplications to engineering 264 65 Application to Biophysics 267 Examples 270 Problems 277 ‘Appendix A: Indirect Lagrangian Representations 281 AL Inditet Lagrangian representations within Bred local variables 281 ‘A2 lnotopie transformations ofa Lagrangian 289 ‘AG Indirect Lagrangian representation vie the use ofthe teansformation theory a ars: ‘Ail Analytic Newtonian systems 307 ‘82 Anus estensions of Lagrangian and Hamitonian functions 10 complex variables | 309 ‘A. The Cauchy-Kovalewki theorem 312 ‘Ad Kobussen’s treatment of Darbour’s theorem of universality for one- dimensional systems 314 |A.S Vanderbaowhede's funcional approach tothe averse problem 317 ‘AG Symmetries 221 ‘AT Lee's consrustion of symmetries of given equations of motion 324 ‘ALS Firat integrals and conservation laws 327Contents ix ‘A. Noother's consrution of it integral rom given symmetries 333, .10 Isotopic transformations, symmetes, and Hirst integrals 338 ‘A.M! Lack ofa unique elatonship becwoon space-time series end physical laws 340 ‘A.2 Classification ofthe breaki mechanics 44 Examples 348 Probloms 336 so space-time symmetees in Newtonian References 359 Index 365Preface In the preceding volume;' I identified necessary and sufficient conditions for the existence ofa representation of given Newtonian systems via a variational principle, the so-called conditions of variational seif-adjointness. ‘A primary objective of this volume is to establish that all Newtonian systems satisfying certain locality, regularity, and smoothness conditions, ‘whether conservative or nonconservative, ean be tzeated via conventional variational principles, Lie algebra techniques, and symplectic geometrical formulations. This volume therefore resolves @ controversy on the repre- sentational capabilities of conventional variational principles that has been lingering in the literature for over a century, as reported in Chart 13.12 ‘The primary results ofthis volume are the following, In Chapter 4,°1 prove 14 Theorem: of Direct Universality of the Inverse Problem. It establishes the existence, via a variational principle, ofa representation for all Newtonian systems of the class admitted (universality) in the coordinates and time variables of the experimenter (direct universality) The underlying analytic ‘equations turnout to bea generalization of conventional Hamilton equations (those without external terms) which: (a) admit the most general possible action functional for first-order systems; (b) possess a Lie algebra structure Jn the most general possible, regular realization of the product; aud (c) "aot (1978). As was the case for Voom the ferences ae Hite a heed of hie ‘ame rit in cronoogial once and then nsiphabt order Al rernee tothe prseding volume have the peta“ e, Section Tt, Egoation (hts) Seep her aed 16 rerf dane hn he atic fo he prosat *"To sess the comity with the te chapters of Volume I these of this volume are numbered 5,308 6aii Preface characterize a symplectic two-form in its most general possible local and ‘exact formulation. For certain historical ressons, indicated inthe text, Ihave called these equations Birkhof's equations. In Chapter $ 1 present the transformation theory of Birkhoff's equations. Essentially, it emerges that, while Hamilton's equations preserve their structure only under special classes of transformations (the canonical and the canonoid), Birkhof's equations preserve their structure under arbitrary, generally noncanonical, transformations. I then present a step-by-step ‘generalization of the Hamiltonian wransformstion theory. In addition, I point out that Birkhoff’s equations can be obtained from Hamilton's ‘equations via the nse of noncanonical transformations. The inverse reduction ‘occurs instead via the use of Darboux’s transformations of the symplectic ‘geometry. This allows the proof in Chapter 6 of the Theorem of Indirect Universality of Hamilton's Equations, according to which conventional Hamilton equatioas are unable to represent Newtonian systems at large in the reference fame of their experimental observation; nevertheless, 2 reprorontation oan always be achieved via use of the transformation theory. ‘As has been known since Gulile’s time, physics requires that abstract ‘mathematical algorithms admit a realization in the frame of the observer. ‘The inability of Hamilton's equations to satsiy this fundamental requirement confirms the need for their Bickhoflian generalization. ‘The analysis presented in these volumes therefore establishes that the treatment in the frame of the observer of Newtonian systems with unre- stricted dynamical conditions requires the use of generalized analytic {formulations for the most general possible first-order Pfaffian action and of ‘generalized geometric formulations for the most general possible local and ‘exact two-forms, These occusrences render inevitable a reinspection of Lie’s theory (enveloping associative algebras, Lie algebras, and Lie groups) to achieve a form which is directly compatible with the generalized analyt and geometric formulations—that is, a form which is classically of non- canonical character and quantum mechanically of predictable nonunitary character. This study is conducted in the final stage of a program where the existence of generalized algebraic formulations is shown. These formulations essentially consist of a reformulation of Lies theory that is directly applicable to the most general possible associative envelopes, the most general possible nnon-Hemiltonian/Birkhoffian realizations of the Lie product, and the most general possible noncanonicel/nonunitary structures of the Lie groups. By keeping in mind that Lie's theory was developed for the simplest possible associative product X/X, of the envelope, the simplest possible form X,X, — X,X, ofthe Lie product, and the simplest possible structure exp 0'X, ofthe Lie groups, the need for te reformulation under consideration i self evident, I bave called the emerging formulations isotopic generalizations, ‘where the term “isotopic” expresses the preservation ofthe primary analytic, Lie, or symplectic character. Tn this way, we see the emergence of the foundations of a Birkhoffian Generalization of Hamiltonian Mechanics whichPreface xi 1 applies to a class of physical systems broader than that for which Hamiltonian Mechanics was conceived—systems with action-ata- istance, potential, self-adjoint forces, as well as contact, non- potential, non-self-adjoint forces; 2, is based on an isotopic generalization of the analytic, algebraic, and geometric methods of Hamiltonian Mechanics; and 3. is capable of recovering Hamiltonian Mechanics identically when all nnon-self-adjoint forces are nll |A number of applications to systems of ordinary differential equations in Newtonian Mechanics, Space Mechanics, Statistical Mechanics, Engineering, and Biophysics are presented during the course of our analysis, with more specific treatment appearing in Chapter 6. With the understanding that quantum mechanical profiles are beyond the scope ofthis volume, T have briefly indicated the existence of an isotopic generalization of Heisenberg's equations, as well as of a number of related quantum mechanical aspects, for the description of particles under action-at--dstance, potential interactions, as well a5 contact, nonpotential interactions, which are conceivable under mutual wave penetration, and overlap. The rather old (and currently dormant) problem of the generalization of Quantum Mechanics is therefore brought (o life in an intriguing and direct way by the Birkhoffian Generalization of Hamiltonian Mechanics. Regrettably, for the sake of brevity I have been foreed to ignore several additional, equally intriguing developments such as the extension of Birkhoffian Mechanics to field theory—a study which has already been initiated in the literature ‘The mathematically inclined reader should be informed from the outset that T have given priotty of presentation to methods and insights, not only in focal coordinates but also within a single fixed system of variables, those relative (othe observer. The use of transformation theory is presented only as.asecond phase of study Finally, generalization via coordinate-iree, global, ‘and geometric approaches is presented as a more advanced approach. This style of presentation implies a reversal of the priorities of contemporary ‘mathematical studies, particularly those of geometric character, but it is dictated by specific pedagogical and technical needs. On pedagogical grounds, my teaching experience has suggested that itis best to expose students first to geometric structures in specifi loel variables and show thatthe essential geometric properties persist under arbitrary (but smoothness- and regularity-preserving) transformations of the local vari- ables. Then the students may be brought, in @ progressive motivated way, to advancod coordinate-free techniques. ‘The technical reasons for giving priority to formulating the methods in local variables are even more pressing than the pedagogical ones. Infact, the crucial inability of conventional Hamilton equations to represent New tonian systems in the frame of the observer can be identified only via the local formulation ofthe theory because, atthe abstract, coordinate-tre level, * Kobusen (197.iv Preface Hiamiltons ang Birkhofs equations are indistingushable. At any cate, & primary function of the Inverse Problem is to provide methods for the computation of an analytic representation of specie differential equations in specific local variables. Clearly, this task can be accomplished most ‘efletively via the local formulation ofthe theory. “These priorities should not be interpreted as denying the need for global techniques. On the contrary, thes tochniques will be quite useful throughout ‘ur study, particulary in proving the main theorems. ‘This work reflects the organization of Volume I: main text, a series of charts illustrative examples, and problems. The mai text is devoted to the simplest possible presentation ofthe technigues in loeal variable. The certs complement the presentation through more advanced topics in ‘Abstract Algebras, Functional Analysis, Differential Geometry, and other disciplines. The examples are intended to iustrate only the most important aspects. Finally, the problems are designed to test he student's understanding of the basic ideas and methods and to evaluate the student's capability for practical aplication. “The relevance ofthe analysis presented in this and the preceding volume can be indicated essentially as follows. Within the context of Theoretical Physics, the methods presented permit the idenifcation and treatment of a now class of interactions calle “closed non-selfadjoint "(Chapter 6). These interactions verify the conventional conservation laws of total quantities (sure), yet the internal forces are partially of action-a--distanc, potential type and partially of contact, nonpotential type (non-seadjointness).° ‘The interactions of primary intrest in contemporary physic (eg, letro- ‘magnetic interactions and the unified gauge theories of weak and clectro- ‘magnetic interactions) ur out tobe of the closed selfadjoin type upon the extension of the methods to relativistic and field theoretical settings (Sencili 1977a,be, and 1978). In essence, the transition from closed self-adjoint to non-self-adjoint interactions is given by the replacement of poinlike constituents with éxtended constituents under sufficiently small mutual distances. Points can only interact at a distance, thus admitting only solfadjoint interactions Estended objects, on the contrary, whether particles or waves, admit the additional contact interactions for which the notion of potential encray has no physical basis. Thus they are of the non-slFadjoint type. The former interactions are of Lagrangian/Hamiltonian type, while a necessary condi- tion for te later interactions tobe truly non-sel-ajoinis that they are not of Lagrangian/Hamiltonian type in the frame ofthe observer; yt they can be treated via the Birkhofian Mechanics and related isotope generalization » Aa in Voluse 1 tbe ems “cat wed jis matic! see of “guing” he sudent though te main ess of ore advanced opt, whe providing sete releencs fr sb- Sequent suds.Preise av of analytic, algebraic, and geometric methods. The conceptual and technical advances are self-evident. A rather forceful Newtonian example of closed non-self-adjoint inter- actions is given by our Earth. I'considered isolated from the rest of the un verse, the Earth verifies the conservation of total physical quantities, but the interior motions are of the non-sel-adjoint type, asis the case for satellites during reentry into our atmosphere, spinning tops with drag torques, etc. Tis often argued that nonpotential forces are due to the “immaturity” of the Newtonian description, and that the locel/potential/Lagrangian or Hamiltonian naturo is recovered in full when passing to elementary con- stituents of matter. This view has been eriticized in recent times because itis ‘based on the pointlike abstraction of the elementary constituents and because it ignores the experimentally established conditions of mutual wave over- lappings for all interior problems under strong interactions, such as the structures of nuck, of strongly interacting particles (hadrons), and of stars. ‘At any rate, the idea that the experimentally established ‘nonpotential [Newtonian interactions can be reduced to a large collection of potential interactions has no practical computational value owing tothe large number (of constituents of macroscopic bodies). It has no experimental support at this time, is therefore merely a scientific belie, and when subjected to an actual mathemsatical study, is aficted by a host of consistency probiems such as the need to recover nonpotential dynamics via a large collection of potential ones. In different terms, the Newtonian description ofthe structure of our Earth ‘with ils established potential and nonpotential forces in local or nonlocal” ‘treatment, is a model of invaluable guidance in the study ofthe more complex structures of nuclei, hadrons, and stars, rather than knowledge to be by- passed via pointlike abstractions of the elementary constituents. To put it quite candidly, Thave conducted most of those studies because ofthe poss- ibility that our Earth can be viewed as a Newtonian image ofthe structure of hhadrons, in the same way as our solar system is seen as @ Newtonian image of the structure of atoms, ‘Once the closed non-sel-adjoint interactions are acknowledged either as an experimentally established realty (classical mechanics) or asa possibility (particle physics), the relevance of the methods of these volumes becomes selovident. In fact, by recalling thatthe broader interactions considered, by conception, cannot be directly treated via Hamilton’s (or Heisenberg’s) Mechanics, the methods employed in these volumes permit the use of rigorous analytic algebraic, and geometric techniques which would otherwise be precluded. Besides the evident classical applications, the methods are potentially useful for the future experimental resolution of the problem ofthe structure of strong interactions; tha is, whether theultimate structure ofthe 21, should be sre tht this volume in goer and Bikhofs equations in parts, treat el nosed! ftracioas, The mola! noose nactons demand {eration of DirshoffTarsymploae farmalatins, ilo the so-alot Llewdmisble Fomiaons (ee Char 4)avi Prefuce universe can be reduced to a collection of points, or a substantially more ‘complex reality must soouer or later be acknowledged. ‘The relevance of the methods in Engineering is equally self-evident. Engineering systems are non-sel-adjoint as a rule and ate sel-adjoint only in very special cass, am referring to computer or electri systems inclusive of internal losses, trajectory problems with follower forces, etc. The tech- niques presented in these volumes allow the computation of an action