betaFeedback

WorksPhysics and Astronomy

Canonical and Canonoid Transformations of Hamiltonian Systems on qq-Cosymplectic and qq-Contact Manifolds

Abstract

qq-Cosymplectic and qq-contact structures equip presymplectic and precontact geometry with additional one-forms and a normalized family of Reeb vector fields. We study canonical and canonoid transformations relative to these data, with particular attention to their origin in constrained mechanical systems. In the qq-cosymplectic and uniform qq-contact cases, we establish local normal forms, characterize infinitesimal automorphisms, and obtain Noether-type relations between symmetries and conserved or dissipated quantities. For a one-parameter group of canonoid transformations with generator XX, we show that [X,XH][X,X_H] is Hamiltonian and determine its Hamiltonian function. In the uniform qq-contact case, simultaneous constant rescalings of the defining forms provide canonoid transformations for every Hamiltonian. The geometric constructions are illustrated by regular relativistic mass-shell constraints and by global Lagrangian descriptions of a charge in magnetic monopole fields. These examples distinguish parameters along constraint orbits from physical time and clarify the role of gauge directions. They also recover the total angular momentum of free relativistic particles and the Poincaré angular momentum of the charge–monopole system. For two monopoles, we construct a non-uniform 22-contact structure and prove a topological obstruction to a uniform structure with the prescribed gauge generators as Reeb fields. Irrationally related magnetic charges additionally obstruct the intermediate quotient that would produce a single contact manifold. This example identifies a limitation of the uniform theory and motivates its extension to non-uniform structures.

Introduction

Degenerate Lagrangians and Hamiltonian constraints lead naturally to geometric descriptions in which the relevant closed two-form has a nontrivial kernel. The foundations of this approach were developed by Dirac and by Anderson and Bergmann [15, 3, 20]. Symmetries, their generators, and their relation to constraints were subsequently studied in detail within generalized Hamiltonian dynamics [28, 26]. These results provide the background for the present work. Our purpose is to study the consequences of specifying additional geometric data on the kernel and on a complementary distribution, as occurs in qq-cosymplectic and qq-contact geometry.

On a qq-cosymplectic manifold, the closed two-form Ω\Omega defines a presymplectic structure with qq-dimensional kernel R\mathcal{R}. The closed one-forms λ1,…,λq\lambda_1,\ldots,\lambda_q determine a complementary distribution ξ=⋂iker⁡λi\xi=\bigcap_i\ker\lambda_i and a distinguished frame R1,…,RqR_1,\ldots,R_q of R\mathcal{R}, normalized by λi(Rj)=δij\lambda_i(R_j)=\delta_{ij}. Thus the structure includes a choice of transverse directions and a normalization of the characteristic directions. In the qq-contact case, each λi\lambda_i is a precontact form, with characteristic distribution ker⁡λi∩ker⁡dλi\ker\lambda_i\cap\ker d\lambda_i. The forms are related by the requirement that all dλid\lambda_i have the same kernel R\mathcal{R} and are non-degenerate on the common complement ξ\xi. Uniformity imposes the further condition dλ1=⋯=dλqd\lambda_1=\cdots=d\lambda_q.

These additional data affect the transformation theory. Preserving a presymplectic form preserves its kernel, but need not preserve a chosen complementary distribution or the individual Reeb fields. For a Hamiltonian constant along the kernel, the choice of ξ\xi selects a unique horizontal solution of the presymplectic Hamiltonian equation. Similarly, preserving one precontact form does not in general preserve an entire qq-contact structure. We therefore formulate canonical transformations, Hamiltonian vector fields, and canonoid transformations with explicit reference to the defining one-forms. The physical examples below explain how such data can arise and why their local and global properties require separate consideration.

Geometric and dynamical background

Two earlier constructions are directly relevant to the present work. Dubrovin, Giordano, Marmo and Simoni [16] construct Poisson brackets from a presymplectic form and a flat generalized connection. The qq-cosymplectic bracket used here is obtained by taking the horizontal projector determined by the closed one-forms λi\lambda_i; see Remark 2.3. Alekseevsky, Grabowski, Marmo and Michor [1] describe symplectic and Poisson structures on cotangent bundles of Lie groups and principal bundles in terms of momentum mappings, Maurer–Cartan forms, and connections. Their principal-bundle construction provides a geometric framework for the Cartan forms in our monopole examples; see Remark 3.3. These works supply the underlying bracket and bundle constructions. Our transformation theory keeps explicit the additional ordered one-forms and the normalized Reeb fields, and imposes compatibility with these data.

