WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

Nonclassical Electromagnetic Dynamics

CONSTANTIN UDRISTE University Politehnica of Bucharest Department of Mathematics Splaiul Independentei 313 060042 Bucharest ROMANIA udriste@mathem.pub.ro DOREL ZUGRAVESCU Institute of Geodynamics Dr. Gerota 19-21 020032 Bucharest ROMANIA dorezugr@geodin.ro FLORIN MUNTEANU Institute of Geodynamics Dr. Gerota 19-21 020032 Bucharest ROMANIA orin@geodin.ro

Abstract: Our paper is concerned with effects of special forces on the motion of particles. 1(2) studies the singletime geometric dynamics induced by electromagnetic vector elds (1-forms) and by the Euclidean structure of the space. 3 denes the second-order forms or vectors. 4 (5) describes a nonclassical electric (magnetic) dynamics produced by an electric (magnetic) second-order Lagrangian, via the extremals of the energy functional. 6 generalize this dynamics for a general second-order Lagrangian, having in mind possible applications for dynamical systems coming from Biomathematics, Economical Mathematics, Industrial Mathematics, etc. KeyWords: geometric dynamics, second-order vectors, second-order Lagrangian, nonclassical dynamics.

Single-time geometric dynamics induced by electromagnetic vector elds

It is well-known that the motion of a charged particle in the electromagnetic elds is described by the ODE system (Lorentz World-Force Law) m d2 x dx =e E+ B , 2 dt dt

Let U be a domain of linear homogeneous isotropic media in the Riemannian manifold (M = R3 , ij ). Maxwells equations (coupled PDEs of rst order) div D = , rot H = J + t D, t = time derivative operator div B = 0, rot E = t B with the constitutive equations B = H, D = E,

x = (x1 , x2 , x3 ) U R3 , where m is the mass, and e is the charge of the particle. Of course, these are Euler-Lagrange equations produced by the Lorentz Lagrangian 1 dxi dxj dxi j L1 = mij + eij A eV. 2 dt dt dt The associated Lorentz Hamiltonian is m dxi dxj ij + eV. 2 dt dt A similar mechanical motion was discovered by Popescu [7], accepting the existence of a gravitovordx tex eld represented by the force G + , where dt G is the gravitational eld and is a vortex determining the gyroscopic part of the force. The papers [5], [8], [9], [12], [14], [15] conrm the point of view of Popescu via geometric dynamics. H1 =

on R U , reect the relations between the electromagnetic elds: E H J B electric eld strength magnetic eld strength electric current density permitivity permeability magnetic induction (magnetic ux density) D [C/m2 ] = [As/m2 ] electric displacement (electric ux density) [V /m] [A/m] [A/m2 ] [As/V m] [V s/Am] [T ] = [V s/m2 ]

1.1

Since div B = 0, the vector eld B is source free, hence may be expressed as rot of some vector potential A, i.e., B = rot A. Then the electric eld strength is E = grad V t A.
ISSN: 1109-2769 31

Single-time geometric dynamics produced by vector potential A

To conserve the traditional formulas, we shall refer to eld lines of the vector potential A using dimensional homogeneous relations. These curves are
Issue 1, Volume 7, January 2008

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

solutions of the ODE system m dx = eA. dt

where the unit measure for the constant is [kgm3 /V s2 ]. The associated least squares Lagrangian is dxi 1 L3 = ij m B i 2 dt = 1 dx m B 2 dt m
2

This ODE system and the Euclidean metric produce the least squares Lagrangian 1 dxi L2 = ij m + eAi 2 dt = dx 1 m + eA 2 dt m
2

dxj B j dt .

dxj + eAj dt .