fune- tional for all these systems, by treating them via well-established methods such asthe canonical perturbation theory and the Hamilton-Jacobi theory. ‘The reader can now see the relevance of the Inverse Problem for other disciplines, such as Space Mechanics, Statistical Mechanics, Biophysics, ete. ‘This volume originated in the following way. In Volume I, I reported on. studies of the integrability conditions for the existence of a Lagrangian as available in the literature and presented my own work on the independent existence of a Hamiltonian (ic, existence of a Hamiltonian without prior Knowledge of a Lagrangian). I also presented my methods forthe computa- tion ofthese functions from the equations of motion, when the integrability conditions ure verified. {identified the capability of these functions to repre- Sent both potential and non-potentil forces and treated a number of com- plementary aspects, ‘While conducting these studies, I became aware that the tiolation of the integrability conditions forthe existence of a Lagrangian, or independently, of @ Hamiltonian, isthe rule in practical cases and that their verification is the exception. Although T have no available evidence, I believe that this restrictive character of the conditions of seladjointness has been known since the early studies on the Inverse Problem in the last century, and this resulted in the subsequent lack of significant attention to the problem in both the mathematical and physical literature, as reported in the Introduc- ton to Volume I. ‘Clearly, in order to reach a level of practical effectiveness had to “solve the Inverse Problem.” That i, I had to identify methods capable of turning all non-self-adjoint systems ofthe class admitted into equivalent self-adjoint forms for which an action functional can (atleast formally) be computed. Ate a considerable library search (in addition to that reported in Volume 1, T succeeded in tracing efforts back to Mayer (1896); additional relevant contributions were made by Davis (1931) and Havas (1957). All these con- tributions deal specifically with the Indirect Lagrangian Problem and aze reported in the Appendix along with the Newtonian reduction of my fleld theoretical studies on the topic (Santili (1977), ‘Even though the methods permitted the construction of Lagrangian representations for genuine nonconservative nonpotential systems, the lack of direct universality of the Inverse Lagrangian Problem was 500n e- tablished.® This situation called for additional efforts, At this point the should dicate rm the ots tht bes itatlons fer special othe Lagrangians atetepoy oie pi se cpt sox on ein erPreface ii Independent Inverse Hamiltonian Problem became crucial. In essence, in Volume I, 1 had established the symbiotic character of the conditions of selfadjointaes for fest-order systems by ensuring: 1) Ue derivability from a variational principle, 2) the Lie algebraic character, and 3) the symplectic. seometrc structure. As indicated in Volume 1, Hamalton’s equations axe ‘only a particular case of these conditions of selF-adjontness. The existence of more general equations preserving the analy, algebraic, and geometric character was then ensured. I thereface proceeded to the identification of these broader equations as the most general possible form permitted by the conditions of selhadjointness. In this way, T “rediscovered” equations which had been proposed by Bickhoff (1927) without the alebraic and geometric techniques ofthe Inverse Problem. Theis diseot universality for systems of firs-order ordinary differential equations resulted from an Unexpected property identified by Havas (1973) Unlike second-order systems, first-order sjstems alvays admit (under suficient topological conditions) a regular matric of integrating factors which produces an ctivalent seladjoint form. The Birkhofian representation ofthe systems {5 then consequential, and its explicit form can be computed via the Converse ‘ofthe Poincaré Lemma Section 11.2). These stutis ate presented in Chapter 4 jointly with & number of complementary topics such as the indirect Birkhoffan representation of Hamilton's equations, the algebraic signifi cance ofthe selfadjoininess inducing and preserving iansformations of the equations of motion, ee "Tine next step of my studies was predictable and consisted of reducing a 1 shoul be sroed tat se ovon of aie ep of Delion 42 is agurdant fore symplaaie tron Infact the condita of egy alone sft forthe cha lonatin of sympleciev forms. This pots apt i istag the fe, same ‘bere inthe abaract cordintefteefeatent of eeomety, thatthe comet one Urpoeracy of form doesnot quran he geome hacen of alr wo Strained Newton jem, bca the need fran aden replay eonaton of pe eratBirks Equations 35 3. jis the geometical structure of Birkhof’s equations. To see this late point, onc ean write the tensor in the disjoint coordinates a) @p) (4221) Birkhof’s equations can then be written in the unified notation on Be TM 9,008" OL Qevredn 4222) or explcity, “ (42.280) SMA" = 5 (28 @R,) ., (25) E-Bene HEL 2ye002m (42230) ‘The first term i identically null (along a possible or actual path) because of the self-evident property that 0B OR), _ (0B aa" * “ae aa" ‘The last 2n terms of Equations (42.23) coincide with Equations (42.1). By recalling the symbiotic characterization of analytic, algebraic, and ‘geometrical aspects by the conditions of self-adjointness, al the above listed ‘properties of Birkhoff's equations can be synthetically expressed via the following property 0. (4224) Proposition 4.2.1. (Self-Adjointness ofBirkhoff's Equations). Necessary and sufficient condition fora general nonautonomous first-order system Clty aa? + Dilla) =O, = BQeeeIn (4225) hich is analytic and regular na star-shaped region 3* of pointsofR x T*M t0-be selfradjoint in His that tis ofthe Birkhoffian type, ie, Cpa? + Dy & (4226 Pnooe Conditions (4.1.32) re tho integrability conditions for wo-orms (4.152) ie, es Pareto) ee eee C= seulaae 9 a,36 BiekhotTs Equations tobe closed, ie, ae, =0 4228) (Gee the review atthe end of Section 41). The regularity condition implies that Ci of ‘aimal rank and, therefore, ia contact fom. The applicability ofthe convers ofthe Poineae# Lemna 1.22 impliss that form is exact, that is, 2 primitive one-foom Ry con x TM exists sch that Creat, 229) “Tho use of Equations (62.20)-(42.24)complates he proof that Equation (4228), under the conditions of sell-adoininess, necrasarily have Birkhofan structure (4226) The scene is tevaly established by the Dirt Poineaeé Lemma L121 (0.ED) ‘Thus Birkhof’s equations are self-adjoint in a way parallel to the self adjointness of Hamilton's equations. However, while Hamilton’s equations ‘area particular formadmitted by the conditions of selF-adjoininess, Birkhof’s equations are the most general possible form, & direct verification that Bitk- hof's equations verify al of conditions (4.1.32) is instructive (Problem 4.2) ‘The Calculus of Diferential Forms, a8 used for the proof of Proposition 42.1, provides aot only the integrability conditions for a two-form on Rx TT*M to be an exact contact form, but also a solution forthe primitive one- form, A straightforward use of the techniques reviewed in Scction L12 Equations (11.230) in particular) permits the proof of the following corollary. Corollary 42.18 (First?® Method for Computing the Bikhoflian Fun- tions from the Equations of Motion). Under tke condition of Proposition 42:1, the Birkhofian functions R = (—B.R,) can be expressed in terms of Equations (42.25) according to the rules R= [fe C60) ae We proved earlier in this section the contact geometric character of Birkhof's equations. For completeness, we must also point out the following difference between contact two-forms of Birkhoffian type and those most ‘commonly treated in the contemporary literature. The former posses, in ‘eneral, an explicit tine dependence in their symplectic substructure, while such a dependence is generally absent in the later. ‘The difference originates from the fact that contemporary treatments of contact two-forms have been usually patterned along the structure of Hamilton's equations, In this ease, one starts from the fundamental sym- plectc structure on T°M LQevo2m 4230) os itm dat = dp nak oq, dat» da i a 2 \aar~ ae" SAD Mp Neo 23) 2 Acditona methods wil be prevented in Sestions 44 andDBirkhotTs Equations 37 ‘and thea performs the prolongation (Chart 4.4) into the contact two-form on Rx TM ste» day = (RE _ ORE) ae y= yd de (E-3} Ra, y= O1 2.0.20 22a) ,) | a=Ga 423% whieh is the geometric structure of Hamilton's equations, asthe reader cen verily by particularizing Equations (42,21)-(4.224) for the canonical case R= R° = (p, 0). The point is that the symplectic structure (42.31) does not possess an explicit time dependence, and thie feature persists undot prolongation to form (4.2.32) ‘The situation is altered by Birkhof’s equations. As we shall see in the snext sections and inthe examples atthe end ofthis chapter, the computation ofa Birkhoffian representation for given Newtonian systems can be generally achieved in practice via functions R, with an explicit time dependence. Specific applications in mechanics therefore demand, in genera, the initc ation of the geometric study via symplectic two-forms with an explicit dependence on time 2; = 109 (t,dddat 4 dar = fonds) yp 5fPME2 PR ae ae, 23) ok38 Bickhofs Equations regular Bickhoffian representations) via the knowledge of functional de- ppendences (42.11a), under which we write ot = [eth a. — Be) fasten. 9% + 96.2 DH — BEEN] Jafntonmenore stn mtn B a Bie Sih) Plan Mn} farteon prs won] fetenen (4234) ‘where the M's are the implicit functions in the 5s, Equation (4.1.23), as characterized by the irs set of Equations (4,218) under regularity condition (42.12), One can seein this way that the Lagrangians are of second-order type (ic. dependent on the accelerations), although of the totally degenerate type “The variation of action (4.2.34) then yields equations aL dab ab ae ar* aar oF -- arp + arrge 4) at ae * 2,4 _aW 6238) aT ae which characterize a system of second-order differential equations, contrary to the expectation of their being of third-order, Furthermore, the system is linear in the acceleration whenever the ¥"s are independent of the velocities. “These results permit inspection of the Inverse Lagrangian Problem in a now light. In fact, the lack of direct universality of first-order Lagrangians may be due to the restrictions imposed by the first-order character. The direct ‘universality of Bitkholl’s equations and images (4.234) and (4.2.35) then ‘make it possible for the Tnverse Lagrangian Problem to become directly 2% Recht a oral dagenrete Lagrangian oscurs when each cmon of te Mesa i ldeoy tus Tt the ce foc condone Lagann whe tye nea In seconde onder deatBirkholTs Equations 39 univers fF systems of secondorder diftrental equations under the tolargooet ofthe Lagrangian othesoftaossondanie tally dona ‘ype toordng to heer, 2 dal oe ae ap ae gag L= Kile git + Wl 4g). (4.2.36d) ‘The study ofthis problem eft the interested reader Problem 49, ‘We conclude ts ston vith a ew sora omar geen 2.1) have boon sno, earn a det ot tndecsmgtch say peed tutors. Fits the equations coin ata as thet tucune concoct with Lagrange eguaionenfrtorder lly generate Lagrasgnns Ht by assuming for“ Lagrangian” the expres 6a DH + Bl 4D, (42.36) 10,0, 8) = ~Ryt, da" + Bl) 239 Lagrange’ equations coincide with equations (491), a (28, a, a aw — at Naar Ne (% a) Coo) However, the use of the terms “Lagrange’s equations” for Equations (42.1) would be misleading, particularly for the analysis of this volume. In fact, our objective is to seek s generalization of phase space formulations, hile Lagrange’s equations were conceived for configuration space formula. tions, and this sptit has persisted to this day.?" Additional studies more directly related to Equations (42.1) are those by Pfaff (1814). In fact, the primitive one-form leading to the equations is Pfaff's form (or action), as recalled in regard to Equation (4.2.14). However, itdoes not appear that Pfaff identified the true meaning of Equations (42.1) as bonafide analytic equations of mechanics. ‘These latter properties were identified in full by Bitkhoff (1927) who also provided explicit examples of applications to mechanical systems. Perhaps the best way to illustrate this historical point is through Birkhof’s original words (loc. cit, p. 89): “Suppose now that we take an extended Pinffan veriational problem SL [Extnont zine ay sateen yet en ren Elg-a)e 3 Lagrangians of type (4.237) repent fonder stems 20d as ch tey at along the pn spice ater tan the configuration pac) formation of mechani 2m 3)40 Biskhors Equations We propose to conser these equations inthe ease when there is an equilsium point athe oiga, under th assumption that the 2m analyte fonctions Xa such thatthe schew-symmetie datrminant {snot Oat the origin, The constant terms in the series fr the functions X, may obviously be omitted throughout tis clear thatthe Hamiltonian equadons appeat asa particular cae of hese Pfaffian equations (12). As willbe shown the following chapter, this generalization ‘ofthe Hamilton's equations poseeses the ame property of antomatialy fuling All ofthe conditions for compote staaliy, ones the obvious conditions for Sst ‘order sablty are satisie. Hence, rom this point of view, the Pfaff equations Seem as signifeant for dynamics athe Hamiltonian equations, although more ‘ener in type. Moreover, they poses the addtional advantage of maintaining their Plan form under an arbitrary transformation ofthe formal group” [Notice the clear identification of Equations (42.1) by BirkhofT as being (1) derivable from a variational principle; (2) a generalization of Hamilton's equations; and (3) "as sigoificant for dynamics as Hamilton's equations”? Specie illustrative applications were provided later on in Birkhoff's memoir land they are still recommendable for study. The extension to the nonautono- ‘mous case was provided soon after the quoted passage. FFor these reasons, Equations (42.1) were called “Birkthof’s equations” by Santilli (1978) and this terminology was subsequently adopted by a number of authors. Additional studies on the equations which deserve ‘mention are presented here. The equations were briefly indicated by Whittaker (1904) (whieh is the only reference known to the author for the period be- tween the studies of Pfall and Birkhof). After 1927, the equations were studied in more detail by Feraud (1930). Lee (1945), Pauli (1953), ané Martin (1959) studied them to a considerable extent, but primatily for quan tum mechanical considerations, More recently, Hughes (1961) considered the equations for relativistic treatments. [All the references quoted above treat equations of type (42.1). The alge- braic-geometric character of tensor (424) has been studied by numerous ‘authors, beginning with De Donder (1927) and Cartan (1971). “The variational selé-adjoininess of Equations (42.1) was identified by ‘Santill (1978e) by reaching the frst unified treatment of the analytic, alge- braic, and geometric properties of the equations. This author also initiated the first stady (see Santili (1978) of (a) the applications of the equations to the representation of local nonconservative Newtonian systems; (b) the 2 Notices Bikol emphasis on the repuaity of he equations The rade xn now se sat Defguion £2 apts nanber of feomsti an technol splementtions, has bees onesial to chine with Biko gia ve as closely a posible. Nous also Bikol s ‘atenacelogancs in esreseng the egal condton Tot its expr the orig, ‘Titise ct understanding tha is pesevation a ber points aonastwd under replatBitkhof?s Equations 41 equations’ consequential role for the Inverse Problem; and (c) the identitice- tion oftheir direet universality. The variational seltadjointness ofthe equax tions was subsequently studied by Sarlet and Cantrijn (1978) and Sarlet (1979) (foliowing a private communication