Canonical transformations are standard tools for comparing Hamiltonian descriptions [9, 11]. Canonoid transformations preserve the Hamiltonian form of a specified dynamics while allowing its Hamiltonian description to change; their study in symplectic geometry includes [14, 30, 12, 32]. Related constructions in cosymplectic, locally conformal symplectic, contact, and cocontact geometry connect transformation theory with symmetries, first integrals, and dissipation laws [4, 5, 6, 31, 36].

Structures defined by several one-forms have been studied in connection with integrability, thermodynamical models, and Hamiltonian dynamics [17, 23]. The qq-cosymplectic structures considered here are those of [23, 24], while the qq-contact setting is developed in [2, 21, 22, 25, 37]. Integrability, reduction, and Noether-type results are already available in these settings [23, 24, 25, 37, 18]. A broader framework allowing several two-forms with a common kernel is discussed in [18]. We use this framework when the forms arising from separate magnetic contributions must be distinguished from the form governing the total dynamics.

A related classical construction replaces the one-parameter group of time translations by a Lie group GG. This idea was already considered by J. Schwinger in his lectures on quantum field theory [33]. Hamilton–Jacobi theory in this setting was studied by Mukunda [29] and by Marmo, Morandi, and Mukunda [27]. For a Hamiltonian action of the Abelian vector group Rq\mathbb{R}^{q} with an equivariant momentum map, a regular momentum level is presymplectic, with kernel generated by the independent infinitesimal action fields. Equipping this level with closed one-forms whose restrictions to the kernel form a coframe gives a qq-cosymplectic structure. The existence of such forms globally is an additional requirement. This construction relates the present geometry to the classical theory of Hamiltonian group actions. Our transformation results concern the Abelian setting: the Reeb fields commute and generate a local Rq\mathbb{R}^{q}-action. Non-Abelian extensions require different structure equations and will not be dealt with in this paper.

Physical examples and their interpretation

Relativistic constraint dynamics. The no-interaction theorem [13] exhibits the restrictions imposed by the standard canonical world-line assumptions in a single-time relativistic particle description. Constraint formulations provide a way to accommodate interacting particles while retaining relativistic covariance [20, 8]. We consider NN independent generalized mass-shell constraints on T∗((R4)N)T^{*}((\mathbb{R}^{4})^{N}) and restrict to a regular region in which these constraints Poisson-commute. This commutativity is a stronger assumption than the first-class property alone. The constraint surface has dimension 7N7N, and the restricted symplectic form has a kernel of rank NN.

Functions χ1,…,χN\chi_{1},\ldots,\chi_{N} whose differentials are independent on this kernel determine an NN-cosymplectic structure wherever the corresponding transversality matrix is invertible. Its Reeb fields are normalized linear combinations of the constraint fields. Their parameters describe directions along constraint orbits. In the free-particle example, our normalization identifies these parameters with constant multiples of the individual proper times. For interacting systems, an interpretation in terms of proper time requires a separate analysis of the chosen constraints and clock functions.

  • A charge in magnetic monopole fields.* Magnetic monopoles provide a setting in which the topology of the configuration space affects the existence of a global potential [19, 7, 35]. For a point monopole, the magnetic two-form on Q=R3∖{0}Q = \mathbb{R}^{3}\setminus\{0\} is closed but not exact. It therefore admits no global vector potential on QQ and no global magnetic Lagrangian of the usual potential form on TQTQ. A global Lagrangian can instead be defined on TPTP, where P→QP \to Q is a principal U(1)U(1)-bundle. Its Legendre image is a constraint submanifold of T∗PT^{*}P, on which a normalized Cartan form is contact and its Reeb field generates the gauge action [7, 35]. The particle still evolves with one physical time parameter. Gauge-invariant observables descend to the magnetic phase space TQTQ, and the lifted Hamiltonian dynamics projects to the Lorentz force equation.