This gives the Euler-Lagrange equations (single-time magnetic geometric dynamics) m where dx d2 x 2 = rot B + fB + t B, dt2 dt m 1 1 fB = ij B i B j = ||B||2 2 2

The Euler-Lagrange equations associated to L2 are m d2 xi =e dt2 Aj Ai xi xj dxj e2 fA + et A, dt m xi

1 fA = ij Ai Aj 2 is the energy density produced by A. Equivalently, it appears a single-time geometric dynamics d2 x dx e2 m 2 =e B+ fA et A, dt dt m B = rot A,

where

is the magnetic energy density. This is in fact a dynamics under a gyroscopic force or in J-vortex. The associated Hamiltonian is 1 dxi H3 = ij m B i 2 dt = m dxj + B j dt

a motion in a gyroscopic force [12], [14], [15] or a B-vortex dynamics. The associated Hamiltonian is 1 dxi H2 = ij m eAi 2 dt = m dxj + eAj dt

m2 dxi dxj ij 2 fB . 2 dt dt

1.3

Single-time geometric dynamics produced by electric eld E

m2 dxi dxj ij e2 fA . 2 dt dt Remark. Generally, the single-time geometric dynamics produced by the vector potential A is different from the classical Lorentz World-Force Law because 1 L2 mL1 = e2 ij Ai Aj + meV 2 e and the force fA t A is not the electric eld m E = V t A. In other words the Lagrangians L1 and L2 are not in the same equivalence class of Lagrangians.

The electric ow is described by m dx = E, dt

where the unit measure of the constant is [kgm2 /V s]. It appears the least squares Lagrangian dxi 1 L4 = ij m E i 2 dt 1 dx = m E 2 dt m
2

dxj E j dt ,

1.2

Single-time geometric dynamics produced by magnetic induction B

with Euler-Lagrange equations (single-time electric geometric dynamics) m where d2 x dx 2 = rot E + fE + t E, dt2 dt m 1 1 fE = ij E i E j = ||E||2 2 2
Issue 1, Volume 7, January 2008

Since we want to analyze the geometric dynamics using units of measure, the magnetic ow must be described by dx = B, m dt
ISSN: 1109-2769 32

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

is the electric energy density; here we have in fact a dynamics in t B-vortex. The associated Hamiltonian is 1 dxi H4 = ij m E i 2 dt = m dxj + E j dt

2.1

Potential associated to electric 1-form E

m2 dxi dxj ij 2 fE . 2 dt dt Open problem. As is well-known, charged particles in the magnetic eld of the earth spiral from pole to pole. Similar motions are also observed in laboratory plasmas and inferred for electrons in metal subjected to an external magnetic eld. Till now, these motions where justied by the classical Lorentz World-Force Law. Can we justify such motions using the geometric dynamics produced by suitable vector elds in the sense of present paper and the papers [5], [7]-[12], [14], [15]?

Let us consider the function V : R U R, (t, x) V (t, x) and the Pfaff equation dV = E or V the equivalent PDE system = Ei , i = 1, 2, 3. xi Of course, the complete integrability conditions require dE = 0 (electrostatic eld), which is not always satised. In any situation we can introduce the least squares Lagrangian 1 L9 = ij 2 V + Ei xi V + Ej xj

1 = ||dV + E||2 . 2 These produce the Euler-Lagrange equation (Poisson equation) V = div E, and consequently V must be the electric potential. For a linear isotropic material, we have D = E, with = div D = div E. We get the potential equation for a homogeneous material ( = constant) V = . For the charge free space, we have = 0, and then the potential V satises the Laplace equation V = 0 (harmonic function). The Lagrangian L9 produces the Hamiltonian 1 H9 = (dV E, dV + E), 2 and the momentum-energy tensor eld T ij = V xj L V xi
i L9 j

Potentials associated to electromagnetic forms

Let U R3 = M be a domain of linear homogeneous isotropic media. Which mathematical object shall we select to model the electromagnetic elds; vector elds or forms? It was shown [1] that, the magnetic induction B, the electric displacement D, and the electric current density J are all 2-forms; the magnetic eld H, and the electric eld E are 1-forms; the electric charge density is a 3-form. The operator d is the exterior derivatives and the operator t is the time derivative. In terms of differential forms, the Maxwell equations on U R can be expressed as dD = , dB = 0, dH = J + t D dE = t B.