by Santill, These authors also identified a Birkhotfian generalization of the Hamilton-Jacobi equations hich will be reviewed in Chapter 6, Additional studies, eg, on the trans- formation theory of Birkhof’s equations, were conducted by Kobussen (1978 and 1979), ‘The application of Birkhof’sequationsto Space Mechanies was studied by Broucke (1979), while the application to Biophysics was studied by Lumsden ‘and Trainor (1979). These and other applications ae reviewed in Chapter 6 Furthermore, a quantum mechanical generalization of Hoisenberg's equations which leads (o Birkhoffs equations under the correspondence principle has been proposed by Santili(1978d), and itis also reviewed in Chapter 6 Finally, the resder should Keep in mind that Bitkhof’s equations have ‘been studied by the author and presented inthis vobime forthe treatment of local non-potential forces (or interactions). In this way, the condition of Aerivabilty from a potential inherent in most of the Hamiltonian treatment is removed, but the locality condition persists, {In principle, Birkhofl’s equations might be studied for the possible repre- sentation of non-local nonpotential systems (or interactions), via integro~ diferential functions ofthe type = Roos [dimsiae, 7 (4.2.39) 2 on0 + faswiae, whieh, via the reduction to second-order form presented earlier i this section, ‘can represent nonlocal nonpotential Newtonian systems of the type me Finn) [de Kenesb e024) ‘This use of Birkhof’s equations, even though computationally con- As weal, tha ase czas rater reenty in past sppliaions.46 Bickhots Equations notion of Hamiltonian vector fel, via the inner product for the autonomous 0, = -aB 43.19) \with corresponding generalizations for the semi-autonomous and the non- In the local coordinates needed for practical applications to specific systems, Equations (4.3.14) read? 4622, OB os dB = — Sede 43.13) “The generalization of Hamiltonian structure (4.3.5) s0 as to preserve the underlying geometry (and therefore, the algebra) is solf-ovident. ‘As recalled earlier, first-order systems are generally given in form (43.6), in which case the Birkhoffian definitions given above do not apply directly. ‘This imitation is resolved via the following definition, Definition 43.1. A general covariant form (4.36), which is well defined, analytic, and eogular in a neighborhood # ofa regular point of the variables, admits & representation in terms of Birkhof’s equations when a regular rmatex (H3( a)) of mokipicaive functions which are analytic in d¥ and 2 set of Biskhoffan functions R(t, a) and B(t, a) which are also analytic in exist such that the following identities hol in in a given ordering: [@-B}-E-), eer nak HAD Qov2n 43.19) ‘The representation is called direet when (i) isthe unit mateix; otherwise, it is called indirect. Finally, the representation is called nonautonomous, semi- ‘autonomous, of autonomous when Birkhof’s equations are of the corre- sponding type in the sense of Definition 4.2.1 Ic is understood that the integrability conditions for representations (4.39) and (4.3.16) are equivalent. 4.4 Isotopic and Genotopic Transformations of First-Order Systems {As indicated in the preceding section, particularly in Definition 43.1, the construction of « Birkhoffin representation is essentially dependent on the capability of writing first-order systems in a self-adjoint form, and, more 21 importnt to ses evn thir introectry geome lee thatthe liferation eon te Hatin and Btn ator fs st forthe global coordina erm tows of geome. Ths pout tate the ned focal foxmlatios aid the ieficny oF thecooriaate fre approach a mechs, coosidered aloe. Fot mote deals te eo sce care ttm ced of i chap,‘otopic and Genotopic Transformations of Firs-Order Systems 47 specifically, on the capability of turning a given non-selhadjoint system info an equivalent self-adjoint form. A technical context for the rigorous treatment ofthese transformations is provided by the theory of Abstract Algebras and, in particular, by the so- called isotopies and genotopies. In this section we shal deal with the simplest possible part of the topic, that dealing explicitly with first-order systems. ‘A more technical treatment within the context of Abstract Algebras will be provided in Chart 5.2, Definition 4.4.1.°° An equivalence transformation of a system of ffereatial equations (or first- or higher-order) is called self-adjoint (non-selfadjoint) isotopic or self-adjoint (non-selfadjaint) genotople, depending oa whether the transformation preserves, in the isotopic case, of induces, in the genotopie ‘ase, the seliadjointness (non-selF-adjointness). ‘This definition has been given for all possible equivalence transforma tions, including those charaetcrized by the transformation dievry. Ine restrict the transformations to a fixed system of local variables, the only possible transformations are those characterized by the multiplication of a regular matrix of factor functions. The classification ofall possible cases from, the viewpoint of the variational self-djointness, then leads to the following possibilities. Cua + Dan = 0 (Cha + DDy (ata) (ud? + Doysn = 0 (Cha? + Day = 0, 4a.1b) (Cai? + Doron = 0- (Chl? + DP yaa = 0, (ao) (Crud! + Degg = 0+ (Cha? + DE san = 0, Gata) Ch = HC DI=HD,. (Aste) According to Defi — transformations (44.1a) are self-adjoint isotopic because they preserve the selradjointness of the original system; — transformations (44.15) are selfudjoint genotopie because they trans form a non-selF-adjoint system into a self-adjoint one, by therefore inducing the self-adjointness; — transformations (42-1c) are non-seffadjoin Isotopic because they pre- sotve the original non-selFadjointness ofthe systems; and, finally, 44.1, the above cases can be identified as follows: 2° RM Santil(177e) Witha minimalamount of inguistizese, he Grek or“ intope” Is Yoo, tnx, which means "sane coniguration The term =genoopie® has bees Sugg tole by Me Cal anil rom ths Grek er enor meting “induce coafguraton= PE Nauta the tet “agunalene™ doesnot posse ts traditional theta nest ‘within the conte of wansormatons (41) bets, onan anal and togla atk Ae Selese, the equlalnes character ofthe tenomnation alo depends on the spo oo ‘See For isan, (iH) may gorerae an equvalenc ransormation ofthe sxond- oder ‘ste Fi.) = Oba ot the first order sem Ff) = 0. Novell an Tocris consiered a this voluse se equrleno ransormaton by «constaon. This Imples lle seunary restriction on mapa ortions emsring he guano the ‘iil and antral systema We sal terete appiy the frm “equivalence entra tion to teansformatios of ype (4) a wel48 BickhotTe Equations transformations (4.4.1d) are non-self-adjoint genotopic because they transform a sel-adjoint system into a non-sel-adjoint form, by therefore inducing the non-selF-adjointness. ‘A few examples may assist the reader in identifying the type oftransforma- tions under consideration. They are given below for the second-order case, with the understanding that the first-order case follows similar patterns. “The equivalence transformation of the radial equation of a particle in @ central force field (studied in Example 13.5, p. 1212) wes) | [ter = 8 ovo 1 ye = ae + me HO) ~fe ler wre = [Be an is a selfadjoint genotopic transformation, The inverse transformation is then of non-selfadjoint genotopic type. What is important for this analysis js thatthe multiplication of the equation of motion by the term 1/m? clearly eaves the solution of the system unaffected (equivalence transformation), yeti isnot trivial from the viewpoint ofthe existence of a Lagrangian repre- Sentation, Infact, this representation exists (and is well-known) forthe sel- ‘adjoint form of the equations of motion, yet does not exis for the equivalent nnon-self-adjoint form. Notice thatthe form of the equation of motion origin- ating from Newton's second law (mi? ~ F = 0) is self-adjoint inthis case. “The equivalence transformation of the equation of motion of a particle subject to a linear velocity-dependent drag force (Studied in Example 13.1, .1.206) 0 443) (Fla <0 £ te va} (#0m=1 isa setfadjoine genotopi transformation, In this cas the transformation inducing seleadjonines is provided by te factor I/F which the reader can very through (12), Notce tha, in tis ease, the equation of motion Digtating ftom Newton's seco lw is nonaleadjoint For addtional illustrations ofthe self-adjoint genotopc transformations, sverefer the reader tothe eximples atthe en ofthis chapter (well 8 those I the end of Chapter Sand ofthe Append) ‘Th css of sal sdjoin isotopy fe similar to that of slhadjointgenotopy. For instance, the equivalence transformation of tho ope-dimensioaal harmon oselator (conservative cae), UF + Flag = 0 (G+ PULP + awd (asa)Isotopic and Genotopic Transformations of First-Order Systems 49 preserves self-adjointness. The reader is encouraged to verify this, The equivalence transformation of the particle with linear velocity damping (aonconservative case), Frode, also preserves self-adjointness. According to Definition 44.1, (44.4) and (4.5) ate therefore cases ofsal-adjoint isotopy. Equivalence transformations which induce the non-seliadjointness are particularly useful in generalizing variational principles so as to represent directly non-selF-adjoint systems, and they will be considered later. The same ‘transformations are also useful forthe non-Lie study of Newtonian systems (eg, that of Lie-admissible type), but this later approach will not be con sidered here In conclusion, the behavior of given system of differential equations under conditions of variational selF-adjointness is highly sensitive o the way sn which the system is written. In particular, the multiplication by a regular {and thus invertible) matrix of functions, while leaving the implicit functions {and thus the solutions) unaffected, generally alters the variational character ofthe system, early, the transformations that are important forthe Birkhoffian repre- sentations are those of self-adjoint (isotopic and genotopic) type. Their integrability conditions are identified in the following theorem where the terms “self-adjoint transformations” represent both the isotopic and the ‘genotopic ones as) ‘Theorem 4.4.1 (Self-Adjoint Transformations of First-Order Systems). Consider a first-order system Colt d + Da) 0, 22 1,2.052n 446) which ts welldefined, analytie, regular, and either non-self-adjoint or self adjoin in a region & of the variables (a). A necessary and sufficient condi- tion forthe transformation in & {56 ML Cadt, ada? + D4Ct a) ICA, aa” + DG, a)] = 0, (44.72) C= Can DE= Hey det) #0 (44.7) {0 be self-adjoint is that al the conditions Ch Cy = 0. (4480) ah | OCH | 0C%, Tet SE + Sno, (44st) ach, _ ov eG (4486) Byes are identically verified in &.50, Bickhots Equations ‘Asindicated in Section 4.1, in practice one often constructs firsta covariant normal form (4.131) for the possible identification of a Hamiltonian. When this form is not self-adjoint (and 2 Hamiltonian does not exist), one can search for 2 Birkhoffian representation. In this ease the following particu larization of Theorem 4.4.1 is useful Corollary 4.4.12. When system (44.6) isthe (not necessarily self-adjoint) ‘covariant normal form 49) conditions (48) forthe construction ofa slfadjoin general form Ale Nod” ~ El Dan = [Cot ob" + Dy en, 4108) Ge (44.106) reduce to ey + Hay (44.11) alg Fon (44.116) 4119) Sometimes, self-adjoint transformations admit a tensor Hf with constants elements The following particular case is then useful. Corollary 4.4.1b. When all elements of the matrix (Hf) are constants, ‘conditions (4.4.11) become tt _ A129) hy = -8, (44.12) (44.126) ‘The algebraic and geometrical implications of the self-adjoint transforma tions can be pointed out, ina preliminary way, via the following theorem. Its proof isa direct consequence ofthe algebraic and geometric meaning of the conditions of sel-adjointeness (Section 41) and, as such, is ignored here ‘Theorem 4.4.2. (Lie and Symplestic Character of the Sef-Adjoint Trans- formations). Under integrability conditions (84.8), transformed systems (442) have a Lie algebraic and a symplectic geometric structure, in the sense thatthe brackets a As = 4omegS, cm = (et ae 2a (a3)Isotopic and Genotopic Transformations of First Onda Systems 51 ‘are Lie, and the ovo-forms cr = ICM addat 0 dat (44.14) are symplecti. Recall that Theorem 441 applies whether or not the original system is selEadjoint. The following clasification of Theorem 44.2 then follows. (A) Selfadjoint isotopic sransformations. In this case the original brackets 2A co BB aa” ae a (481 Cals (4419) are Lie, andthe exterior two-form r= Hyde 1 dat 4419 is symplectic. We therefore have the Lie algebra preseroing transformation of the brackets 2A oy 9B 4.81 24 com 2B BowZ aay With corresponding symplectic preserving transformation of the two:form Cz = Hy da” 0 da” > Ct = $Cf, da® da”. (4.4.18) {B) Self-adjoint genotopic transformations. In this case the original brackets 2A oy 2B 4,5 how 4419) are not Lie, (2g, the tensor C*" isnot totally antisymmetric), Consequently, the tensorial two-form C1 = Cw da @ da? 44.20) cannot be reduced entirely to the exterior form (4.4.16)** and, as such, is not symplectic. In this ease we have the Lie algebra inducing transformation of the brackets (asa 2 neal fom Section L12 tha the leno product @ i nets eymietic nor an symmeii whlehe exteis product «is ttaly atiymmeti ree follow th, whenever ‘tensor Cy, santo he eater wo forme redaoe automata othe escior oe cording there Cn @ ae HC — Cor» dot Ha HCN da = Cd at the tensor Cinna aniymuntri i eduetion tt pole, ai th andelyng geometry ss notsymplete52. Birkhots Bquatons with corresponding symplectic-inducing transformation of the (tensorial) ‘wosform c Ch dat nda’ (44.22) ‘Note that al transformations (44.17), (44.18), (4421), and (4422) occur, by construction, within one single fixed system of local variables. Asa result, ‘the transformations express the algebraic and geometric degrees of freedom of the specific reference frame of the observer. Also, since no change of variables is involved, the transformations are a new algebraic and geometric type, which willbe studied in more detail in Charts 42. tis remarkable thatthe identification and teeatment of these new trans formations isa direct result ofthe conditions of variational selFadjointness. ‘By comparing Theorems A.LI and 441, we see a considerable similarity in the construction of self-adjoint transformations for second- and fist- order systems. Nevertheless, a deeper study reveals a rather profound dif- ference at the basis ofthe universality ofthe Inverse Problem, as well as of ‘number of important properties, [As stressed in the Introduction, second-order systems do not necessarily admit a sel-adjoin cransformation within a fixed system of local variables. AS ‘resulta Lagrangian or a Hamiltonian for the representation ofa Newtonian system i the coordinate and time variables of the experimenter does not necessarily exist ‘The situation for first-order systems is diferent. (Havas (1973). In fact, ‘swe shall show, first-order systems always admit a sel-adjoin transformation within fixed local variables. AS a result, whenever a Hamiltonian does not exist, Bickhoffian representation can be established. de? @ dat > CE ‘Theorem 4.43 (Universality of the Self Adjoint Transformations of First-Order Systems). Local, analytic, regular and even-dimensional systems of first-order ordinary differential equations always admit at least one self-adjoint ransformation in the neighborhood of a regular point of their variables PxoOr. To prove the theorem itis sucient to consider the ease when the functions of Equations (44.10) possess an explicit ime dependence. Equations (4411) always ‘mis a solution in the neighborhood ofa regular point because they canbe writ in the equivalent Cauchy-Kovalevsi form oh x 4423) and the functions, are analyte, Moreover, the intial conditions ean be chose in such ‘way that the tensor fn, bas acl