For two monopoles, the corresponding construction uses a principal U(1)×U(1)U(1)\times U(1)-bundle and produces an eight-dimensional constraint manifold. The normalized forms associated with the separate monopoles define a non-uniform 22-contact structure. The total Cartan form, which contains both magnetic contributions, determines the physical dynamics. There are two gauge directions and one physical time parameter.

Two different global issues occur in this example. If the magnetic charges have an irrational ratio, the one-parameter subgroup along which the total Cartan form would descend has dense orbits in the gauge tori. Its orbit space does not give the smooth seven-dimensional contact manifold obtained in the rational case [35]. The quotient by the full gauge torus remains the magnetic phase space. Independently of this irrationality condition, we prove that the topology of the monopole bundles prevents a uniform 22-contact structure whose Reeb fields are the prescribed gauge generators. The example therefore distinguishes the obstruction to an intermediate contact quotient from the obstruction to uniformity.

Results and scope

We develop the transformation theory for qq-cosymplectic and uniform qq-contact structures, keeping the additional one-forms explicit. The local normal forms and Noether-type statements are presented as adaptations of established geometric methods to these specified data. The main results concern the compatibility conditions for transformations, their infinitesimal generators, and the geometric restrictions exhibited by the physical examples.

In the qq-cosymplectic case, canonical transformations preserve both Ω\Omega and each λi\lambda_i. We distinguish this requirement from preservation of the Reeb frame and the complementary distribution; see Proposition 4.2. We also characterize horizontal generators of generalized canonical transformations by locally Hamiltonian vector fields. For a Hamiltonian HH, we consider the family EHa⃗=XH+∑iaiRiE_H^{\vec{a}}=X_H+\sum_i a_iR_i, where a⃗∈Rq\vec{a}\in\mathbb{R}^q. The Noether theorem relates horizontal infinitesimal symmetries to functions conserved by every member of this family. These functions are invariant under each Reeb field as well as under XHX_H. This simultaneous conservation condition does not require the different fields EHa⃗E_H^{\vec{a}} to commute.

In the uniform qq-contact case, a symmetry preserving HH and all the defining one-forms gives a tuple (λ1(V),…,λq(V))(\lambda_1(V),\ldots,\lambda_q(V)) satisfying a common dissipation law. Uniformity also imposes a restriction: for such strict symmetries, the differences λi(V)−λj(V)\lambda_i(V)-\lambda_j(V) are locally constant. We retain the tuple formulation to express compatibility with the whole family of one-forms.

For both structures, let XX generate a one-parameter group of canonoid transformations with smoothly varying Hamiltonians KsK_s, normalized by K0=HK_0=H. Writing κ=dds∣s=0Ks\kappa=\left.\frac{d}{ds}\right|_{s=0}K_s, we prove the necessary infinitesimal relation [X,XH]=XX(H)−κ[X,X_H]=X_{X(H)-\kappa}. The equations determining κ\kappa depend on the geometric structure and, in the uniform qq-contact case, must hold simultaneously for all defining one-forms. We also give an explicit class of canonoid flows valid for every Hamiltonian: those generated by vector fields satisfying LXλi=λi\mathcal{L}_X\lambda_i=\lambda_i for all ii.

The examples recover familiar conservation laws within these descriptions: the total angular momentum of free relativistic particles and the Poincaré angular momentum of a charge in a monopole field. The two-monopole example has a different role. Its non-uniform structure and the obstruction just described identify a physical situation beyond the uniform transformation theory developed here.

Organization of the paper

Section 2 recalls the geometric structures, establishes local normal forms, and describes the Jacobi structure associated with a uniform qq-contact manifold. Section 3 develops the constraint and monopole constructions. It also gives a local passage from qq-cosymplectic to uniform qq-contact structures under a normalization of a primitive of Ω\Omega, together with a global realization for free relativistic particles. Sections 4 and 5 contain the canonical and canonoid transformation results and their relations to symmetries and conservation or dissipation laws. Section 6 discusses the extension to non-uniform structures, the integration of infinitesimal conditions, and further questions suggested by the examples.