The constitutive relations are D = E, B = H, where the star operator is the Hodge operator, is the permitivity, and is the scalar permeability. The local components Ei , i = 1, 2, 3, of the 1form E are called electric potentials, and the local components Hi , i = 1, 2, 3, of the 1-form H are called magnetic potentials. Since the electric eld E, and the magnetic eld H are 1-forms [1], in the Sections 4-5 we combine our ideas [8]-[16] with the ideas of Emery [3] and Foster [4], creating a nonclassical electric or magnetic dynamics. Finally, we generalize the results to nonclassical dynamics induced by a second-order Lagrangian (Section 6).
ISSN: 1109-2769 33

V xj

V + Ei xi

i L9 j .

2.2

Potential associated to magnetic 1-form H

Now, we consider the Pfaff equation d = H or the PDE system = Hi , i = 1, 2, 3, : R U R, xi (t, x) (t, x). The complete integrability conditions dH = 0 are satised only for particular cases. If we build the least squares Lagrangian 1 L10 = ij 2 Hi xi Hj xj

Issue 1, Volume 7, January 2008

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

1 = ||d H||2 , 2 then we obtain the Euler-Lagrange equation (Laplace equation) and consequently must be the magnetic potential. The Lagrangian L10 produces the Hamilto1 nian H10 = (d H, d + H). 2

2.3

Potential associated to 1-form potential A

for an extremum. The Euler-Lagrange equation, 2z 2z + = 0, shows that the extremals are harx2 y 2 monic functions. Of course, J has no global miniz z mum point since the PDE system = y, = x x y is not completely integrable. 2) All previous examples in this section are single-component potentials. Similarly we can introduce the multi-component potentials.

Since dB = 0, there exists an 1-form potential A satisfying B = dA. Now let us consider the Pfaff equa tion d = A or the equivalent PDEs system = xi Ai , i = 1, 2, 3, : R U R, (t, x) (t, x). The complete integrability conditions dA = 0 are satised only for B = 0. If we build the least squares Lagrangian 1 L11 = ij 2 Ai xi Aj xj

Second-Order Forms and Vectors

Now we will concentrate on certain geometric ideas that are very important in the physical and stochastic applications. To avoid too much repetition, M will denote a differentiable manifold of dimension n, and all the functions are of class C . Let xi = xi (xi ), i, i = 1, . . . , n be a changing of coordinates on M . Then we introduce the symbols
i Di =

1 = ||d A||2 , 2 then we obtain the Euler-Lagrange equation (Laplace equation) = 0. The Lagrangian L11 gives the 1 Hamiltonian H11 = (d A, d + A) and the 2 momentum-energy tensor eld T
i j

xi , xi

i Di j =

2 xi . xi xj

For a differentiable function f : M R we use the simplied coordinate expression f (xi ) = f (xi (xi )), the rst order derivatives f,i , f,i and the second order derivatives f,i j , f,ij . These are connected by the rule (f,i , f,i j ) = (f,i , f,ij )
i Di i Di j j i Di Dj

= xj

Ai xi

i L11 j .

(1)

Open problem. Find interpretations for the extremals of least squares Lagrangians 1 L12 = ||dA B||2 2 1 = ik jl 2 Ai Aj Bij xj xi Ak Al Bkl l x xl

The pair (rst-order partial derivatives, secondorder partial derivatives) possesses a tensorial change law that the second derivative, by itself, lacks. This pair was used in the classical works like contact element or like jet. If a curve is given by the parametric equations xi = xi (t), t I, then the preceding diffeomorphism modies the pair (, x x)T as follows x xi = xi Di , xi = xi Di + xi xj Dij , i i i xi xi xj =
i Di i Dij j i Di Dj