severe, sy, Cy a the itil me = fy, Ths the Cauchy-Kovalewki theorem i verified. This ensues the existence ofan (analytic) >e-me proof dea withthe particular cat of Buaton (4101 extension othe genera case (42) el othe treed Teaer (Problem 43}.Isotopic and Genotopic Transformations of Frs-Onder Sysiems $3 solution Mf suc that. has cue tretue at limes as can be sen in the formal expression Melia) = Ham = [aE [ar + Ch (4429 Altensor of rank two with cur steueture werify Equations (44.118) and (44110) ee Problem 42), and this completes the proof of the thearem, (Q.E.D.) “Theorem 443 is remarkable inasmuch ast establishes that system (4.4.11) ‘of partial difeental equations in the unknown functions for Sxed terms >, and &, is always consistent, despite its overdetermined character, with 8 similar case occurring for the more general system (44.8). The geometric implications ofthe solutions ar also ineiguing. Recall that the systems considered are nonautonomous and that, a8 such, they can be more properly described va the contact geometry. Real from Section 42 that contact two-forms, in their current general formulation, have atached symplectic forms without an explicit time dependence. Theorems 4.1, 442, and 443 establish instead tha the presence ofan explicit time dependence in the symplectic form is rather natural in mechanics ‘The following method forthe explicit construction ofa slladjoint form 4ue to Hojman (1981) i important on both formal and practical grounds. Proposition 44.1 (A Method for the Construction of a SelAdjoint First-Order Form). Consider @ contravariant, first-order, normal form, =P B= 122m (4425) which is analyte inthe neighborhood # of a regular point of the variables, and suppose that 2n independent first integrals, a), ia) us (44.268) dex(al"/2a"y) # 0 4260) pe eee eee 6, ax(l- o re Coho? +f) <0 ca 00, _ 00) a 29 = (3), 7 ee on is self-adjoint in 2.St Biekhols Bquations ‘The proof of Proposition 44.1 is left as an instructive exercise for the interested reader (Problem 44). The construction of the Birkhofian func- tions via the method of the proposition will be presented in the next section ‘The reader should keep in mind that (as was the case for Theorems 4.4.1 and 4.42) Theorem 443 also applies whether or not the original system is seléadjoint, In the former case, an analytic representation already exists, While in the later case, i is induced by the transformation, By recalling the remarks following Theorem 44.2, we can conclude this section by stating that the selPadjine {soropic (genotopic) transformations preserve (induce) the derivabilty of he system from a variational principle, its Lie algebra character, and its symplectic geometric structure 4.5. Direct Universality of Birkhof?’s Equations Definition 42.1 aad Theorem 4.4.3 are euficent for setablishing the direct universality of Birkhof's equations. Nevertheless, for the sake of complete- ness, we shall give belowa more direct proof based on the Cauchy-Kovalevski theorem (o establish the existence ofthe Bitkhofian functions. ‘Theorem 4.5.1°7 (Direct Universality of Birkhofs Equations for Local Newtonian Systems), All! local, analytic, regula, fcesdimensional, wa- constrained or holonomic, conservative or non-conservative, and selfadjoint (or non-self-adjoin systems in fst-order form always admit, in a star-shaped neighborhood of a regular poin: of their variables, a representation in terms of Birkhof's equations in the coordinate and time variables of the exper rmenter. PROOF. Unconstrained systems ofthe cass admitted ae given by the escatialy non- seltadjont systems (Definition 4.1) In the (Cartesian) coordinate and time variables ofthe experimenter, thay canbe wetten"® {Uiaie~ falls Desa ~ Falls ts nanan — Flt Danan = kRh2S RQ WN eB | AS) sad they do not admit « Lagrangian (or a Hamilionan) representation in the local ‘arabes considered. "To radco the systems to equivaleat fs-ordr forms, introduce the physical (gener ally non-canonical) near momentum Pie Mahe 4s) 2 Tue Lagrangian veson of the theorem (ee Chart 42) was formulated and peoven by a 1973) Thora 4a the given Biola ven was given by Sani 180) Nore the sts aed ot acess) te Newlin the) may have celeron dependent Feces), The tnorea ly extendable tosses of ede higher than one vate raluetion Sade form presented in Chat 3, "ein gutions (451) ao seaman onthe repented ander exeDirect Universality of Bikol" Equations 5S sa realization of prescriptions (4.23). The contravariant normal Forms (128) are ‘hen given by © asa) (2) Gessner srry + 22%.) and can be writen in wife notation o-() ‘Our proof ofthe theorem consists of showing that, under the assumed smoothness (stay asa) and regularity conditions, the fundamental equations (439) for & Birktotian repre: Rea) 284) | AR,0) it ee Waheed 455) salwaysadmita solution in the fonctions R, and B, that, soltionexissforanarbieary Tunetiona eepencence ot the vector noid. 3 Such, the proo stomatal extends 10 arbitrary prescriptions (2.23) other than the physial selection (4.5.2), a well a «© acbitrary second-order systems ofthe class admitted, eg, holonomiesysoms in the senzral form (49) By recalling particularzation (4.22), the conventional Hamilton Fepresenation of systems with potential forces i tiv subease. (Case 1. The Funetons R, have am Explicit Dependence on Time (Now-autonomous Case). For ay sven fnction B, equations (4.55) ae ofthe Cauchy-Kovalevki type, as one «an se by writing them in the form aR, _(2R,_8R,).,_ 2B ar Gee) aso “Then, under the assumed smoothness regulary, and cat coaditions, Theorem | of ‘Chart 3 holds, anda soltion alway exit, "Notie that this ease applies also when the vcr Held is autonomous, by therefore being suficientes to prove the theorem. Neverthe oben ine with contemporary formulations of contact two-orms?” the case of autonomous functions (2), whee applicable, is relevant ‘Case 2. The Functions donot ave an Explicit Dependence on Time (Se-eutonomous sn Autonomous Cass). The fundamental equations fra irkhoflan epresenttion ‘are now given by Equation (4.3.10), fr the Semiautonomous cate, and by Equation (44.1), for the autonomous ease. However, bth sts of equation ae ofthe Cauchy ‘Kovalev form, gone ean ee by writing them i the form. asp under ideaication asa) nd this completes the proof ofthe theorem. (Q-ED.. > See the remach in Seton 42 folowing Coola 4S.$6 Biekhot's Equations ‘A number of physical and mathematical properties deserve inspection. First, we would like to identify the nature of the “direct universality” of Birkhoff's equations. This can be done by identifying all the mathematical algorithms at hand, that is, the local coordinates ¢ and a = (f, p) and the function B (the meaning ofthe functions R, wil be identified shortly) Corollary 41a. The direct universality of Birkhoff’s equations for local unconstrained Newtonian systems in Euclidean space is characterized by the following properties. 1. The loeal variables ¢ and ean be the time and Cartesian coordinates actually used by the experimenter, 2. The variables p can be the physical linear momenta mt, 3. The function B(, a) can be the physical energy Ein that ithe sum of the Kinetic energ} and ofthe potential energy of al self-adjoint forces. Note thatthe physica energy can be equivalently defined as the total energy of the maximal self-adjoint subsystem of (4.5.1), Needless to say, the total energy is generally non-conserved because of the presence of contact non- potential interactions. The definition i also introduced to stress the distinc- tion from the familiar canonical Hamiltonian (which, as pointed out in (Chart A.1I is often “conserved” while the system is nonconservative). ‘The reader can now appreciate the importance ofthe direct universality of Corollary 45.18. In fact, lacking a precise physical identification for all the ‘mathematical symbols at hand, one risks drawing mathematically correct ‘conclusions which are physically meaningless. This situation becomes even ‘more pronounced when one confronts the problem of quantizing non- potential interactions, asis expected for mutual penetration of wave packets? In thisease the canonical momentum “p,” the canonical angular momentam “M" =r X Page and the canonical Hamiltonian “HI” do not represent the physical linear momentum, the physical angular momentum, and the physical eneray, respectively, asa necessary condition forthe existence of¢ Hamiltonian representation. But then, the attempt to preserve conventional quantum mechanical settings (eg, the spectrum of "H” interpreted as “energy levels,” ‘or the spread “Ap” interpreted as “uncertainty in the momentum,” etc) risks being sterile. Tt is hoped that the reader will begin to see a reason for this volume’s emphasis on achieving analytic representations of Newtonian systems frst in the variables and functions of direct physical meaning. Once this has been achieved, one can then study mathematical topics (Such as nonlinear, experimentally unrealizable transformations of the coordinates), by mini- ‘mizing possible physical inconsistencies. Corollary 45.1 essentially states that one can first identify the quantities t,8,p,and B= By directly with physical quantities, and then search for a Birkhoffian representation. But che quantity B represents, inthis case, only “© This problem wil be touche in the ebarts of Chapter 6Divet Universality of Birkhof's Equations 57 potential forces. This creates the need for identifying the ways in which Bitkhof’s equations represent the remaining nonpotential forces. At this poiat the geometric or algebraic structure of Birkholl’s equations acquizes a direct dynamic content. Corollary 4.5.18. Under the conditions of direct universality of Corollary 45a, all nonpotential (non-selPadjoint) forces are represented by the covariant symplectic tensor aR, aR, 0,40) = Fs - Bs, (459) or, equivalently, By the eontravariant Lie tensor QH(ea) = (IM). 45.10) [In particular, when the Birkhoffian represents kinetic energy onl, all acting forces are entirely represented by the geometric or algebraic tensor. “The diferenc between the Hamiltonian and BirkhofBan sine evolutions (en, forthe autonomous case) 4a) = 34 on lam, (asia) 4 cay = [A BY sth) ‘can now be understood. In the conventional Hamiltonian case, all forces (whether potential or not) are represented by the Hamiltonian. In fact, the fundamental Lie tensor oo" has constant elements and therefore does not carry a direct dynamic content. Inthe transition to the Birkhoffian case the situation is different insofar asthe Birkhoffian represents ony part ofthe acting {forces, while the remaining forces are directly embedded in the structure ofthe ‘rackets of the time evolution Equivalent, we can say that the direct universality of Birkhoff’s equa- tions in the final analysis, isa consequence ofthe utmost possible use ofthe underlying geometry and algebra. When the realizations of symplectic two forms and Lie brackets are restricted to canonical forms, the capacity to ‘represent unrestricted systems in the coordinates of the observer is lost. For an explicit illustration of this important function ofthe geometry and algebra, we recommend that the reader consult the examplesat the end ofthis fand the next chapters, with particular reference to Example 41 on the Hamiltonian ond Birkhofflan representations of the (Newionian) electro- ‘magnetic interactions, ‘After having identified the admissible physical meaning of the focal vari- ables, the Birkhoffian functions, and the underlying geometrcfalgebraic tensors, the next objective is to characterize physically the space in which [Birkhof’s equations act. The relevance of the characterization will pointed‘58 Birks Equations ‘out soon. Note that, on mathematical grounds, the problem has been solved in Section 4.2 via the identification ofthe local variables af with a chart of the cotangent bundle T*M. It is advisable to compare the most salient physical properties ofthe cartier space of Birkhol’s equations with those for the Hamiltonian case. Corollary 45.16. While she variables rand p ofa Hamiltonian representa tion are canonically conjugated, ie, they verify the canonical rule (4s.2) and span a phase space, the variables rand p ofa Birkhoffian representation ‘are not canonically conjugated because, in general, oB to as) ‘As aresuit the space ofthe Birkhoffan variables, a = (ep) isnot necessarily ‘a phase space; i will be referred to as a“dymamie space" In particular, while ‘the phase space can be equipped with a fundamental Lie algebra structure which represents directly the fimdamental Posson brackets ern (MEM) (4°) a this structure is inapplicable ro the dynamic space and must be replaced with the general Lie algebra structure, "(a)" which now represents the ‘generalized fundamental brackets ervey = e299 = (GeO) a ($ *4). esto As a result, components of coordinates and moments with different indices ‘commute inthe phase space, Grl=(ord=Cord=0 eh 4516, but they do not generally commute inthe dynamic space oo eee ee aay ‘The loss of the conventional phase space and its replacement with a more general space has a rather deep impact in mechanics. For an idea, consider the problem of quantizing non-potential interactions when the classical equations are given by Birkhof’s (rather than Hamilton's) equations. Under “We extade the nonautonomous case bcauso of he lack of alba ch hasPs equations (Cant, te of BikeDirect Unversiliy of Bitkhots Equations 59 these circumstances, conventional physical laws for potential interactions, such as Heisenberg's uncertainty principle, ArAp = 3h, (4518) cannot even be consistently formulated, let alone applied, trivially, because of the loss of the quantum mechanical version of fundamental brackets (4514) ‘The differences between the dynamic space and the conventional phase space constitute one of the best mathematical formulations of the physical differences between the potential and nonpotential interactions. Jointly, the differences illustrate the essentially misleading nature of the Hamiltonian formulations when applied to nonpoteatial interactions, unless proper care is used for the physical interpretation of the algorithms at hand. Infact, if Hamiltonian representations are used for non-potential systems, conventional quantum mechanical images are expected to apply, leading to principle (4.5.18) (because now commutation rules 4.5.14) hold). This con- ‘lusion is matheraatically correct, but its physical interpretation isin doubt because, as indicated earlier, a necessary condition for the existence of the ‘Hamiltonian representation is that the quantity “p” does not represent the physical linear momentura, Inthe transition to the Birkhoffian representation of the same system, according to the direct universality of Corollary 45.1a, insidious physical eccurronces of the type considered are removed by con struction. However, a generalization of basic physical laws which is more irectly compatible with the underlying generalized algebra and geometry appears unavoidable ‘Occurrences of this type should not be surprising. Hamilton's equations (without external terms") have been developed throughout this century for the study of potential interactions. Birkhoff’s equations have been redis- covered for the study of fundamentally more general interactions. The fate of the underlying physical laws is then predictable. ‘After having identified some preliminary physical aspects of Birkhoffian representations, the next objective is to determine methods for computing the Birkhoffan functions from the equations of motion. Notice that first method is provided by Corollary 42.1a. However, the method is of somewhat formal inspiration and, as such, cals for reformulation into an operational version. The identification of additional methods and their interpretation is also desirable. Corollary 45.14. Suppose that a second-order Newtonian system is given, and its equioalent contravariant first-order form (4.5.4) has been constructed tia physical prescriptions (4.52). Some methods for the construction of the [Birkhoffanfintions Ry(, a) and BC, a) from the equations of motion are the Sollowing. * seefootnate of the atoduation.6 Bickhots Equations Method 1.