Preliminaries

All manifolds, differential forms, and vector fields are smooth and real. We use the contraction convention (ιXΩ)(Y)=Ω(X,Y)(\iota_X\Omega)(Y)=\Omega(X,Y). The purpose of this section is to fix the geometric data and sign conventions used below, and to explain their relation to presymplectic and precontact geometry. We include proofs of the local normal forms and basic Hamiltonian identities for completeness.

A common geometric setting

Definition 2.1 (qq-structured manifolds; cf. [18]). Let n,q,mn,q,m be positive integers and let MM have dimension 2n+q2n+q. A qq-structure on MM consists of one-forms λ⃗=(λ1,…,λq)\vec{\lambda}=(\lambda_1,\ldots,\lambda_q), two-forms Ω⃗=(Ω1,…,Ωm)\vec{\Omega}=(\Omega_1,\ldots,\Omega_m), and a splitting TM=R⊕ξTM=\mathcal{R}\oplus\xi such that:

(i) the λi\lambda_i are pointwise linearly independent and ξ=⋂i=1qker⁡λi\xi=\bigcap_{i=1}^{q}\ker\lambda_i;

(ii) each restriction Ωj∣ξ\Omega_j|_{\xi} is non-degenerate;

(iii) ker⁡Ωj=R\ker\Omega_j=\mathcal{R} for every j=1,…,mj=1,\ldots,m.

A manifold endowed with these data is called a qq-structured manifold. We use this terminology to collect the algebraic data shared by the structures considered below. No closedness condition is imposed at this stage. The case m=qm=q is the framework of [18]; here the number of two-forms is allowed to differ from the number of one-forms. This is useful in Proposition 3.3, where dλ1\mathrm{d}\lambda_1, dλ2\mathrm{d}\lambda_2, and the physically relevant form dθL\mathrm{d}\theta_L occur together, so that q=2q=2 and m=3m=3. We write (M,λ⃗⊕Ω⃗,R⊕ξ)(M,\vec{\lambda}\oplus\vec{\Omega},\mathcal{R}\oplus\xi) for these data. The bundles R\mathcal{R} and ξ\xi have ranks qq and 2n2n, respectively, and are called the Reeb distribution and the horizontal distribution. The nowhere-vanishing qq-form λ1∧⋯∧λq\lambda_1\wedge\cdots\wedge\lambda_q is the characteristic form.

Remark 2.1. Every Ωj\Omega_j has rank 2n2n. Its restriction to ξ\xi makes ξ\xi a symplectic vector bundle, a statement of fibrewise linear algebra that does not require dΩj=0\mathrm{d}\Omega_j=0. Non-degeneracy of the individual restrictions does not imply non-vanishing of their mixed wedge products. For example, on R4\mathbb{R}^4, let a=dx1∧dy1a=\mathrm{d}x_1\wedge\mathrm{d}y_1 and b=dx2∧dy2b=\mathrm{d}x_2\wedge\mathrm{d}y_2. Both a+ba+b and a−ba-b are non-degenerate, whereas (a+b)∧(a−b)=0(a+b)\wedge(a-b)=0.

Proposition 2.1. For each j=1,…,mj=1,\ldots,m, the vector bundle map

♭j:TM⟶T∗M,♭j(X)=ιXΩj+∑k=1qλk(X)λk\flat^j:TM\longrightarrow T^*M,\qquad\flat^j(X)=\iota_X\Omega_j+\sum_{k=1}^{q}\lambda_k(X)\lambda_k

is an isomorphism. It therefore induces an isomorphism of C∞(M)C^\infty(M)-modules between vector fields and one-forms.