1 1 L13 = ||dE + t B||2 + ||dH J t D||2 2 2 1 1 + ||dD ||2 + ||dB||2 , 2 2 which are not solutions of Maxwell equations. Remarks. 1) There are a lot of applications of previous type in applied sciences. One of the most important is in the material strength. In problems associated with the torsion of a cylinder or prism, one has to investigate the functional J(z()) =
D

xi xi xj

The pair (acceleration, square of velocity) is suggested by the equations of geodesics xk (t) + k (x(t))xi (t)xj (t) = 0. ij Consequently, when rst and second derivatives come into play together, then matrices of blocks such
34 Issue 1, Volume 7, January 2008

z y x

z +x y

dxdy

ISSN: 1109-2769

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

as K=
i Di i Di j j i Di Dj

, K
1

i Di

i Dij j i Di Dj

is a second-order object determined by gij = ij 1 2 (Ei,j + Ej,i ). If ij is symmetric, then gij is a symmetric tensor eld; if we add det(gij ) = 0, then gij can be used as a semi-Riemann metric. In this context the equality 1 (Ei , ij ) = (Ei , (Ei,j + Ej,i )) + (0, gij ) 2 shows that the valuable objects 1 (Ei,j + Ej,i ) 2 come from electricity and mate with gravitational potentials gij . Consequently they have to be electrogravitational potentials. The preceding potentials determine an electric energy second-order Lagrangian, Lel = Ei (x(t), t) d2 xi (t) dt2

are useful for changing the components of pairs of objects. Denition. Any pair (i , ij ) admitting the changing law (1), where (i ) is an 1-form and (ij ) is symmetric, will be called a second-order form on M . A second-order vector eld on M is a second-order differential operator, with no constant term. A second-order form can be written as = i d2 xi + ij dxi dxj , where the components i and ij = ji are smooth functions. A second-order vector eld writes Xf = ij Dij f + i Di f , where the components ij = ji and i are smooth functions. The theory of second-order vectors or forms appears in Emery [3], with applications in stochastic problems, and in Foster [4], suggesting new point of view about the elds theory. We use these ideas to dene meaningful second-order Lagrangians (kinetic potentials) and to study their extremals.

Dynamics induced by secondorder electric form

1 dxi dxj + (Ei,j + Ej,i )(x(t), t) (t) (t). 2 dt dt If the most general energy second-order Lagrangian is given by Lge = Ei (x(t), t) d2 xi dxi dxj (t)+ij (x(t), t) (t) (t), dt2 dt dt

Let Ei be the electric potentials. The usual derivative Ei,j may be decomposed into skew-symmetric and symmetric parts, Ei,j 1 1 = (Ei,j Ej,i ) + (Ei,j + Ej,i ). 2 2

and the gravitational energy rst-order Lagrangian is Lg = gij (x(t), t) dxi dxj (t) (t), dt dt

The skew-symmetric part (vortex) 1 mij = (Ei,j Ej,i ) 2 is called Maxwell tensor eld giving the opposite of the time derivative of magnetic induction. The symmetric part 1 (Ei,j + Ej,i ) 2 is not an ordinary tensor, but the pair (Ei , 1 (Ei,j + Ej,i )) 2

then Lge = Lel + Lg . To simplify, let us take E = E(x). Since the distribution generated by the electric 1-form E is given by the Pfaff equation Ei (x)dxi = 0, the electric energy Lagrangian is zero along integral curves of this distribution. Also the general energy second-order Lagrangian for E = E(x), ij = ij (x) determines the energy functional
b

Ei (x(t))
a

d2 xi dxi dxj + ij (x(t)) (t) (t) dt. dt2 dt dt (2)

We denote 1 ijk = (kj,i + ki,j ij,k ) 2 (the Christoffel symbols of ij ), Eijk =

35

is a second-order object [3]-[4]. Let ij be a general object such that (Ei , ij ) is a second-order object. The difference 1 (0, ij (Ei,j + Ej,i )) 2
ISSN: 1109-2769