*> Identify B with he total energy Ey inthe sense of Corollary 45:1a, and then solve the Cauchy-Kovalevski equations (4.5.6) in the Sanctions Ry. ‘Method 2° Construct a self-adjoint covariant general form BueoreThions0 Se) hema of Sn 4Phe fn ee ey nivo=[frein}, asap onthe thon prob hae mon [fale B)eofe aan Method 3. Suppose that 2n independent first integrals 1%, a) (in the sense of Proposition 44.1) are known. Then functions R,, from Equations (44.286), are given by sean ont a and the Birkhoffian is given by Be,a) (4523) A few comments are in order, The first mothod is clearly inspired by the desire to have a direct physical meaning for all local variables and functions. However, the method leads to functions R, possessing, in general, an explicit dependence on time, as evident from Equations (45:6). The geometric ‘implications ofthis dependence have been indicated in Section 42, and the algebraic implications are pointed out in Chart 4.1. The reader shoud be fully aware ofthese implications before passing to applications (eg, quantiza- tion) A method for attempting the elimination ofthe explicit dependence on time will be worked out shorly. ‘The second method is recommended when no physical condition is {imposed on the meaning of the Birkhoffian and on the preseriptions for the ‘construction of the first-order form. It is often preferable in practice, clearly, because of the greater freedom in the Bitkhoffian functions. Notice that, compared with Proposition 42,1, the method stresses the need to compute first the Rfunctions and then the Birkhoffian, as clearly expressed by the ‘contribution of the former to the latter according to Equation (4.5.21). This Sans (9780) Homa (580,Diner Universality of Bikhof's Bquations 61 necessary procedure is somewhat hidden in formal method (42.30). This second method is more readily set for the semi-autonomous ease, that i, for the representation of nonautonomous Vector fields by autonomous R-fune- tions and the consequential elimination of the problematic aspects of Chart 4. ‘The third method is conceived to provide a first interpretation of the re~ sults. Tn fat, it essentially emerges thatthe Birkhoffan functions ar fnctions ‘of a maximal independent set of fst integrals. It should be stressed that none of the methods guarantees the capability of actually constructing the Birkhoffan functions in the needed explicit form. In fact, the solutions aze often expressed by power-series expansions, Theorem 45.1 guarantees their convergence and therefore, the existence of @ solution, ‘bu the computation of the sums in the needed explicit closed form may often turn out to be beyond practical computational capabilites Example 44 has been included to illustrate the practical dfficultes in the construction of the Birkhoffian functions. The possibilty of constructing approximate Birkheffian representations (that ia, representations provided by the first terms of convergent power-series expansions) should not be over- looked. In fact, Physics is intrinsicelly an approximation of nature. The important point is to identify the degree of approximation which can be accepted for the case at hand.*? This line of study is left to the interested reader (Problem 4.5) ‘After the identification of the methods for the construction of the Birk- ‘hoffian functions, the next problem is to study their degrees of fedom, that is, their fanetional arbitrariness for fixed implicit functions (or solutions). Corollary 45.16. A given second-order Newtonian system verifying the conditions of Theorem 45.1 always admits infinite varieties of equivalent Birkhogfan representations all in the same time and coordinates of the experimenter. Some ofthe functional degrees of freedom are the following. (Class {. An iayinite variety of preseriptions (4.1.23) exists, ie, ee MG KEL GH LN G52) forthe construction of equivalent normal forms, one for each selection ofthe “arbitrary functions M,, (subject to regularity conditions (4.1.24), For each of these infinitely diferent possiblities, Theorem 4.51 applies and the corresponding Birkhoffian representations are equivalent, in the sense that they can all be reduced to the same second-order system. + another important iference betwen methods (2.30) and (4520)-(4521) ita inthe former the efstorntion les tat fe, whieh aoreiton absent re eter. Tar ntnoy, he approximation of Newtonian terme athe facto = T Vand = 7° V shouldbe foeced bose W Recall imple the eksence of perpetual ation ‘our eovtoament Alon hese nes cs onpotenn Bikholian approximates aware scopable becuse provider 4 guantnte tentment of ha soncogurative carte? of ‘esate undetandingi tht hy ne mess shoud sucha Bukbotian epesntation be ‘oneldeed eominal fo charter, owing to he move false soploalniegal abd not potetalinonsa adjoint matare ofthe sym sade at the ed of Sesion 82© Birkhoms Equations Class 2. For each prescription (45.24) and each selfadjoine covariant enera form, an infinite variety of diferent isotope transformations exits 56 MM, ae + TYG, a)a}sa = 0, deueX) #0 4.525) RG, a) mua m0= m0 samy All the corresponding Birkhofian representations are equivalent inthe sense ‘hat Birkhof's equations are he same fr all posible functions (4.5.26), ie, pa a aa" JO Var * ae in, om), (om am (2: — (+88). wan Note that degrees of freedom for Class actually imply the initiation of the transformation theory. In fact, starting from the physical vatiabies a . B), p = mi, the degrees of freedom imply the transition to the different variables a’ = (, y) where y now is no longer subject to the condition of rect physical meaning. Cleary, the transition a -+ a isa particular type of transformation in the cotangent bundle. As such, it will be studied in the next chapter. We have included the case here to stress the property that ‘Theorem 45.1 is consistent for all possible prescriptions (4.524), whether physically inspired or not ‘The depsees of freedom of Class 2are a direct result of the methodology of the Inverse Problem and can be constructed via the following rule of Birk- ofan isotopy Nu - (B+ Vike 4528) Clearly, hero we have a cotangent bundle image of the Lagrangian iso- topies of the Appendix. Intriguingly, this image permits the achievement of the following new interpretation of the isotopic degrees of freedoin (whoseDirect Univeral of Birkhof?'s Equations 63 Lagrangian counterpart is still unknown at this time). Recall from Section 44 the construction of a selF-adjoint form through the use of arbitrary func- tions of independent frst integrals (Proposition 44.1), Recall also from Corollary 4.5.ld thatthe construction results ina method for the computation of Birkhoffian functions (Equations (45.22) and (4.523)). By reinspeeting these results wesee that different Birkhofianfunctions which can be constructed bla rule (45.28) can represent the arbitrariness of finctions G,(T) tn the first integrals 1, as well asthe functional degrees of freedom of the first integrals themselves The gauge degree of freedom of Class 3 is trivial and can be best proved by ‘writing Birkhof’s equations inthe Lagrangian form (4.2.32), where we haves situation similar to that of Equations (A.2.3) i.e, Lm Rit, ad" + BG, a) + L! = — RY a + BIG) =-R@+B- Od) 452%) (4.5290) The applications ofthe degrees of freedom ofthe Biskhoflan representa- tions are intriguing. Below, a few representative cases are given. Tin Chart 16 we recalled a rather old and controversial aspect of mech- anies. It consists of the fact that, on one side, Hamilton's equations possess 2 symplectic structure in a rather lear and direct way while, on the other side, the variational principle from which the equations arc derived, the con- ventional Hamilton's principle in phase space 8 faune— mB) ~ 9, 4530) does not appear to possess a symplectic character in an equally clear way. ‘Apparently, this i the bass ofa tendency in contemporary circles of mathe- ‘maticians to ignote the treatment of mechanics via vasiational principles ‘and restrict the geometrical study to the analytic equations themselves. ‘Two seemingly independent resolutions ofthis controversy were proposed in the preceding volume. The fist is given by the reformulation of Principle (4.530) via the unified notation a = (rp) after which the geometric character of the integrand of the principle (as the contact canonicel one-form) is ‘expressed more transparently, together with the corresponding character of the analytic equations, ie, 8 facege ~ moet) » 8 [ene dar — Hany 4530)6 Dirkbofs Equations Independently from that, we introduced in Chart 13.6 a reformalation of Principle (4.5.31) with an explicit symplectic structure in the intengrand of ‘he action tel aesording to te equation 8 [lation econ) = [al (ona — Z2)se | = 0 (sa Evidently, even though the integrands of principles (45.31) and (4.532) are diferent, the underlying analytic equations are the same, and we can write 5 facrsa — mB) = 6 faderona’ — VB. (4533) Inspected within the context of Corollary 456, this degree of freodom results in being trivially given by the gauge 2G aa RO RS = RO + e "p oy Another application of Corollary 45.eis given by the possible removal of the explicit time dependence in the R functions. This ean be done by erans- forming non-autonomous representations Ry(t, a) and B(t, into an equivalent ‘semiautonomous forms R.(a) and Bt, a). A formal solution can be obtained via the degrees of freedom of Class 3, and reads aa nico mio~ ne, o- fawr a sas In this way, one can first compute the functions R, and B as they originate naturally rom Corollary 4.5.1 (with an explicit dependence on time), and ‘then attempt to eliminate such a dependence via rule (4.5.38). Note that ifthe original Birkhoffan represents total energy, the new Birkhoffian eainot preserve this physical meaning under transformation (4.5.35). ‘The algebraic implications ofthis are nontrivial, As we shal se in detail in (Chart 4.1, if Bis identified with the physical energy Ey, the Birkhoffan time evolution for non-autonomous systems cannot have @ Lie algebra character. If such algebraic character is desired, the Bitkhofian cannot represent physical energy. To state it in diferent terms, the Lie algebra character of the evolution and the direct physical meaning of the Bickhoffinn, rather sur- prisingly, tun out to be mutually exclusive in a number of cases of Biek- hofian representations. We end ths section wit the representation of Hamilton's equations in terms of Birkhoff's equations. Recall from Section 4.4 that the isotopic tansforma- tions are universal for first-order systems. As a result, they exist also for Hamilton's equations in all possible Hamiltonians. More explicitly, con- sider Corollary 44.12, and suppose that covariant form (8.49) is that of Hamilton's equations with 2, = 2H/ad". Theorem 443 establishes thatDirect Universality of Bickhols Equations 65 Equations (44.11) always admit a non-trivial solution in the isotopic fume- tions; thats they admit a solution (ig) other than the identity forall possible ‘Hamiltonians. in this way we reach the following property Proposition 4.5.1. (Represcatation of Hamilton's Equations in Terms of Birkhof’s Equations). Hamilton's equations in all possible analytic and regular Hamiltonians H(¢, a) always admit an indirect Birkhorfian repre- sentation in a star-shaped neighborhood of a regular point of their variables fee's) =e ie 4339) Upon computation of the isotopic functions for each given Hamiltontan via the solution of Equation (44.11), the Birkhoffan functions are given by [frontend ass [f(t Sea} eas ‘We should stress that the universality of the isotopy exsisfor the transition ftom Hamilton's to Brkhof's equations. The inverse ease isnot universal; thas, for arbitrarily geen fuetins Rand B the decomposition of Bikal’s equation into the Hamiltonian form according 10 rule (4.5.36) does not neces sarily exis in fact, the direct universality of Birkhof's equations for local Newtonian s}stoms, as compared tots ebsence for Hamilton’ equations, i ‘due precisely tothe lack of general existence ofa reduction (4.5.36). Ta Corollary 4.5.16 we have stressed the fact that, general, the conven- tional phase space character of thecarier space of irkhofTsequationsis los. Proposition 45: permits the identifcation of the following. property Whenever Birkhofs equations with a tensor Qt ) other than te funda imental symplectic tensor Cy, adit a factorization of Taniton’s equations according orale (4.5.36), the variables a = (,p) span a phase space, that I, the oariablesv and p are canonically conjugated. ‘Proposition 451 opens up, atleast in principle, new possibilities of r= search for consercaive systems such as the three-body system. In fact, as te study ofthese systems via Hamilton's equations can be considered 33 Virtually exhausted a this time, the representation of the same systems via Birkhof's equations permits the application of new, moze general methods ranging ftom new fits integrals to generalized perturbation techniques. In conclusion, the Birkhoffian generalization of Hamitonian mechanics is useful not only forthe non-potental systems for which twas conceived, but also for the more conventional potential systems.aa pee ty “w= '0 oat ron fonane = sa) +) Sot ton GB) Ae serio {6 BikhofTs Bevations a 0 ed) oN wofesosaou eReDirect University of BickhofTs Equations 67(68. Birkhof’s Equations Chart 441 Lack of Algsbaic Character of Nonautonomous Birkhoft Equations ‘As commonly understoos in the contemporary theory of Abstract Algebos, {an algebra U is 9 vector space of eloments a, 6, 6». over a field Fof atersticp (0, or prime) with sfements 2, | squipped witha balinear (abstract) product ab eastying the right and lft dstrbutive lure a(b +0) = ab + 90, ray (a+ be = 00 + be, 1b and the sealer ows*? (2ayb = a(ab) = o(00) @ for all eloments 2, b, ¢ © U, and 4 €F. Additionally, when the associative aw, [a, b,c) © abe) ~ (ab)e = 0, a is verified forall elemonts 2, 6, ¢ € U, we have an assoviatve algebra: othenvise, we havea nonessgeiatine algebra. Inthe contemporary erature, the term “sigebras” generally represents “nonessociative algebras,” and the same usage is adopted in this volume (unless the adjective “associative” is explicitly used). truly large variety of algebras, hhavo been identified in the mathematical and physica! itersture. They 110 cheractoriaed fist by the distributive and sealer lows (Toe the product to charecterize an algebra), and then by additional, specific laws (also Called ideniies or axioms). ‘The aigobras playing a relevant role fr the analysis ofthis volume are the following, 1. Lie Algebras. They are algebras L over F characterized by the las ab + ba =0, (as) _albe) + B(ca) + fab) = 0. (4b) 2. Lie-Admissible Algebras. They are algebras U over F such thatthe attached algebra U-, whichis the same vector space as U equipped with the product (2, bly = ab ~ bo @ is @ Lie algobra. Associative algebras 4 are cary the simplest possible realzations of Lie-admiesible algebras. Ll algebees € ore Slse Le-admisibla because a, b], = 212, bl,. However, there xis a large number of Lie-admisiblealgebras'which ae not Lio. ‘Thus the Lis-admissible algebres constitu a gororaization of the Ls algebras 3. Jorden Algebras. They are algobras J over F charactrized by the jaws. ab ~ ba = 0, (6a) (92, 0,4] = 0. (eo) + a sdional prope (Gand = a(n) = (ad}e~ 206), ‘ean be proved tbe a consaquence of iw 2)Direct Universality of Bitkhol's Equations 6? 4. Jordan-Admissible Algebras. They are algebras U over F such that the attached algebra U>, which is tha same vector space as U ‘equipped with the pracuct 2. B)y = Hod + 60), 0 is 2 Jordan algebra. Again, associative and Jordan algebras are Jordan-admissible, but the inverse dass not necessary hold, Ths the Jordan-admissiblo algebras are a bons fide generalzction of the Jon algebras. 5. Alternative Algebras. They ace algebras U over F verifying the laws 29D = (0b), ba? = (bapa ® For recent mathematica! and physical studies of these algebras, as well 18 for an extensive bibliography, we relor the interested reader to the Proceedings of the second (1875) and third workshops (1981) on Lior ‘amissible formulations. In physics, the abstract oloments 0,0, ... ara realized via specific uentites, such a9 functions (a) on T*td or operators A on Hilbert spoce 2; the fied is usually assumed to have characteristic coro (em. the fila oF reat numbers orthe field € of complex numbers) and the absiact Droduct ab assumes an explicit form depending on the seiacted relation the elements. Different realizations ofthe produ then yield genaraly sifferent algebras. An important (and often overlooked) point Is that all possible realizations ofthe product must verify laws (1) an (2) to quality 85 the product of an algebra {As an example, the product 0) 910 Boy “ 's_9 fully acceptable classical realization of an algebra of functions on Tema because it verti laws (1) and (2) Le, As(B+0)=AcBHALC, (108) A+B) -C HAH BRC, (108) wed (106) To have a more specific algebra, sultable integrability conditions must be Imposed on the tensor C™. For instance, for the product 4 = B to charac terize @ Lie algebra, the tonsor C™ must very integiabilty conettions (ara), Consider now the autonomous or semi-autonomous Bikhot’s eque tions (4.23) or (4.2.5), The brackets [4, B]* of the time evolution of functions (a) on 7#iM cam on fist, varity laws (1) and (2) and second, are Lie, We ean then say thatthe ‘autonomous end semi-autonomous Birkhoffe equations possess e con, ‘stent algebraic structure and that uch stucture tus out to be that of Ue algebras, Bw Ate) es CO ay © Under property (10) law (2) (a8 wall as thoss of footnote 47) fs ivy sea70 Bitkhot’s Equations This important algebraic property Is lost for the non-autonomous Birkhot's equatons (4.2.6). In fact time evolution (11) now takes the form 28 orwa(®t a of) =e Itis easy to seo thatthe “product” A © B does not characterize an algebra ‘because violates the right sistibutive and scalar laws Ata) = Hace 2) Ac(B+C)#ASBH ALC, (13a) (A+ B)eC= ACH BHC, (130) (o0Aye 8 Arb) ea, (13) ‘As a result, the nonoutonomous Bikhot’s equations do not have an algebraie stricture inthe senae thatthe brackets of their time evalusion doinot quali asthe produc ofan algebra, Note that tis nat a breakdown Of Lie algebras but, more profoundly, tha breakdown ofthe very definition oF algebras ‘The occurrence is perhaps the most significant difference between the (nonsufonomous) Hamicon's and Sitkootr’s equavons. In Teck, Ue former have @ conslstent Lie algebre structure, ‘Rs is familar from the analysts of Sections 42 and 4.5, the occurrence crigiates trom the exolict time-depandonce of the Ay functions. In Iiiguingly, the occurence provides an algebraic motivation for the con. temporary semi-autonomous form of eantacttwo-forms (that is, without fn explicit time dopendonce inthe atached symplectic stiucture "A method has been presanted in Section 4-5 (soo Equations (4.5:95)) {or tanstonming the nonautonomaus (nonaigebraic) time evolution (12) Into the semi-autonomous (algebraic) form (11). Regretadly, however the wenstormetion does not alow tho funetion & to represent the total tenorgy This can be easly seen by nating that, in general tho energy does ‘not depend explicit on time, awhile the nonpotential frees ean have such 3 dopendonce. In this caso, he only possible representation is the non: ‘utonomous one with nanalgebraie tne evolution (12). “Tis ilustate the statament of Section 55 tothe effect that the direct physical meaning of the Bithotfian function, and the algebratc charactor ‘of the time evolution aren general, mutually Incompatible"? Chart 4.2 Algebraic Signficence of Isotopic and Genotopie Trans formations Let Ube on algebra with eloments 2 yc... overa field Fof characteristic verifying the set of axioms 100) KAU Zoeee a “2° This propany ws idetifd by Sans (18786). In pening tthe Le-sdnisiblegneralizton of knot (and of Hasitor's) ty breve See nti veepect Chan 47 Final, nate pie th beckaovn ofthe algbre charac of nw (12), the tenor" mitons (449) To strech nero er, on ot the tonsa nos In races (12) once ofDireet Universality of Bikof's Equations 71 where ab is the product. Construct the new algebra U* which ie the same vector space as U but eauipped with the now product 22d (00) a hare ci @ fixed element of U. U* sealed an iotopie extension or more simply. an isotope of U, whan the new product 2+ b, besides preserving the distributive and sealor laws, algo verfies the identities of Ue. 1906) =0. @ The isotopy is called regular (singuler) when the element ¢ is (is not) inverble. A regular isotopy is invertible in the eanse that 31D (acjen1b = ob, 6 More generally, wo shell define isotopy as any transformation ofthe ‘Product ofan sigebra via elements of the algebra isol andor of the fed Which preserves 1) the original algebra es Vector field. 2) th distributive {and scalar laws, and 3) the identities of the original algebre ‘As a simple example, let af be the associative algabra of matrices A, 0, 6, owes tho field eof real umber, equlpped with tne Conven tional product af mavices AB. Let C be an element of The vanstorme- tion of tho product AB Ae B= ACE 6) {or all A, 8 ¢ of and fixed C charactorizes an isotope 2 of the essocietive algebra of (Samill (19784) because the new product A + 8 ie stil ‘seociatve. In this case thore le no need to spacty the association (AC)B oF A(CB) because, trom the associativity law, (AC) ~ ACB) As an example of nonessociatve isotopy, lst L be 9 Lie aigabra of matrices A.C... over Rand product (A.Bly = AB ~ 8A, © Let C be also an clement of Z, Then the transformation ofthe product (A.Bly = AB ~ BA-+ 1A, Ble = ACB ~ BCA o characterizes an isotope L* of the Lie algebra L. ’An example of algebraic isotopy in Newtonian mechanics is given by the transition from the conventional tothe generalized Poisson brackets (Gentil 19780) 24 a8 2 tatiy= Ber 2 og, toein tn one (0) = (oq)-* = (82 - = (@.0,00"= 53 =( a 1 Sever! possiies ao concave, he fet through tho assumption of sn ‘sscition diferent than tat of Equation (2), Lea" = act) Oterposabetia fre gion by combinations ofthe type a*b = (ab * bidsy ie ona feed Semon thr of Iho alates, oF oF he fi, ofboth,‘72. Bikhol's Equations Indeed, the generalized Lio tensor can be decomposed into (0) = Ae(a)o™ ® and therefore obtained as 2 maaification of the original tensor va elements Pa) of tho algebra, Fecal thatthe coveriant tensors, 2nd Oar the ggometic tensors of corresponding general first-order syetams. When both the original and the Final systems are sef-adjoin. the reservation of the Lie algebra is tneured by Theorem 44.2 The elgebrate significance ofthe self-adjoint isotopic wensformations fog d! + Flt Mag 8 O-> (AGC 8 + TAME Nealan = 0 (10) 1s therefore that of characterizing 2 regule Lie algebra icotopy. However, isotopy (10) isthe basis ofthe generalization of Hamilton's into Bukhott's tequstiona. In thie way we reach the following result Proposition 1. The generalization of Hamitn’s equations into the {autonomous of semiautonomous®*) Birkho's equations [ant 222] =o fasan 262] <0 on is the analtic counterpart of the algebraic notion of Lie isotopy. ‘As we shall seein Chapter 6, Proposition 1 turns out to be crucial for the identification of @ possible quantum mechenical inage of Birkhofts ‘equations. "aftr identifying the analytic meaning of isotepy, our next task is tO identify its mesning for aymmoiies, fis integrals, and conservation las, Recall from Chart A.10 that wo Lie algebras L and L* are eeled isotopically Felated. when thay are eymmetty slgebras of two isotopically related Lagrangians leading to the same frst intaprale (or conservation laws) ‘The algebras L and L* are generally nonisomorphic; but (2) they have the fame dimension r; (2) they are defined on the same cartier space (the Ceonfiguration er phase space); and (c) they coincide as vector spaces: ‘thats the generators of tho two algebras.are the same, The only possibly for the two algebras L and L* to be generally non-isomorphie therefore ‘Secure when the products are diferent. Their isotopic relationship in the Slgebraie enco intoduced inthis charts then consequential “An example is useful here, Consider Lagrangian (3) of Chart .10, ie = af(onk2 + my? + nda) ~ (hat + hy? + fa2y). (12) It possesses the $0(3) symmetry algebra with conserved generators Mn teX MP, FRO kemye (18) and commute (MMI =i, My, MM, My M= M14) Consider now the isotopically mapped Lagrangian (4) of Chart A10, her limit = m2 + may ~ (ex ~ by? * kaQ) (18) 2 Wo excide the monauronomous case bacaoe of tho sso he alge inthe time evoition (Chee 1)Direct Universality of Birkholl's Equations 73 breaks the $03) symmetry and possesses insteed Lorente symmetry '$0*(3) = $0(2.1) isotopically related to $O(3): that i, SO(2.7) i the symmetry of the new Lagrangian * which febds, via Noothor’s theorem, 10 the conservation of the generators of SO(3), Now, the carrer spaco (the space of the Cateslon coordinetes x,y,z and moments p,.Py..) hae remained unchanged by construction, nd the Lorentz algebra’ $02.1), to be consistently defined for the case at hand, must be defined in tems of the generator of SO(3) (that i, vie the angufer momentum components). This. if possible if and. only. i $0(2.1) fe realized via an isatopy of the product of §0(2). ‘A study of the case (Santi (19782))indicats that ascution exists, and Its gen by the commutation rules of SO(2.1) (MyM) My, (MM) =~, UO, MJ = M, (18) defined via the Lie isotopy 5010) 0 = Mom BML 80% = sor2s) + ri, ay or (a7) The reader can now see the nontrivial implications of isotopic general leatione of oduct in regard to Lies theory, a6 wall a the need for 2 suitable reformulation of the theory itself. In fac, the nooded broader ‘theory must permit the formulation of a Lie group, say SO(2.1), in terme fof the gonazators, the base manifold, and the parameters of @ generally ‘onisomorphic group. say SO(3). This is not readily permitted by the fvaileble conventional formulations of Li's theory, a8 we shall seein the ‘hans of Chapter 8. [As stressed in Section 4.4, tho self-adjoint isotopic transformations ae only part of the tansformations permitted by the conditions of self ‘adjointness. A second important class's given by the sell-adjint gono- topic wansformetions. The algebraic meaning of the later tensformations Is the following. Let Ube an algebra over» field F verifying axioms (1), where a6 is the product. Construct anew algebra U* which i the same véctor space as ‘eavipped wih ane of the folowing new products 2 Hb = (aepb, a(cb), (80) + afc), ete (13) 23s wa shall ein Chapter 5, the Lornts group $0(21) nat only must bo Gefined in tegns of me angler momentum component ut th passer ans ‘these ofthe group of otatons, thts the Ear angles.74 Birkhol's Equations ‘where o,o,.. afe fixed slements of U. U” ie called a genotopie extension ‘orsimply a ganotope of U whan the new product a # b verfis the dist butive and sealer laws, but violates axioms (1) and verifies instead © dliferent set of axioms Ha # 6) ken, (9) ‘When the genotopy is inverible in the sense of Equation (4), iti called regular; otherwise tis called singular. ‘More generaliy, we define a6 genotopy any transformation of the product of an algebra via elements ofthe algebra tel and/or of tho el lahieh (1) preserves the orginal algebra a5 vector field (2) varies the eG) €8 (Chart 12.1). A set of local coorcinates wil be. denoted by = (et). For a point me U = M, lm) = (°, .. #"), where (ch is considered ae 2 map fom Mt Definition 1. A tangent vector X,, at @ point m « Ms linear function from (0) (ihe space of @ functiSns definad in tne neighbornood of m) 10 R sitisying the rues Xplaf + B0) = 2X, (0) + IX (9). Xqifa) = Hm) X ig) + atm) % yl 0 WgeexM; 4 Bek may eanault for inanes, Dedesker (1981), Stenberg (1968), Abvebarr bd IMarssen (1987) Loans ar Sted (1988) outa (1970), Spivak (1070-73) Edelan (1972), Gullemin snd Stormborg (1977), Thing (1878), Amol (1978) and Shaye (1970)78. Bickhos Equations ‘A numbor of atomative definitions of tangent vectors exis inthe Inereture, ‘The following (equivalent) definition is relovant fr our objectives, Definition 2. A tangent vector X. at m ¢ Mis an equivalence cass of ‘curves which are tangent to each other at m, ie Xq™ [rly and 7, i and only if a a gerne @ (7000 Jno s-0 By combining Definitions 1 and 2, we ean interpret X,, a8 an operator performing the mapping (Mf) "> B for which Kaif ‘Tho tangent space TM ot m ¢ M is the vector space of all tangent vectors at m. The 2n-dimensionsl manifold TM = Urey TaiMf can be equiped with o'@~ stucture ina natural way, and condtiite the tangent bundle over ff, ihere the fiber at each point isthe tangant space st that The cotangent space TzM st m ie the dual of TaM, and itis the space of ail linear functionals on 7M. The. Zn-dimensionel manifold THM = {ney TM can again be eqUlpped with 9 €" stuctute, and fi called the Gotongient bundle. 7 65 (MA), we can define an element of 7% M, called diferental of ¢ atm, by one = ens “ ine hii ose utente ‘the dual of dx’|,, is then give by (in the sense of Definition 1) a/ax,. Uae Pind 27a neta ey teed, x=ae 2, ® neh nfo tt pont sundae. Wo bo hoe My %a) @ OX ER Kall) Bea remnun Le Fost 9 me fm 2 mano Hf aro ant Hettinger eae TEXn) = Yee — Yem(9) * XplFeog), —gete(M). (7) [A tensor of convavaviant order r and covariant order s, #2, of type (©) (Char 12.1), on a vector space Vie gn (7 + s)-multingsr map times time a Bae x Fe eV 6 where V* isthe dual of V Whon Vis identified with 7,., we have the tensor bunle TM) = Loa 77g whee Tie sat ofl (8) tensors over 7 leary, F3(MM) = TM, To = Tem, end T3(M) = R.Direct University of Birkhof's Equations 79 ‘An (7 8)-tensor fed is 2 ¢° map iM TM), Bm) eT Th). ® Most important for our ancysis are the (1.0)-tontr fields on #, called Vector folds, They characterize 2 map trom Mf to TM. Equally important ‘are the (0, 1)-tonsor fields on M called one-forms. They characterize 8 ‘mapping thom Mt to Th “The set of all p-forms on M (Section |.1.2) is denoted by A&(M) Hence 2 « Aa(i) mean a(m) @ A&(Ts Mi). Equations (8) therefore sive {2 Tocal expression for vector fais X ahd one-forms 0. In pantculoe 0) = 0) £400) (10) A curve y:/ > fat m is an integral curve of » vector field X at m if X(y(0) = T,{t 1) forall 1, where T, i the tangent map of y. Suppose that (U. @) is @'chart at m, em) = (x5, ag). and (o> 7) (et). (0), then x is an integral Suevs of at mf ond only i (4°, 20(0) satistos the system of first-order ordinary diferntial equations = HO), 20) = ay In this way we reach the fist contact withthe analysis ofthis volume, ‘As worked out in Volume |” and reviewed in Section 4.1, Newtonian systems can always be writen in the normal form 21), =a, am (ry) «2 via arbitrary presritions forthe charactaaation of 7 new variables y. The {quanttes =" have besn refered to in tha mein text ae vector fells, mercly to xpress the Newtonian ‘character of transforming. 26 contravaiant vectors, We now lean that they can be Interpreted ae vector ils In the ‘somatic sen a3) Equation (10) {is then the ditferal equation comesponding to the {geometrical definition of) vector field = We should keepin mind that, whether the Newronian or the geometrical dotinition is used, those vector fields characterize. autonomous. nM onsorvatve systems. The extension to the ronautonomove cassis ee evident. Autonomous conservative eystoms are, of oovise, not excluded, 85 0 partculer case of tis broader prysical context, From the existonce theory of ordinary differential equations (Section 1.11) we can soe that, at every point m = M, there oxists'& unique integral ‘cue of a vector fiald X at m. For all m =U © Wf, these integral curves Sefine a focst, one-parameter pseudagraup of vansformations on M which bbecomes a focal one-parameter group it the interval of me in which its defined is independent of m,or tis the whale real ine. The vector fied X ‘then acquiras the mesning of generator of this psoudogrovP. Suppose that a @-map F: Mf-> MV given. Then, @ tural map from TR{W) to T3(H) exists called she pull-back of F and given by FE TYAN) + TM), FH), we ARM), (18) CFV Kero Koy) * AEM )TE hs 2 TE)80 Birkhot’s Equations 'F* maps p-forms into p-forms, and itis an (slgebra) homomorphism with respect to the extoror product (also called wedge product in aiferntial ‘geomeuy}. Notice in ‘paricular that, fF is a lifeomorphiem, then yey Given two vector fies X and ¥ on M, one can define a third vector field vi the associative composition law CON =X), We ean, a3) ‘The (nonassociatve} Lie ro De = xY = ve (18) then also defines a vector field which can be written in local coordinates ax) a 1) 2 an We see in this way thatthe sat ofall vector fis on a manifold forms a ie atgebra under brackets (16). {Lat bo a vector field on M end b an (7, 8)-tensor fed. Also, lt F, be the local one-parameter peeudogroup given by the Integral curves of for t 1. Then F,, U > M maps every point m of U ante tho point of M Iving in the integral curve through m a ‘The Lie derivative ofthe tensor field b with respect to the veétor field X is defined by im F262) em) — Bom) L,b(m) : (8) I the tonsor bisa scalar we have Lyf of, wee; (9) Fis 9 vector field ¥ we have LY GM: (20) ‘and forthe ease of one-forms we have (locally) 2h 6 ae wat (re 2 04, 2 en (800) = (900) + 80% 1) @) Finally, by using the global formulation of the exterior dvivative (Section 112), we heve Aya = kyle), Yee APM). (23) ‘The inner product of a voctor fad X and a p-form a, denoted by i OF bby X Jar (Section 4.3) is the (a = 1)-form ok) verifying the properties FR OF mf) = ga) A B+ (1982.8 (eB). » (24) (28)Direct Universality of Birkhof's Equations 81 In this way we reach te following Important property of the Lie drivative bea = ila) + liga) lode + aX 10), Vas Aen (26) “To establish a lnk with the weatment of this volume, iis significant to Idntiy the explicit form of these properves in local coordinates. From the definitions given above, we have. Gal = 20061), were, @ Thus ° e oe vend, 28) a ayleheot noo, and Gad 2 AaDY #2507. (2) jya XS = 2a Xt (20) ‘This yields the expression of the inner product in local coordinates of Seation 4.3, Equations (43) and (4.317), i, 210g = $1404 2 doh oP tdes, 19) 310, = 45130, oes (a1) (32) hhave play a rather crucial role in the mein txt forthe constuction of @ Hamitonian and a Bitkhoffian rapresentation, respectively, In this case, the forms 1, and 0; are the fundamental symplectic form, and a generat (ut local ad execs) symplectic form, respactvey. Chart 4.6 Global Treatment of Hamilton's Equations Lot M boan n-dimensional, ¢° manifold with local coordinates (q°,....) (the configuration spece of Newtonian Mechanics), and lot TM and 7¥At be its tangent and cotangent bundles, respectively. The bundles TM and T*td are customarily used for the cherecteraation of the Lagrangian and Hamitenian formulations of mechanics, respectively. Weare hore interested in tho later eae. ‘point of Tf consis of @ couple (m,/), where m is @ point of Mf, ‘ane / belongs to T*M. The projection mop (Chart 12.2) 2: T°M—> Mt ‘maps the wholo fiber (m, TM) onto m. As a 2n-dimensional monifol, Te can bo equipped with"s fundamental one-form, called a canonical orm, 98 follows, Consider the mapping OTM TT, BeANTEM), o82 Bickhof's Equations defined by the following properties. Let Xian be an arbitrary tangent vector to FHM at (mf). Then 8(, DX 9) = ATE 9D) @ ‘hare Te: T(7*M) Tid is the tangont map of tho projection x. Let (2) = (a, 2) ba local coordinates for T*M. The vector fald Xp, then seguites the Focal form ann aten & 91002 © . tha) =H fy : nA ® thn) “BA © wim on de Rio ® N= 0.0). ‘where @ is the notation generally used in the Fteature of differential ‘geomety/®* while AP is that used In the Inerature of the Birkhoffian techanies, Clealy, the two-form = dO ~ 6p, 4a! = $y da" dot = ORE Open Van tom = (Sree yr) a is nowhere degenerate and closad. It isthe familar fundamental sym plectc form. The 2n- T°M for which one ean define 2°02 = 2. The form 0, = 0-08 0 dt (13) for some function @ on R * THA can then be proved to be an exact ceantact form, ‘A semi-autonomous globolly Birkhoffien vector fad is then any vector field % on (Re Tei, 9) verifying the propertios 14, = ¥ 10, =0, (14) dt() = 1, (ab) ‘This is the desiad global treatment of Equations (5). Indeed, according 10 Equation (18), the Biehotfian hes an explicit time dependence, but the Substcture & does not possess, n oral coordinates, such a dependence. Cleary. Equation (14) above and Equation (#9) ofthe precedina chart, aro equivalent. No distincvon can be mace therefore between semi= ‘autonomous Birkhotfian vector fais and the nonautonomous Hamitonian ‘ones at the coorcinatetre level Wie consider now the general cago of Equation (6), which inchudes all preceding cases, whether Hemitonian of Bikhotfian, and inttoduce fan aritrary one-form A on R* T°M subject to the coneltion that the ‘esaciated twa-form vio exterior derivative a= ah 18) is of maximal rank. We shall call a general, globel, ikhoian vector fold ‘any nonautnomous vector field Xon & * 7° veriying the properties hah =% Jah =o, (160) ati = (180) ‘The equstions sbove proyide the desied global treatment of Equetion (6). Indood, the one-form A can be written in local coordinates A= Aidt and charecterizes precisely the intogrand of the variational principle for Systems (1) (Section 4.2). The inner product of the vector Held withthe textorior derivative of form (15) then yields precisely Birkhot’s equations (6) in our unified notation | = Alt, 9)08" ~ Bit ade, B= Ry, (17) MpB)IB"= 0, WR OAZ 2, 3 (GO), (180) om oR, ee (18) PRUs Be (186) +3) The structure & % T*fd has beon introduced for definition (16) mainly tokeep in touch with the physica insight, thet is, to associate Ue with ‘On more general geometric grouncl, suon en essocition is lest, in the88 Dickhot’s Equations once that the equations ean be defined In an atbitvary (2n + 1)-dimen- jonal marifold 1 equipped with closed snd exact two-orm of maximal Fank. Time would be associated then with the space of the (null) co- Seterminant of maxima rank, a¢ we shall soo betor in Section 6.3. ‘We shall now study the nonconservative nature of systams (1). The problem consists of identifying a geometric charactorzation of the anaray {ate of variation in time. This i ackioved through the Lie derivative (Chart 444) In general, the Lie derivative of the Brkhotfian (8) with respect to¢ slobally Birkhofian vector field is givan, in local coordinates, by 28 OMe , 38 ae or * or Now construct a Brkhotin representation of systems (1) according to Corotny 45:18, whereby the Bichottian i the foal energy BUt a) = Exot 2) = Eygelt, 6, B) = Tl) + Ute), (20). that Is, the Blbotfian ig the Hamitonin of the maximel self-edoint Subsystems of systems (1). Thie energy le necessary nonconserved wing tothe provenco of nonconsrvalva forces, The desea geometie {haracterzatian of the energy rate of vetlatian in tme ie then given bythe Particularization of rule (19) 48 (19) . Fs ao, 9) Be 4 Ob Balt a) = LaEag = Set oes 9) Be 4 Mes, (any OE as 4 @ oe ‘An instructive exercise for the intersted reader isto vary that law (21) may provide e description ofthe energy rete of varation.®” However, the reader should keep in mind that low (21) does not possess a Lie algebra ‘Stucture for the general non-autonomous case (Char 1) Evidently, nonconservation faw (21) admits, es 2 particular case ‘conservation law El ExFior * (Exe Euaiy * © 2) In this caso, the law does possess a Lie algebra structure, but we are ealing with trly pariculer Newtonian systems (the autonomous, con Sorvative essentially self-adjoint systems indirect only opresantations) ‘We consider now the peculiar aspect of the Birkholfian realizations of contact two-forms mentioned in tho text that is, the expen time do pendence of the symplectic substructure, This occumence creat hhumaber of technical problems, such as 1) the region of definition of the {wo-form; 2) the applicable version of the Poincaré lorama, and 2) the proper formulation ‘of the transformation theory. We consider here problems 1) and 2). Problom 2) is studied in the next chapter. Consider contact structure (18) inreazation (18). It israther natural to think of a star-shaped rogion inthe variables 2” at each fixed valu of ime, but then one nt isa mechanism whereby a time varies, different regions ‘of time are smoothly connected. Also, 98 recalled In Section 4.2, a star-shaped region does not necessary remain this ‘yp Undar an arbitrary tanstormation, 57 This canbe vee, for instance, by using the Bikhoian representation of example 42Direot Universality of Birkho#'s Equations 89 In order to overcome these ditficultos, Sarlet ond Cantrin (19788) Introduced the notion of & "region deformable to 9 curve”™—that is, one ‘hich a) is topologically equivalent to @ star-shaped region, (b) allows {2 smooth connection between rogions at ferent values of time, and (c) preserves iis topological cheracter under arbitrary iffeomorphisme (ransformations) Note from the outset that the approach by Savlet and Cantijn is a palural generalization ofthe” doformabilty toa point” by Flanders (1963) ‘We shall thorefore take thie opportunity to review the formulation of t rect and ofthe converse of the Poincaré Lemma which apply to structure (18). This provides an altomative approach to that by Lovelock and Fund (1875) roviewod in Section 11.2 and which, a8 is now familiar, is based fn the notion af star-shaped ragion Let © denate an open subset of fe, and F*(0) the sot ofall * p-forms (Geetion 11.2) on 0. An eiement # of F9(0) aesigns ta avery 2.@ 08 ar alternating mapping AP APR) Ay, agl@ldats A A ae (a) Let be a subset of R= Re, snd put Bpm eRn| (a) €) 2) Definition 1. i smoothly deformable to 2 curve (monotonieslly increasing inthe ¢-ditection) ia family of mappings gee Spek, = 10.4) (ea) Foue.a) bee - ‘vats such that () o,(1, 2) = 2. 9,(0 2) = 29, forall @¢ Z, and whers Si, ag) i fad 8 (i) the ap @: / => 48 of clase e™ with ‘apoat to all argumens Cleary, the propeny of being smoothly deformable to 8 curve is pre served by all images 9" of 9 undor class , invertible transformations ‘Also, regions at diferent values of time are smoothly connected, Finely, ‘the topolagieal equivalence of Definition 7 with the notion of star-shape Is alo ensured, ‘Starting from a family of p-forms AP « F#(2,) given by Ap = Ag) ft O00 no neo (28) foreach esuch het, # », where, (8) oe given 6" functions on $9, we can define parametiep-form on, st? © FB), by #60) = AHO). Meo) e4. 2 As a staightlorward extension of the definitions of Section 11.2, we hove 5: 01) > F018), (23) that is the exterior derivative of parametiic p-toms can be wstton (actually, can be defined by) (at) a) = (dAey(@), 9)90 Birkhod's Equations whore d is the exterior dorvative in FP(E,). A paremetic pform of? is ‘exact when 2 paramottic (p ~ 1) form a!” exist, called primitive form, Such that iP = Bole 0) ‘Similarly, 24° ie closed if and only it Sao = 0, oy ‘The following property is 2 simple, direct generalization of Lemma 424 Lemma (Direct Poincaré Lemma). Every exact peramatle p-form is closed. “The proof of the following property is. on the contrary, nontvial, For bovily, we Toler the intrested reader to Serlet and Cantrjn (le. ci) Lemma 2 (Converse of the Poincaré Lemma). Let # be 9 subset of x fe that is smaothly deformable to a curve. Let sf” be 2 parametic form of @ that fs closed. Then sto fe exect on 8. In conclusion, the existence theorems of the Inverse Problem studied inthis volume, percularly these for Birkhoffian ‘epresontaions, ean be subjected 19 9 dua! approach. One ean fist use conventional star-shaped Fegions in (2n + 1) dimension for ® contact approach to the forms con- sidered. This isthe case, for instance of Corollary 4.2.12. Alternatively, we fan consider the forms at a fixed value of time. In this latter case the patamatic approach utined in this chart apple, This is the case of ‘Theorem 4.5.1. Thisattor approach will be tacitly implemented throughout ‘our analysis whenover considering symplectic stactures with an expla time dependance, Chart 4.7 Lie-Admissble/Symplectic- Admissible Generalization of Birkhof's Equations for Nonlocal Nonpotental Systeme Inthe toxt of this chapter we established the universality ofthe Lie algebras {and ofthe symplectic (or contect) geomet for local Newtonian systems. ‘A'few words on the limitations of these mathematical tools in physics are nom in order to prevent & possible expectation of their terminal mmethodologicel charecter. Stated explicitly, after having ideatiied rather substantial eapablties,¢isimportant wo point out that the Lie elgebres end the symplectic gsometiy dont provide the final frmlation af mechanics, (On the contrary, they characterize only one stage of an ever continuing process of mathematical and physical advances, Consider the problem of interactions, The olfetiveness algebra and of the symplectic geometry forthe treatment oft magnetic Interactions fs well-known. In Example 41 wo shall show thet the formulations considered apply to the charactorization of the electio- ‘magnetic intoactions, not only inthe conventional (Hamitonian) form, but alsin their most general possible (Birkhatfian) form. This efectiveness persists for tho more goneral interactions of contemporary physics, such 85 tho unified gauge theories of weak and electromagnetic Interactions. All these interactions, whethor Newtonian or quantum file thoortial, ‘ate of local/ifferental and potental/solt-adjoint type. In foc, all theseDirect Universality of Biehot's Equations 91 Interactions are characterized by 8 conventional, local, Lagrangian func- tion oF operatr-valued eistabutton ue = Livy * Le ‘Rocent studios®® have identified 2 humBGt of indutficiences of these inorecions in several branches of physics. Infact, the systoms of New- tanian Mechanics are, more properly, of the nonlocal tyne (Section 41), inthe sense thot they demand integro-dfferomial equations to represent the interactions a all pols of @ surface volume. The systems of Statistica! ‘Mechanics, whother-lassieal or quantum mechanical, are also nonlocal ‘whenever the extended character ofthe constituents is taken into eccount, together with theirineastie collisions. Along quite si Hines, (of Particle Mechanics are olso nonlocal in their more adequate treatment. ‘This fs the ease in particular for the strong interactions, because of the nead for mutual penetration of tha wave packets of partcles.°° ‘When tho Lie algebra snd the symplectic geomoty are considered in this context, they emerge possessing rathor brocise limitations, In fact the symplectic geometry is io the final analysis, a. local/diferenti ‘Geometry, that is. @ geometry which, when realized ina local chart ‘tdmits ordinary (or parti) ilfrentsh equations. As a result, no possibilty is known at thie time for an effeote footmant of nonlocal systems via the symplectic yummy i is canton formulation. ‘we can therefore say that the Le algebra and the symplectic (or contact) geomery, rather than providing the ultimate formulations of mechani, Drovide Instead a mere approximation of the lacalinonpotental type, with {he understanding thet more goneral algebraic and geometric structures are expected to enst for nonlocalinanpotential treatments in thie chart ve raviow the mein ideas of tho possibilty of genertizing ‘Bdmisibleaigobras and the symp ‘The most general orm of unconstvained Newtonian eystes in Euclidean spoce known at tis time is given by the so-called integro-diferentil, Varlationally non-soll-edjoit systems. These ara systems with 2 super- postion of locsi/aitferntial and nonlocal/integral forces. which are ‘erivable and nonderivable from a potential. By using a self-evident ‘notation, the systems can be written®® leit =[ff ohne aed]

You might also like

• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  From Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  Rating: 4 out of 5 stars

  4/5 (5834)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  From Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brene Brown

  Rating: 4 out of 5 stars

  4/5 (1093)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  From Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  Rating: 4.5 out of 5 stars

  4.5/5 (852)
• • Magazines
  • Podcasts
  • Sheet music
• Principles: Life and Work

  

  From Everand

  Principles: Life and Work

  Ray Dalio

  Rating: 4 out of 5 stars

  4/5 (612)
• The Glass Castle: A Memoir

  

  From Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  Rating: 4.5 out of 5 stars

  4.5/5 (1720)
• Grit: The Power of Passion and Perseverance

  

  From Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  Rating: 4 out of 5 stars

  4/5 (590)
• Sing, Unburied, Sing: A Novel

  

  From Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  Rating: 4 out of 5 stars

  4/5 (1194)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  From Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  Rating: 4 out of 5 stars

  4/5 (903)
• Shoe Dog: A Memoir by the Creator of Nike

  

  From Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  Rating: 4.5 out of 5 stars

  4.5/5 (541)
• The Perks of Being a Wallflower

  

  From Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  Rating: 4.5 out of 5 stars

  4.5/5 (2106)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  From Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  Rating: 4.5 out of 5 stars

  4.5/5 (350)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  From Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  Rating: 4.5 out of 5 stars

  4.5/5 (474)
• Bad Feminist: Essays

  

  From Everand

  Bad Feminist: Essays

  Roxane Gay

  Rating: 4 out of 5 stars

  4/5 (1029)
• The Outsider: A Novel

  

  From Everand

  The Outsider: A Novel

  Stephen King

  Rating: 4 out of 5 stars

  4/5 (1872)
• Her Body and Other Parties: Stories

  

  From Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  Rating: 4 out of 5 stars

  4/5 (824)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  From Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  Rating: 4.5 out of 5 stars

  4.5/5 (122)
• The Emperor of All Maladies: A Biography of Cancer

  

  From Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  Rating: 4.5 out of 5 stars

  4.5/5 (271)
• Angela's Ashes: A Memoir

  

  From Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  Rating: 4.5 out of 5 stars

  4.5/5 (443)
• Brooklyn: A Novel

  

  From Everand

  Brooklyn: A Novel

  Colm Tóibín

  Rating: 3.5 out of 5 stars

  3.5/5 (1948)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  From Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  Rating: 3.5 out of 5 stars

  3.5/5 (405)
• A Man Called Ove: A Novel

  

  From Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  Rating: 4.5 out of 5 stars

  4.5/5 (4772)
• Steve Jobs

  

  From Everand

  Steve Jobs

  Walter Isaacson

  Rating: 4.5 out of 5 stars

  4.5/5 (809)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  From Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  Rating: 3.5 out of 5 stars

  3.5/5 (2259)
• The Art of Racing in the Rain: A Novel

  

  From Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  Rating: 4 out of 5 stars

  4/5 (4214)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  From Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  Rating: 4 out of 5 stars

  4/5 (98)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  From Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  Rating: 4.5 out of 5 stars

  4.5/5 (266)
• A Tree Grows in Brooklyn

  

  From Everand

  A Tree Grows in Brooklyn

  Betty Smith

  Rating: 4.5 out of 5 stars

  4.5/5 (1930)
• Yes Please

  

  From Everand

  Yes Please

  Amy Poehler

  Rating: 4 out of 5 stars

  4/5 (1905)
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  From Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  Rating: 3.5 out of 5 stars

  3.5/5 (231)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  From Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  Rating: 4.5 out of 5 stars

  4.5/5 (234)
• The Woman in Cabin 10

  

  From Everand

  The Woman in Cabin 10

  Ruth Ware

  Rating: 3.5 out of 5 stars

  3.5/5 (2526)
• Wolf Hall: A Novel

  

  From Everand

  Wolf Hall: A Novel

  Hilary Mantel

  Rating: 4 out of 5 stars

  4/5 (3973)
• Fear: Trump in the White House

  

  From Everand

  Fear: Trump in the White House

  Bob Woodward

  Rating: 3.5 out of 5 stars

  3.5/5 (738)
• John Adams

  

  From Everand

  John Adams

  David McCullough

  Rating: 4.5 out of 5 stars

  4.5/5 (2410)
• On Fire: The (Burning) Case for a Green New Deal

  

  From Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  Rating: 4 out of 5 stars

  4/5 (74)
• The Light Between Oceans: A Novel

  

  From Everand

  The Light Between Oceans: A Novel

  M.L. Stedman

  Rating: 4.5 out of 5 stars

  4.5/5 (789)
• Manhattan Beach: A Novel

  

  From Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  Rating: 3.5 out of 5 stars

  3.5/5 (880)
• The Constant Gardener: A Novel

  

  From Everand

  The Constant Gardener: A Novel

  John le Carré

  Rating: 3.5 out of 5 stars

  3.5/5 (104)
• The Unwinding: An Inner History of the New America

  

  From Everand

  The Unwinding: An Inner History of the New America

  George Packer

  Rating: 4 out of 5 stars

  4/5 (45)
• Rise of ISIS: A Threat We Can't Ignore

  

  From Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  Rating: 3.5 out of 5 stars

  3.5/5 (137)
• Little Women

  

  From Everand

  Little Women

  Louisa May Alcott

  Rating: 4 out of 5 stars

  4/5 (105)
• Iraq Security Forces

  

  Document1 page

  Iraq Security Forces

  Fatima Om Salman

  No ratings yet
• Eldy Notes

  

  Document265 pages

  Eldy Notes

  Fatima Om Salman

  No ratings yet
• CM Solutions

  

  Document7 pages

  CM Solutions

  Fatima Om Salman

  No ratings yet
• Maxwell's Equations and The Principles of Electromagnetism TQW - Darksiderg PDF

  

  Document451 pages

  Maxwell's Equations and The Principles of Electromagnetism TQW - Darksiderg PDF

  Fatima Om Salman

  100% (1)
• Zhilin 2006 APM Eng

  

  Document271 pages

  Zhilin 2006 APM Eng

  Fatima Om Salman

  No ratings yet
• Lim. Yung-Kuo. Problems and Solutions On Solid State Physics, Relativity

  

  Document364 pages

  Lim. Yung-Kuo. Problems and Solutions On Solid State Physics, Relativity

  Fatima Om Salman

  No ratings yet
• A Mathematical Approach To Control System PDF

  

  Document666 pages

  A Mathematical Approach To Control System PDF

  Fatima Om Salman

  No ratings yet
• CM Lectures

  

  Document94 pages

  CM Lectures

  Fatima Om Salman

  No ratings yet
• Moullin PrinciplesOfElectromagnetism

  

  Document322 pages

  Moullin PrinciplesOfElectromagnetism

  Fatima Om Salman

  No ratings yet
• Lawden AnalyticalMechanics

  

  Document46 pages

  Lawden AnalyticalMechanics

  Fatima Om Salman

  100% (1)
• IB Physics Review

  

  Document159 pages

  IB Physics Review

  Reeja Mathew

  71% (7)
• Maxwell's Equations and The Principles of Electromagnetism TQW - Darksiderg PDF

  

  Document451 pages

  Maxwell's Equations and The Principles of Electromagnetism TQW - Darksiderg PDF

  Fatima Om Salman

  100% (1)
• Maxwell's Equations and The Principles of Electromagnetism TQW - Darksiderg PDF

  

  Document451 pages

  Maxwell's Equations and The Principles of Electromagnetism TQW - Darksiderg PDF

  Fatima Om Salman

  100% (1)