References

  1. [1]D. Alekseevsky, J. Grabowski, G. Marmo and P. W. Michor, Poisson structures on the cotangent bundle of a Lie group or a principle bundle and their reductions, J. Math. Phys. 35 (1994), no. 9, 4909–4927. https://doi.org/10.1063/1.530822.
  2. [2]U. Almeida, Contact Anosov actions with smooth invariant bundles, PhD thesis, Universidade de São Paulo (2018). https://doi.org/10.11606/T.55.2018.tde-01112018-110622.
  3. [3]J. L. Anderson and P. G. Bergmann, Constraints in covariant field theories, Phys. Rev. 83 (1951), 1018–1025. https://doi.org/10.1103/PhysRev.83.1018.
  4. [4]R. Azuaje and A. M. Escobar-Ruiz, Canonical and canonoid transformations for Hamiltonian systems on (co)symplectic and (co)contact manifolds, J. Math. Phys. 64 (2023), 033501. https://doi.org/10.1063/5.0135045.
  5. [5]R. Azuaje and A. Bravetti, Scaling symmetries and canonoid transformations in Hamiltonian systems, Int. J. Geom. Methods Mod. Phys. 21 (2024), 2450077. https://doi.org/10.1142/S0219887824500774.
  6. [6]R. Azuaje and X. F. Zhao, Canonical and canonoid transformations for Hamiltonian systems on locally conformal symplectic manifolds, J. Geom. Phys. 222 (2026), 105761. https://doi.org/10.1016/j.geomphys.2026.105761.
  7. [7]A. P. Balachandran, G. Marmo, B. S. Skagerstam and A. Stern, Magnetic monopoles with no strings, Nucl. Phys. B 162 (1980), 385–396. https://doi.org/10.1016/0550-3213(80)90346-6.
  8. [8]A. P. Balachandran, G. Marmo, N. Mukunda, J. S. Nilsson, A. Simoni, E. C. G. Sudarshan and F. Zaccaria, Unified geometrical approach to relativistic particle dynamics, J. Math. Phys. 25 (1984), 167–176. https://doi.org/10.1063/1.525987.
  9. [9]R. Bogdanović and M. S. Gopinathan, A canonical transformation of the Hamiltonians quadratic in coordinate and momentum operators, J. Phys. A 12 (1979), 1457–1468. https://doi.org/10.1088/0305-4470/12/9/013.
  10. [10]J. F. Cariñena, J. Fernández-Núñez and E. Martínez, A geometric approach to Noether’s second theorem in time-dependent Lagrangian mechanics, Lett. Math. Phys. 23 (1991), 51–63. https://doi.org/10.1007/BF01811294.
  11. [11]J. F. Cariñena, F. Falceto and M. F. Rañada, Canonoid transformations and master symmetries, J. Geom. Mech. 5 (2013), 151–166. https://doi.org/10.3934/jgm.2013.5.151.
  12. [12]J. F. Cariñena and M. F. Rañada, Canonoid transformations from a geometric perspective, J. Math. Phys. 29 (1988), 2181–2186. https://doi.org/10.1063/1.528146.
  13. [13]D. G. Currie, T. F. Jordan and E. C. G. Sudarshan, Relativistic invariance and Hamiltonian theories of interacting particles, Rev. Mod. Phys. 35 (1963), 350–375. https://doi.org/10.1103/RevModPhys.35.350.
  14. [14]D. G. Currie and E. J. Saletan, Canonical transformations and quadratic Hamiltonians, Nuovo Cimento B 9 (1972), 143–153. https://doi.org/10.1007/BF02735514.
  15. [15]P. A. M. Dirac, Generalized Hamiltonian dynamics, Canad. J. Math. 2 (1950), 129–148. https://doi.org/10.4153/CJM-1950-012-1.
  16. [16]B. A. Dubrovin, M. Giordano, G. Marmo and A. Simoni, Poisson brackets on presymplectic manifolds, Int. J. Mod. Phys. A 8 (1993), no. 21, 3747–3771. https://doi.org/10.1142/S0217751X93001521.
  17. [17]M. de León and J. Bajo, A geometric description of some thermodynamical systems, J. Phys. A: Math. Theor. 58 (2025), 175203. https://doi.org/10.1088/1751-8121/acdc14.
  18. [18]M. de León, C. Sardón and X. Zhao, A geometric approach to the (generalized) Noether theorem for Hamiltonian systems on (non)-uniform α-contact manifolds, arXiv:2608.05681 [math.SG] (2026). https://arxiv.org/abs/2608.05681.