1 (Ek,ij + Ej,ik Ei,jk ) . 2

Issue 1, Volume 7, January 2008

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

Theorem. The extremals of the energy functional (2) are described by the Euler-Lagrange ODEs gki d2 xi dxi dxj + (ijk Eijk ) = 0, dt2 dt dt x(a) = xa , x(b) = xb . (geodesics with respect to an Otsuki connection [6]). Proof. Using the second-order Lagrangian L = Ei d2 xi dxi dxj + ij , dt2 dt dt

is the Maxwell tensor eld (vortex) giving the sum between the electric current density and the time derivative of the electric displacement. The pair Hi , 1 (Hi,j + Hj,i ) 2

is a second-order object. If (Hi , ij ) is a general second-order object, then the difference 1 gij = ij (Hi,j + Hj,i ) 2 represents the gravitational potentials (a metric) provided that ij is symmetric, gij is a (0,2) tensor eld and det(gij ) = 0. Consequently the valuable ob1 jects (Hi,j +Hj,i ), which come from magnetism and 2 mate, have to be magnetogravitational potentials. The preceding potentials produce the following energy Lagrangians: 1) magnetic energy second-order Lagrangian, Lma = Hi (x(t), t) d2 xi (t) dt2

the Euler-Lagrange equations Lxk d d2 L dxk + 2 L d2 xk = 0 dt dt dt dt2

transcribe like the equations in the theorem. To a Lagrangian there may corresponds a eld theory. Consequently we obtain a eld theory having as basis the general electrogravitational potentials. The pure gravitational potentials are given by 1 gij = ij (Ei,j + Ej,i ). 2 We dene ijk = ijk Eijk . It is veried that ijk = gijk + mijk , where gijk are the Christoffel symbols of gij , and mijk = mij,k + mik,j is the symmetrized derivative of the Maxwell tensor 1 mij = (Ei,j Ej,i ). 2 Corollary. The extremals of the energy functional (1) are described by the Euler-Lagrange ODEs gki d2 xi dxi dxj + (gkji + mkji ) = 0, dt2 dt dt x(a) = xa , x(b) = xb . (geodesics with respect to an Otsuki connection [6]).

dxi dxj 1 (t) (t); + (Hi,j + Hj,i )(x(t), t) 2 dt dt 2) gravitational energy rst-order Lagrangian, Lg = gij (x(t), t) dxi dxj (t) (t); dt dt

3) general energy second-order Lagrangian, Lge = Hi (x(t), t) d2 xi dxi dxj + ij (x(t), t) (t) (t). dt2 dt dt

Dynamics induced by secondorder magnetic form

These satisfy the relation Lge = Lma + Lg . To simplify, let us take H = H(x). Since the distribution generated by the magnetic 1-form H is given by the Pfaff equation Hi (x)dxi = 0, the magnetic energy Lagrangian is zero along integral curves of this distribution. Also the general energy second-order Lagrangian for H = H(x), ij = ij (x) determines the energy functional dxi dxj d2 xi (t) + ij (x(t)) (t) (t) dt dt2 dt dt a (3) determined by a second-order Lagrangian which is linear in acceleration. We denote Hi (x(t)) 1 ijk = (kj,i + ki,j ij,k ) 2
36 Issue 1, Volume 7, January 2008
b

Let Hi be the magnetic potentials. The usual derivative Hi,j may be decomposed into skew-symmetric and symmetric parts, 1 1 Hi,j = (Hi,j Hj,i ) + (Hi,j + Hj,i ), 2 2 where 1 Mij = (Hi,j Hj,i ) 2

ISSN: 1109-2769

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

(the Christoffel symbols of ij ), Hijk = 1 (Hk,ij + Hj,ik Hi,jk ) . 2

is a Maxwell tensor eld (vortex). The pair 1 i , (i,j + j,i ) 2 is a second-order form. If (i , ij ) is a general second-order form, then we suppose that the difference 1 gij = ij (i,j + j,i ) 2 represents the components of a metric, i.e., ij is symmetric, gij is a (0,2) tensor eld and det(gij ) = 0. The preceding potentials produce the following energy Lagrangians: 1) potential-produced energy Lagrangian, Lpp = i (x(t), t) d2 xi (t) dt2

Theorem. The extremals of the energy functional (3) are solutions of the Euler-Lagrange ODEs gki d2 xi dxi dxj + (ijk Hijk ) = 0, dt2 dt dt x(a) = xa , x(b) = xb . (geodesics with respect to an Otsuki connection [6]). To an energy Lagrangian there may corresponds a eld theory. Consequently we obtain a eld theory having as basis the general magnetogravitational potentials. The pure gravitational potentials are (components of a Riemann or semi-Riemann metric) 1 gij = ij (Hi,j + Hj,i ). 2 We introduce ijk = ijk Hijk . It is veried the relation ijk = gijk +Mijk , where gijk are the Christoffel symbols of gij , and Mijk = Mij,k + Mik,j is the symmetrized derivative of the Maxwell tensor 1 Mij = (Hi,j Hj,i ). 2 Corollary. The extremals of the energy functional (3) are solutions of the Euler-Lagrange ODEs gki d2 xi dxi dxj + (gkji + Mkji ) = 0, dt2 dt dt x(a) = xa , x(b) = xb . (geodesics with respect to an Otsuki connection [6]).

1 dxi dxj + (i,j + j,i )(x(t), t) (t) (t); 2 dt dt 2) gravitational energy Lagrangian, Lg = gij (x(t), t) dxi dxj (t) (t); dt dt

3) general energy Lagrangian, Lge = i (x(t), t) which verify Lge = Lpp + Lg . The Pfaff equation i (x)dxi = 0, i = 1, . . . , n denes a distribution on M . The valuable objects 1 (i,j + j,i ) 2 are the components of the second fundamental form of that distribution [6]. The potential-produced energy Lagrangian is zero along integral curves of the distribution generated by the given 1-form = (i (x)). In the autonomous case, the general energy Lagrangian produces the energy functional d2 xi dxi dxj (t) + ij (x(t)) (t) (t) dt. dt2 dt dt a (4) This energy functional is associated to a particular Lagrangian L of order two. We denote i (x(t)) 1 ijk = (kj,i + ki,j ij,k ) 2
37 Issue 1, Volume 7, January 2008
b

dxi dxj d2 xi + i,j (x(t), t) ; dt2 dt dt

Dynamics induced by secondorder general form

Now we want to extend the preceding explanations since they can be applied to dynamical systems coming from Biomathematics, Economical Mathematics, Industrial Mathematics etc. Let i be given potentials (given form). The usual partial derivative i,j may be decomposed as 1 1 i,j = (i,j j,i ) + (i,j + j,i ), 2 2 where 1 Mij = (i,j j,i ) 2

ISSN: 1109-2769

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

(the Christoffel symbols of ij ), ijk = 1 (k,ij + j,ik i,jk ) . 2

[3] M. Emery, Stochastic Calculus in Manifolds, Springer-Verlag, 1989. [4] B. L. Foster, Higher derivatives in geometry and physics, Proc. R. Soc. Lond. A423 (1989), 443455. [5] D. Isvoranu, C. Udriste, Fluid ow versus Geometric Dynamics, 5-th Conference on Differential Geometry, August 28-September 2, 2005, Mangalia, Romania; BSG Proceedings 13, pp. 70-82, Geometry Balkan Press, 2006. [6] T. Otsuki, General connections, Math. J. Okayama University 32 (1990), 227-242. [7] I. N. Popescu, Gravitation, Editrice Nagard, Roma, Italy, 1988. [8] C. Udriste, Geometric dynamics, Southeast Asian Bulletin of Mathematics, Springer-Verlag, 24, 1 (2000), 313-322. [9] C. Udriste, Geometric Dynamics, Mathemat ics and Its Applications, 513, Kluwer Academic Publishers, Dordrecht, Boston, London, 2000. [10] C. Udriste, Dynamics induced by second-order objects, Global Analysis, Differential Geometry and Lie Algebras, BSG Proceedings 4, Ed. Grigorios Tsagas, Geometry Balkan Press (2000), 161-168. [11] C. Udriste, Multi-time dynamics induced by 1 forms and metrics, Global Analysis, Differential Geometry and Lie Algebras, BSG Proceedings 5, Ed. Grigorios Tsagas, Geometry Balkan Press (2001), 169-178. [12] C. Udriste, Geodesic motion in a gyroscopic eld of forces, Tensor, N. S., 66, 3 (2005), 215228. [13] C. Udriste, O. Dogaru, Convex nonholonomic hypersurfaces, The Mathematical Heritage of C.F. Gauss, 769-784, Editor G. M. Rassias, World Scientic Publ. Co. Singapore, 1991. [14] C. Udriste, M. Ferrara, D. Opris, Economic Ge ometric Dynamics, Monographs and Textbooks 6, Geometry Balkan Press, Bucharest, 2004. [15] C. Udriste, Tools of geometric dynamics, Bulet inul Institutului de Geodinamic , Academia a Rom n , 14, 4 (2003), 1-26; Proceedings of the a a XVIII Workshop on Hadronic Mechanics, honoring the 70-th birthday of Prof. R. M. Santilli, the originator of hadronic mechanics, University of Karlstad, Sweden, June 20-22, 2005; Eds.
38 Issue 1, Volume 7, January 2008

Theorem. The extremals of the energy functional (4) are solutions of the Euler-Lagrange ODEs gki d2 xi dxi dxj + (ijk ijk ) = 0, dt2 dt dt x(a) = xa , x(b) = xb (geodesics with respect to an Otsuki connection [6]). To an energy Lagrangian there may correspond a eld theory. We introduce ijk = ijk ijk . After some computations we nd ijk = gijk +mijk , where gijk are the Christoffel symbols of gij , and mijk = Mij,k + Mik,j is the symmetrized derivative of the Maxwell tensor eld M. Corollary. The extremals of the energy functional (4) are solutions of the Euler-Lagrange ODEs gki d2 xi dxi dxj + (gkji + mkji ) = 0, dt2 dt dt x(a) = xa , x(b) = xb (geodesics with respect to an Otsuki connection [6]). Open problems: (1) Find the linear connections in the sense of Crampin [2] associated to the preceding second-order ODEs. (2) Analyze the second variations of the preceding energy functionals. (2) Analyze the symmetries of the preceding second order differential systems (see also [16]). (3) Find practical interpretations for motions known as geometric dynamics, gravitovortex motion and second-order force motion (see also [5]). Acknowledgements: Partially supported by Grant CNCSIS 86/ 2007 and by 15-th ItalianRomanian Executive Programme of S&T Cooperation for 2006-2008, University Politehnica of Bucharest. References: [1] A. Bossavit, Differential forms and the computation of elds and forces in electromagnetism, Eur. J. Mech., B, Fluids, 10, no 5(1991), 474488. [2] C. Crampin, A linear connection associated with any second order differential equation eld, L.Tamassy and J.Szenthe (eds), New Developments in Differential Geometry, 77-85, Kluwer Academic Publishers, 1996.
ISSN: 1109-2769

WSEAS TRANSACTIONS on MATHEMATICS

Constantin Udriste, Dorel Zugravescu, Florin Munteanu

Valer Dvoeglazov, Tepper L. Gill, Peter Rowland, Erick Trell, Horst E. Wilhelm, Hadronic Press, International Academic Publishers, December 2006, ISBN 1-57485-059-28, pp 10011041. [16] C. Wafo Soh, C. Udriste, Symmetries of sec ond order potential differential systems, Balkan Journal of Geometry and Its Applications, Vol. 10, No. 2 (2005), 129-141.

ISSN: 1109-2769

39

Issue 1, Volume 7, January 2008