  19. [19]P. A. M. Dirac, Quantised singularities in the electromagnetic field, Proc. R. Soc. Lond. A 133 (1931), 60–72. https://doi.org/10.1098/rspa.1931.0130.
  20. [20]P. A. M. Dirac, Lectures on Quantum Mechanics, Belfer Graduate School of Science, Yeshiva University, New York (1964).
  21. [21]D. Finamore, Contact foliations and generalised Weinstein conjectures, Ann. Global Anal. Geom. 65 (2024), 27. https://doi.org/10.1007/s10455-024-09957-w.
  22. [22]D. Finamore, Quasiconformal contact foliations, Math. Ann. 389 (2024), 1575–1598. https://doi.org/10.1007/s00208-023-02687-7.
  23. [23]M. Leok, C. Sardón and X. F. Zhao, Integration on q-cosymplectic manifolds, J. Nonlinear Sci. 36 (2026), 56. https://doi.org/10.1007/s00332-026-10279-3.
  24. [24]M. Leok, C. Sardón and X. F. Zhao, q-Cosymplectic geometry, integrability and reduction, arXiv:2509.05998 [math-ph] (2025). https://arxiv.org/abs/2509.05998.
  25. [25]M. Leok, C. Sardón and X. F. Zhao, Noether-type theorems and the generalized Herglotz principle in q-contact geometry, J. Phys. A: Math. Theor. 59 (2026), 375203. https://doi.org/10.1088/1751-8121/aea463.
  26. [26]G. Marmo, N. Mukunda and J. Samuel, Dynamics and symmetry for constrained systems: a geometrical analysis, Riv. Nuovo Cimento 6 (1983), no. 2, 1–62. https://doi.org/10.1007/BF02740487.
  27. [27]G. Marmo, G. Morandi and N. Mukunda, A geometrical approach to the Hamilton-Jacobi form of dynamics and its generalizations, Riv. Nuovo Cimento 13 (1990), no. 8, 1–74. https://doi.org/10.1007/BF02832785.
  28. [28]N. Mukunda, Symmetries and constraints in generalized Hamiltonian dynamics, Ann. Physics 99 (1976), 408–433. https://doi.org/10.1016/0003-4916(76)90101-9.
  29. [29]N. Mukunda, Phase space methods and the Hamilton-Jacobi form of dynamics, Proc. Indian Acad. Sci. Sect. A 87 (1978), no. 5, 85–105. https://repository.ias.ac.in/25374/1/426.pdf.DOI
  30. [30]L. J. Negri, L. C. Oliveira and J. M. Teixeira, Canonical transformations and constants of motion, J. Math. Phys. 28 (1987), 2369–2372. https://doi.org/10.1063/1.527772.
  31. [31]J. Pérez Álvarez, Symmetries and dissipation laws on contact systems, Mediterr. J. Math. 21 (2024), 151. https://doi.org/10.1007/s00009-024-02695-0.
  32. [32]G. Rastelli and M. Santoprete, Canonical and Poissonoid transformations, symmetries and bihamiltonian structures, J. Geom. Mech. 7 (2015), 483–515. https://doi.org/10.3934/jgm.2015.7.483
  33. [33]J. Schwinger, Quantum theory of gauge fields, in Elementary Particle and High Energy Physics, eds. M. Lévy and Ph. Meyer (Gordon and Breach, New York, 1965), p. 189.
  34. [34]K. Yano and S. Ishihara, Tangent and Cotangent Bundles: Differential Geometry, Monographs and Textbooks in Pure and Applied Mathematics, Vol. 16 (Marcel Dekker, New York, 1973).
  35. [35]F. Zaccaria, E. C. G. Sudarshan, J. S. Nilsson, N. Mukunda, G. Marmo and A. P. Balachandran, Universal unfolding of Hamiltonian systems: from symplectic structure to fiber bundles, Phys. Rev. D 27 (1983), 2327–2340. https://doi.org/10.1103/PhysRevD.27.2327.
  36. [36]X. F. Zhao, Canonoid transformation and master symmetries of Hamiltonian systems on locally conformal symplectic manifold, J. Math. Phys. 66 (2025), 082704. https://doi.org/10.1063/5.0255345.
  37. [37]X. F. Zhao and Y. Li, Complete integrability in q-contact Hamiltonian systems, Bull. Sci. Math. 213 (2026), 103928. https://doi.org/10.1016/j.bulsci.2026.103928.

Paper details

vlnc:W_s67p2g4w
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY-SA 4.0
Usage

views · 5

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents