Parametric Nonholonomic Frame Transforms and Exact Solutions in Gravity
Sergiu I. Vacaru

TL;DR
This paper introduces a geometric method using nonholonomic frame deformations to construct exact off-diagonal solutions in Einstein and string gravity, revealing new classes of solutions with potential physical implications.
Contribution
It develops a systematic approach for generating exact solutions in gravity theories using parametric and nonholonomic deformations, extending previous methods.
Findings
Constructed five classes of off-diagonal solutions in vacuum Einstein and string gravity.
Derived solutions describing solitonic pp-wave interactions.
Explored physical implications of solutions based on Schwarzschild and pp-wave metrics.
Abstract
A generalized geometric method is developed for constructing exact solutions of gravitational field equations in Einstein theory and generalizations. First, we apply the formalism of nonholonomic frame deformations (formally considered for nonholonomic manifolds and Finsler spaces) when the gravitational field equations transform into systems of nonlinear partial differential equations which can be integrated in general form. The new classes of solutions are defined by generic off-diagonal metrics depending on integration functions on one, two and three (or three and four) variables if we consider four (or five) dimensional spacetimes. Second, we use a general scheme when one (two) parameter families of exact solutions are defined by any source-free solutions of Einstein's equations with one (two) Killing vector field(s). A successive iteration procedure results in new classes of…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Parametric Nonholonomic Frame Transforms
and Exact Solutions in Gravity
Sergiu I. Vacaru
The Fields Institute for Research in Mathematical Science
222 College Street, 2d Floor, Toronto M5T 3J1, Canada sergiu*-*[email protected], [email protected]
(September 19, 2007)
Abstract
A generalized geometric method is developed for constructing exact solutions of gravitational field equations in Einstein theory and generalizations. First, we apply the formalism of nonholonomic frame deformations (formally considered for nonholonomic manifolds and Finsler spaces) when the gravitational field equations transform into systems of nonlinear partial differential equations which can be integrated in general form. The new classes of solutions are defined by generic off–diagonal metrics depending on integration functions on one, two and three (or three and four) variables if we consider four (or five) dimensional spacetimes. Second, we use a general scheme when one (two) parameter families of exact solutions are defined by any source–free solutions of Einstein’s equations with one (two) Killing vector field(s). A successive iteration procedure results in new classes of solutions characterized by an infinite number of parameters for a non–Abelian group involving arbitrary functions on one variable. Five classes of exact off–diagonal solutions are constructed in vacuum Einstein and in string gravity describing solitonic pp–wave interactions. We explore possible physical consequences of such solutions derived from primary Schwarzschild or pp–wave metrics.
Keywords: Exact solutions; Finsler geometry methods; nonlinear connections.
1 Introduction
Even through, a large number of exact solutions were found in various models of gravity theory [1, 2, 3], there are available only a few general methods for generating new solutions from a given metric describing a physical situation to certain new physical properties and geometric configurations. In quantum field theory, (although approximated) some methods where formulated by using the formalism of Green’s functions, or quantum integrals, the solutions are constructed to represent a linear or nonlinear prescribed physical situation. Perhaps it is unlikely that similar computation techniques can be elaborated in general form in gravity theories. Nevertheless, such approaches where developed when new classes of exact solutions are constructed following some general geometric/ group principles and ideas [4, 5, 6, 7, 8, 9, 10]. Although many of the solutions resulting from such methods have no obvious physical interpretation, one can be formulated some criteria selecting explicit classes of solutions with prescribed symmetries and physical properties.
The technique proposed in works [4, 5] generates exact source–free solutions of Einstein equations and treats spacetimes having one, or two, Killing vectors. The scheme introduced in [4] begins with any source–free solution of Einstein’s equation with a Killing vector and defines a one parameter family (possessing a nontrivial group structure) of exact solutions. Even through starting from a quite simple solution like the Schwarzschild one, the resulting metrics were considered too sophisticated to admit any simple interpretation. In the second work [5], the author proved that the case of two Killing vectors111for instance, a Weyl solution, which is a space with two commuting Killing vectors is more appropriate for physical interpretation. In such a case, one must specify an arbitrary curve (up to parametrization) on a three–dimensional vector spaces associated to an exact four dimensional solution. The so–called parametric transform (forming a non–Abelian group) was defined generating, from a single solution, a family of new solutions involving two arbitrary functions of one variable. A successive iteration of such transforms results in a class of exact solutions characterized by an infinite number of parameters.
Almost 20 years after formulation of the parametric method, a new approach to constructing exact solutions in gravity (the so–called, anholonomic frame method) was proposed and developed in works [6, 7, 8, 9, 3, 10, 11, 12, 13, 14, 15]. One of its distinguished properties is that the existence of Killing symmetry is not crucial for definition of moving anholonomic frames222we shall use both equivalent terms anholonomic and/or nonholonomic; here we note that a local basis is nonholonomic if its vectors do not commute like for the coordinate bases but satisfy some anholonomy relations, see section 2. The first publication [6] contained certain examples of generic off–diagonal exact solutions, in three and four dimensional gravity (in brief, we shall write respectively 3D and 4D). The idea was to take any well–known exact solution (of black hole, instanton or monopole … type) which can be diagonalized with respect to a corresponding coordinate frame and then to deform it by introducing generic off–diagonal metric terms333parametrizing, for instance, certain 2D or 3D solitonic waves; such metrics can not be diagonalized by coordinate transforms in a manner that generates new classes of exact solutions. We note that one could be constructed source–free solutions and more general ones with matter fields, or with string gravity corrections, when extra dimensions and nontrivial torsion fields are considered. Various classes of such solutions were analyzed [7, 8, 9, 3, 10] (they describe nonholonomic deformations of Taub – NUT spaces, locally anisotropic wormholes, black ellipsoid and toroidal configurations, self–consistent interactions of (non)commutative Dirac and/or solitonic gravitational waves…).
The anholonomic frame method works as follows. We take a ’primary’ metric in a 5D (or 4D/ 3D) spacetime. The constructions are more simple if this metric is at least a conformal transform of a well known exact solution with diagonal metric. As a matter of principle, we can consider that the primary metric is a general one on a Riemann–Cartan manifold, not being obligatory a solution of gravitational field equations. By anholonomic frame (vielbein; or vierbein/ tetradic, in 4D) deformations, the primary metric and linear connection structures are transformed into the corresponding ’target’ ones for which the Einstein equations are exactly integrable. We note that the target metrics are generic off–diagonal, depend on classes of integration constants and arbitrary functions on one, two and three/four local coordinates (respectively for 4D/5D spacetimes).
It should be emphasized that the nonholonomic deformations induce nontrivial torsion structures, which can be effectively exploited in string/ brane gravity where the antisymmetric torsion plays a corn–stone role. The method can be applied in a straightforward form to some general classes of generic off–diagonal metrics and linear connections with nontrivial torsion. Haven being constructed certain classes of exact solutions with integral varieties parametrized by a some classes of integration functions, it is possible to constrain the set of such functions when the so–called canonical d–connection (with nontrivial torsion) transforms into the Levi–Civita connection (with vanishing torsion). This way, some more general classes of ’nonholonomic’ solutions can be restricted to define exact solutions in Einstein gravity. Nevertheless, even the metrics defining Einstein spaces depend on various types of integration functions and possess general nonlinear symmetries.
Sure, many of the off–diagonal solutions generated following the anholonomic frame method have no obvious physical interpretation. It is quite a cumbersome task to define the nonlinear symmetries of such spacetimes and decide what kind of physical interpretation may be adequate. Nevertheless, if any initial physical situations were given, it is possible to analyze if such nonholonomic deformations can preserve certain similarity and admit nonlinear superpositions and any new prescribed properties. We distinguish here five special cases (preliminary analyzed in our previous works): 1) The generic off–diagonal metric terms effectively polarize the constants of the primary metric (for instance, the point mass and/or electric, or cosmological constants). 2) The existing horizons (if any) are slightly deformed, for instance, from the spherical to an ellipsoidal symmetry. 3) The symmetry of former solutions can be broken in a spacetime region. 4) One can be changing of topological structure but certain former physical properties and analogy are preserved. 5) The primary solution is imbedded into a nontrivial background (for instance, consisting from a superposition of solitonic and pp–waves).
A rigorous analysis is necessary in order to state what kind of prescribed spacetimes can be generated by a class of nonholonomic transforms from a primary metric. Nevertheless, at least for certain classes of ’small’ smooth deformations, we can conclude that the singular properties (of the curvature scalar and tensor) and topology are preserved even additional nonlinear interactions are present and the symmetries are deformed. In such cases, it is possible to preserve the former physical interpretation but with modified constants, deformed horizons and nontrivial backgrounds.
In order to decide if a new class of generic off–diagonal solutions have nontrivial physical limits to the Einstein gravity, we must take into account various type of black hole ’uniqueness’ theorems and cosmic censorship criteria [16, 17, 18, 19, 20]. The strategy to deal with such solutions is to chose certain type of integration functions and boundary conditions when ’far away’ from the ’slightly’ deformed horizons and finite spacetime regions with nonlinear polarizations of constants and nontrivial backgrounds the Minkowski asymptotic and spherical topology hold true. Here we note that the off–diagonal metric terms, for vacuum solutions, may model certain effective matter field interactions (like in the Kaluza–Klein gravity but, in our case, without linearization for inducing electromagnetic fields and compactification on extra dimension coordinate). In this case, there are introduced the so–called geometric spacetime distorsions [21, 22] (like matter field distorsions for black holes [23, 24, 25]) and the restrictions of the uniqueness theorems and censorship criteria may be avoided. In modern gravity, the solutions with possible violation of mentioned type theorems and criteria and even of local Lorentz symmetry also present a special interest.
The parametric method can be applied if the (peseudo) Riemannian spacetime possesses at least one Killing vector. In the case when there are two Killing vectors, one can be defined an iteration procedure of generating classes of exact solutions involving arbitrary functions on spacetime coordinates labelled by an infinite family of parameters (such parameters are not spacetime coordinates). The set of such parameters can be treated as a specific space of internal symmetries of the solutions of vacuum Einstein equations but there is not clear the complete physical significance of such symmetries. For application of the anholonomic frame method, it is not crucial that the primary metric is a solution with Killing symmetries. The most important point is to define a nonholonomic frame deformation to a special type, off–diagonal, ansatz solving the Einstein equations for a connection deforming ’minimally’ the Livi–Civita connection in order to include the contribution of anholonomy coefficients. Certain constraint on integral varieties of such solutions allow to generate usual Einstein spaces and their generalizations with matter sources and, for instance, string contributions. Such classes of solutions, in general, are not characterized by a a group of parameters. Nevertheless, a number of commutative and noncommutative, Lie algebroid444Lie algebroids can be considered as certain generalizations of spaces with generalized Lie algebra symmetries when, roughly speaking, the structure constants depend on basic manifold coordinates and certain singular maps, anchors, are introduced into consideration, see Ref. [26] for a detailed discussion on definition of such geometric structures as exact solutions in gravity and Clifford algebroid or other nonlinear symmetries can be prescribed for such metrics.
Because the target off–diagonal metrics generated by applying the anholonomic frame method positively do not depend on one spacetime coordinate (but certain coefficients of metrics depend, for instance, on four/three coordinates of 5D/4D spacetimes), for sure, the generated nonolonomic spacetimes555 A manifold is nonholonomic if it is provided with a nonintegrable distribution, for instance, with a preferred frame structure with associated nonlinear connection (such spacetimes are also called locally anisotropic), see Refs. [9, 27, 28] for basic references and applications in modern gravity. possesses a Killing vector symmetry. In this case, a parametric transform can be applied after an anholonomic frame generation of a vacuum spacetime and the resulting vacuum solution will be characterized both by nonholonomic and parametric group structures. If one of the primary or target solutions is at least a conformal transform, or a small nonholonomic deformation of a well known exact solution of physical importance, we can formulate the criteria when certain prescribed geometrical and physical properties are preserved or may be induced. For instance, we can generate black hole, wormhole,… solutions with locally anisotropic parameters, deformed horizons and propagating in nontrivial solitonic/pp–wave backgrounds when the physical parameters and geometrical objects are parametrized by an infinite number of group parameters and possess generalized (for instance, Lie algebroid) symmetries.
The goals of this work is to show how new classes of exact solutions can be constructed by superpositions of the parametric and anholonomic frame transform and to carry out a program of extracting physically valuable solutions. We shall emphasize the possibility to select physically important solutions in Einstein and string gravity.
The paper has the following structure: in Sec. 2, we outline two geometric methods of constructing exact solutions in modern gravity. We begin with new geometric conventions necessary for a common description both of the parametric and anholonomic frame methods. The constructions distinguish the approaches related to Killing vectors and to the formalism of anholonomic frames with associated nonlinear connection structure. Then we formulate the techniques of constructing solutions for five and four dimensional (generic off–diagonal) metric ansatz and analyze the conditions when Einstein foliations can be defined by such solutions.
Section 3 is devoted to the main goal of this paper: elaboration of a unified formalism both for the parametric and anholonomic frame methods of constructing solutions in modern gravity. We start with the geometry of nonholonomic deformations of metrics resulting in exact solutions. Then we study superpositions of parametric transforms and anholonomic maps. Finally, there are proposed two alternative constructions when a class of solutions generated by the parametric method is deformed nonholonomically to other ones and, inversely, when the parametric transform is applied to nonholonomic Einstein spacetimes.
In Sec. 4, we construct five classes of exact solutions of vacuum Einstein equations and with sources from string gravity, generated by superpositions of nonholonomic frame and parametric transforms. We briefly explain the computations and emphasize the conditions when physically valuable solutions can be extracted. In explicit form, such metrics are defined by superpositions of solitonic pp–waves interacting in nonholonomically deformed black hole solutions.
We conclude the paper with some comments and remarks in Sec. 5. The Appendix contains some necessary formulas on curvature, Ricci and Einstein tensors for the so–called canonical distinguished connection. An ansatz with antisymmetric torsion in string gravity is considered and a general solution for nonholonomically constrained components of Einstein equations is considered.
2 Outline of the Methods
In this section, we outline and compare both the parametric and the anholonomic frame methods of constructing exact solutions in gravity, see details in Refs. [4, 5] and [9, 26, 3].
2.1 Geometric conventions
Let us begin with some general notations to be used in this work. We consider a spacetime as a manifolds of necessary smooth class of dimension with and (the meaning of conventional splitting of dimensions will be explained in section 2.3), provided with a metric
[TABLE]
of any (pseudo) Euclidean signature and a linear connection satisfying the metric compatibility condition 666in this work, the Einstein’s summation rule on repeating ”upper–lower” indices will be applied if the contrary will not be stated The components of geometrical objects, for instance, and are defined with respect to a local base (frame) and its dual base (co–base, or co–frame) for which where ” denotes the interior product induced by and is the Kronecker symbol. For a local system of coordinates on (in brief, we can write respectively
[TABLE]
and
[TABLE]
for the indices run correspondingly values of type: and for any conventional splitting
Any local (vector) basis can be decomposed with respect to any other basis and by considering frame transforms,
[TABLE]
where the matrix is the inverse to It should be noted that an arbitrary basis is nonholonomic (equivalently, anholonomic) because, in general, it satisfies certain anholonomy conditions
[TABLE]
with nontrivial anholonomy coefficients For we get holonomic frames: for instance, if we fix a local coordinate basis,
Denoting by the covariant derivative along a vector field we can define the torsion
[TABLE]
and the curvature
[TABLE]
tensors of connection where we use ”by definition” symbol ”” and The components and are computed by introducing into respective formulas (2.6) and (2.7), see [9] and [26] for details and computations related to the system of denotations considered in this paper.
The Ricci tensor
[TABLE]
is constructed by contracting the first (upper) and the last (lower) indices of the curvature tensor. The scalar curvature is by definition the contraction with the inverse metric (being the inverse to the matrix ),
[TABLE]
and the Einstein tensor is introduced as
[TABLE]
The vacuum (source–free) Einstein equations are postulated
[TABLE]
The four dimensional (4D) general relativity theory is distinguished by the property that the connection is uniquely defined by the coefficients following the conditions of metric compatibility, and of zero torsion, This defines the so–called Levi–Civita connection we respectively label its curvature tensor, Ricci tensor, scalar curvature and Einstein tensor in the form and Modern gravity theories consider extra dimensions and connections with nontrivial torsion. We note that in this work we shall consider both nontrivial and trivial torsion configurations. The aim is to show not only how our methods can be applied to various types of theories (string/ gauge/ Einstein–Cartan/ Finsler gravity models) but also to follow a simplified computational formalism related to spaces with effective torsions induced by nonholonomic frame deformations.777The difference between ”boldfaced” and ”calligraphic” labels, respectively for operators on spaces provided with nonlinear connection structure and certain differential forms, will be explained in Sec. 2.3. Certain nontrivial limits to the vacuum Einstein gravity can be selected if we impose the conditions
[TABLE]
even, in general, we have Such cases are considered in Refs [9, 26, 3]. In this paper, we shall follow more restrictive conditions when and have the same components with respect to certain preferred bases, when the equality (2.12) can be satisfied for some very general classes of metric ansatz.
We shall use a left–up label ”” for a metric
[TABLE]
a metric being a solution of the Einstein equations (2.11) for a linear connection with possible torsion In order to emphasize that a metric is a solution of the vacuum Einstein equations, in any dimension for the Levi–Civita connection or for any metric compatible connection satisfying the conditions (2.12), we shall write
[TABLE]
where the left–low label will distinguish the geometric objects for the Ricci flat space defined by a Levi–Civita connection
2.2 The formalism related to Killing vectors
The first parametric method [4] proposes a scheme of constructing a one–parameter family of vacuum exact solutions (labelled by tilde ” and depending on a real parameter )
[TABLE]
beginning with any source–free solution with Killing vector symmetry satisfying the conditions (Einstein equations) and (Killing equations). We denote this ’primary’ spacetime as One has to follow the rule 888For our purposes, in order to elaborate and unified approach to the parametric and the anholonomic frame methods, we introduce a new system of denotations.: The class of metrics is generated by the transforms
[TABLE]
where the matrix is parametrized in the form when
[TABLE]
for
[TABLE]
A rigorous proof [4] states that the metrics (2.15) also define exact vacuum solutions with if and only if the values from (2.17), subjected to the conditions
[TABLE]
solve the equations
[TABLE]
where the Levi–Civita connection is defined by and is the absolutely antisymmetric tensor. The existence of solutions for (2.19) (Geroch’s equations) is guaranteed by the Einstein’s and Killing equations.
The first type of parametric transforms (2.16) can parametrized by a matrix with the coefficients depending functionally on solutions for (2.19). Fixing a signature and a local coordinate system on one can define a local frame of reference
[TABLE]
like in (2.4), for which
[TABLE]
We note that have to be constructed as a solution of a system of quadratic algebraic equations (2.21) for given values and In a similar form, we can determine
[TABLE]
when
[TABLE]
The method guarantees that the family of spacetimes is also vacuum Einstein but for the corresponding families of Levi–Civita connections In explicit form, the matrix of parametric transforms can be computed by introducing the relations (2.21), (2.23) into (2.16),
[TABLE]
where is inverse to
The second parametric method [5] was similarly developed which yields a family of new exact solutions involving two arbitrary functions on one variables, beginning with any two commuting Killing fields for which a certain pair of constants vanish (for instance, the exterior field of a rotating star). By successive iterating such parametric transforms, one generates a class of exact solutions characterized by an infinite number of parameters and involving arbitrary functions. For simplicity, in this work we shall apply only the first parametric method in order to generate other nonholonomically deformed vacuum Einstein spaces. The case with off–diagonal metrics and two Killing vectors is more special; it will be analyzed in our further works.
2.3 The anholonomic frame method
We outline the results necessary for elaborating an approach containing both the parametric transforms and nonholonomic frame deformations. In details, the anholonomic frame method is reviewed in Refs. [9, 3], see also Appendix to [26] containing proofs of basic theorems and formulas.
Let us consider a –dimensional manifold enabled with a prescribed frame structure (2.4) when frame transform coefficients depend linearly on values
[TABLE]
where and and are local coordinates. The geometric constructions will be adapted to a conventional splitting stated by a set of coefficients defining a nonlinear connection (N–connection) structure as a nonintergrable distribution
[TABLE]
with a conventional horizontal (h) subspace, (with geometric objects labelled by ”horizontal” indices and vertical (v) subspace (with geometric objects labelled by indices .999For simplicity, in this work, we shall not enter in the details of the formalism of N–connections and (pseudo) Riemannian and Riemann–Cartan spaces, and of the so–called N–anholonomic manifolds, considered in Refs. [9, 28, 26] and in the Introduction section of Ref. [3]. In an alternative way, for different classes of connections not related to solutions of the Einstein equations, the theory of nonholonomic manifolds and (pseudo) Riemannian foliations is considered in Ref. [27]. We shall use ”boldfaced” symbols in order to emphasize that certain spaces (geometric objects) are provided (adapted) with (to) a N–connection structure
The transforms (2.27) and (2.30) define a N–adapted frame (vielbein) structure
[TABLE]
and the dual frame (coframe) structure
[TABLE]
The vielbeins (2.33) satisfy the corresponding nonholonomy (equivalently, anholonomy) relations of type (2.5),
[TABLE]
with (antisymmetric, ) anholonomy coefficients
[TABLE]
We note that a distribution (2.31) is integrable, i.e. is a foliation, if and only if the coefficients defined by satisfy the condition In general, a spacetime with prescribed nonholonomic splitting into h- and v–subspaces can be considered as a nonholonomic manifold [9, 27, 28].
Let us consider a metric structure on
[TABLE]
defined by coefficients
[TABLE]
This metric is generic off–diagonal, i.e. it can not be diagonalized by any coordinate transforms if are any general functions.101010We note that our –coefficients depending nonlinearly on all coordinates are not those from Kaluza–Klein theories which consist a particular case when with further compactifications on coordinates . We can adapt the metric (2.36) to a N–connection structure induced by the off–diagonal coefficients in (2.37) if we impose that the conditions
[TABLE]
where are satisfied for the corresponding local basis (2.32). In this case where is inverse to and we can write the metric (2.37) in equivalent form, as a distinguished metric (d–metric) adapted to a N–connection structure111111We shall call some geometric objects, like tensors, connections,…, to be distinguished by a N–connection structure, in brief, d–tensors, d–connection, if they are stated by components computed with respect to N–adapted frames (2.32) and (2.33). In this case, the geometric constructions are elaborated in N–adapted form, i.e. they are adapted to the nonholonomic distribution (2.31).,
[TABLE]
where and The coefficients and are related by formulas
[TABLE]
or
[TABLE]
where the vielbein transform is given by matrices (2.27) with and
Any vector field on can be written in N–adapted form as a d–vector
[TABLE]
In a similar form we can ’N–adapt’ any tensor object and call it to be a d–tensor.
By definition, a d–connection is adapted to the distribution (2.31) and splits into h– and v–covariant derivatives, where and are correspondingly introduced as h- and v–parametrizations of the coefficients
[TABLE]
The components with the coefficients defined with respect to (2.33) and (2.32), completely define a d–connection on a N–anholonomic manifold
The simplest way to perform a local covariant calculus by applying d–connections is to use N–adapted differential forms and to introduce the d–connection 1–form when the N–adapted components of d-connection are computed following formulas
[TABLE]
where ” denotes the interior product. This allows us to define in N–adapted form the torsion (2.6),
[TABLE]
and curvature (2.7),
[TABLE]
The coefficients of torsion (2.45) of a d–connection (in brief, d–torsion) are computed with respect to N–adapted frames (2.33) and (2.32),
[TABLE]
where, for instance, and are respectively the coefficients of the –torsion and –torsion In a similar form, we can compute the coefficients of a curvature d–curvatures (see Appendix for the formulas for coefficients, proved in Refs. [9, 3]).
There is a preferred, canonical d–connection structure, on a N–anholonomic manifold constructed only from the metric and N–connection coefficients and satisfying the conditions and and It should be noted that, in general, the components and are not zero. This is an anholonomic frame (equivalently, off–diagonal metric) effect. Hereafter, we consider only geometric constructions with the canonical d–connection which allow, for simplicity, to omit ”hats” on d–objects.121212The preference to the canonical d—connection is motivated also by the fact that it is possible to solve the vacuum Einstein equations for very general ansatz for metric and N–connection structure just for this linear connection. Usually, we can restrict the integral varieties in order to generate solutions satisfying the conditions (2.12), i.e. to construct generic off–diagonal solutions in general relativity . We can verify by straightforward calculations that the linear connection with the coefficients defined
[TABLE]
where
[TABLE]
uniquely solve the conditions stated for the canonical d–connection.
The Levi–Civita linear connection uniquely defined by the conditions and is not adapted to the distribution (2.31). Let us parametrize its coefficients in the form
[TABLE]
where with respect to N–adapted bases (2.33) and (2.32)
[TABLE]
A straightforward calculus131313Such results were originally considered by R. Miron and M. Anastasiei for vector bundles provided with N–connection and metric structures, see Ref. [29] . Similar proofs hold true for any nonholonomic manifold with prescribed N–connection. shows that the coefficients of the Levi–Civita connection can be expressed in the form
[TABLE]
where are computed as in the second formula in (2.35).
For our purposes, it is important to state the conditions when both the Levi–Civita connection and the canonical d–connection (being of different geometric nature) may be defined by the same set of coefficients with respect to a fixed frame of reference. Following formulas (2.49) and (2.51), we obtain the component equality if
[TABLE]
(there are satisfied the integrability conditions and our manifold admits a foliation structure),
[TABLE]
and
[TABLE]
which, following the second formula in (2.49), is equivalent to
[TABLE]
We conclude this section with the remark that if the conditions (2.52), (2.53) and (2.55) hold true for the metric (2.36), equivalently (2.39), the torsion coefficients (2.47) vanish. This results in respective equalities of the coefficients of the Riemann, Ricci and Einstein tensors (the conditions (2.12) being satisfied) for two different linear connections.
2.4 Off–diagonal exact solutions
We consider a five dimensional (5D) manifold of necessary smooth class and conventional splitting of dimensions for and The local coordinates are labelled in the form for and and For our further purposes, we can consider that any coordinates from a set can be of (3D) space, time, or extra dimension (5th coordinate) type.
2.4.1 A five dimensional off–diagonal ansatz
The ansatz of type (2.39) is parametrized in the form
[TABLE]
with the coefficients defined by some necessary smooth class functions of type
[TABLE]
The off–diagonal terms of this metric, written with respect to the coordinate dual frame can be redefined to state a N–connection structure \mathbf{N}=[N_{i}^{4}=w_{i}(x^{k},v),$$N_{i}^{5}=n_{i}(x^{k},v)] with a N–elongated co–frame (2.33) parametrized as
[TABLE]
This funfbein is dual to the local basis
[TABLE]
We emphasize that the metric (2.56) does not depend on variable i.e. it possesses a Killing vector and distinguish the dependence on the so–called ”anisotropic” variable
Computing the components of the Ricci and Einstein tensors for the metric (2.56) (see main formulas in Appendix and details on tensors components’ calculus in Refs. [26, 3]), one proves that the Einstein equations (A.7) for a diagonal with respect to (2.57) and (2.58), source
[TABLE]
transform into this system of partial differential equations:
[TABLE]
where, for
[TABLE]
when the necessary partial derivatives are written in the form In the vacuum case, we must consider We note that we use a source of type (2.59) in order to show that the anholonomic frame method can be applied also for non–vacuum configurations, for instance, when and defining locally anisotropic configurations generated by an anisotropic cosmological constant, which in its turn, can be induced by certain ansatz for the so–called –field (absolutely antisymmetric third rank tensor field) in string theory [9, 3, 26], see formulas (A.10) and (A.11) and related explanations in Appendix. Here we note that the off–diagonal gravitational interactions can model locally anisotropic configurations even if or both values vanish.
Summarizing the results for an ansatz (2.37) with arbitrary signatures (where and and one proves [9, 26, 3] that any off–diagonal metric
[TABLE]
with the coefficients being of necessary smooth class and the indices with ”hat” running the values , where is a solution of the 2D equation (2.60) for a given source
[TABLE]
and the N–connection coefficients and are computed following the formulas
[TABLE]
and
[TABLE]
define an exact solution of the system of Einstein equations (2.60)–(2.63). It should be emphasized that such solutions depend on arbitrary nontrivial functions (with , and and sources Such values for the corresponding signatures have to be defined by certain boundary conditions and physical considerations.141414Our classes of solutions depending on integration functions are more general than those for diagonal ansatz depending, for instance, on one radial like variable like in the case of the Schwarzschild solution (when the Einstein equations are reduced to an effective nonlinear ordinary differential equation, ODE). In the case of ODE, the integral varieties depend on integration constants which can be defined from certain boundary/ asymptotic and symmetry conditions, for instance, from the constraint that far away from the horizon the Schwarzschild metric contains corrections from the Newton potential. Because our ansatz (2.56) results in a system of nonlinear partial differential equations (2.60–(2.63), the solutions depend not on integration constants, but on very general classes of integration functions. A similar situation is considered in the Geroch method but those solutions are also parametrized by sets of parameters not treated as local coordinates.
The ansatz of type (2.56) with but (or, inversely, but consist of more special cases and request a bit different method of constructing exact solutions. Nevertheless, such type solutions are also generic off–diagonal and they may be of substantial interest (the length of paper does not allow to include an analysis of such particular cases).
2.4.2 Four and five dimensional foliations and the Einstein spaces
The method of constructing 5D solutions with nontrivial torsion induced by anholonomy coefficients can be restricted to generate 4D nonholonomic configurations and generic off–diagonal solutions in general relativity. In order to consider reductions for the ansatz (2.56) we can eliminate from the formulas the variable and to consider a 4D space (parametrized by local coordinates trivially embedded into a 5D spacetime (parametrized by local coordinates with In this case, there are possible 4D conformal and anholonomic transforms depending only on variables of a 4D metric of arbitrary signature. To emphasize that some coordinates are stated just for a 4D space we might use ”hats” on the Greek indices, and on the Latin indices from the middle of the alphabet, when the local coordinates on are parametrized in the form for The 4D off–diagonal ansatz
[TABLE]
is written with respect to the anholonomic co–frame where
[TABLE]
is the dual of where
[TABLE]
and the coefficients are necessary smoothly class functions of type:
[TABLE]
In the 4D case, a source of type (2.59) should be considered without the component in the form
[TABLE]
The Einstein equations with source (2.73) for the canonical d–connection (2.49) defined by the ansatz (2.70) transform into a system of nonlinear partial differential equations very similar to (2.60)–(2.63). The difference for the 4D equations is that the coordinate is not contained into the equations and that the indices of type must be changed into the corresponding indices The generated classes of 4D solutions are defined almost by the same formulas (2.66), (2.68) and (2.69).
Now we describe how the coefficients of an ansatz (2.70) defining an exact vacuum solution for a canonical d–connecton can be constrained to generate a vacuum solution in Einstein gravity:
We start with the conditions (2.55) written (for our ansatz) in the form
[TABLE]
These equations for nontrivial values of and constructed for some defined values of and must be compatible with the equations (2.61)–(2.63) for One can be taken nonzero values for in (2.62) if and only if because the the equation (2.61) imposes the condition This is possible, for the sourceless case and if and only if
[TABLE]
see formula (2.65). A very general class of solutions of equations (2.74), (2.75) and (2.77) can be represented in the form
[TABLE]
where and is any function for which and is an arbitrary integration function.
The next step is to satisfy the integrability conditions (2.52) defining a foliated spacetimes provided with metric and N–connection and d–connection structures [26, 9, 3, 28] (we note that (pseudo) Riemannian foliations are considered in a different manner in Ref. [27]) for the so–called Schouten – Van Kampen and Vranceanu connections not subjected to the condition to generate Einstein spaces). It is very easy to show that there are nontrivial solutions of the constraints (2.52) which for the ansatz (2.70) are written in the form
[TABLE]
We solve these equations for if we take any two functions satisfying
[TABLE]
(it is possible for a particular class of integration functions in (2.69) when and are constraint to satisfy just the conditions (2.80)). Then we can consider any for which solve the equation (2.79). In a more particular case, one can be constructed solutions for any and and (or, inversely, for any and Here one should be also noted that the conditions (2.53) are solved in straightforward form by the ansatz (2.70).
We conclude that for any sets of respectively generated by functions and see (2.78), and satisfying (2.80), the generic off–diagonal metric (2.70) possesses the same coefficients both for the Levi–Civita and canonical d–connection being satisfied the conditions (2.12) of equality of the Einstein tensors. Here we note that any 2D metric can be written in a conformally flat form, i. e. we can chose such local coordinates when
[TABLE]
for signatures in (2.70).
Summarizing the results of this section, we can write down the generic off–diagonal metric (it is a 4D dimensional reduction of (2.66))
[TABLE]
defining vacuum exact solutions in general relativity if the coefficients are restricted to solve the equations
[TABLE]
for and where, for instance, and
We can generalize (2.4.2) similarly to (2.66) in order to generate solutions for nontrivial sources (2.73). In general, they will contain nontrivial anholonomically induced torsions. Such configurations may be restricted to the case of Levi–Civita connection by solving the constraints (2.74)–(2.76) in order to be compatible with the equations (2.61) and (2.62) for the coefficients and computed for and see formula (2.65), resulting in more general conditions than (2.77) and (2.78). Roughly speaking, all such coefficients are generated by any (or ) defined from (2.62) for prescribed values (or ) and The existence of a nontrivial matter source of type (2.73) does not change the condition see (2.76), necessary for extracting torsionless configurations. This mean that we have to consider only trivial solutions of (2.63) when two functions are subjected to the condition (2.79). We conclude that this class of exact solutions of the Einstein equations with nontrivial sources (2.73), in general relativity, is defined by the ansatz
[TABLE]
where the coefficients are restricted to satisfy the conditions
[TABLE]
for see (2.65), being compatible with (2.74) and (2.75), for given sources and We note that the second equation in (2.85) relates two functions and In references [6, 7, 10, 8, 9, 11, 26], we investigated a number of configurations with nontrivial two and three dimensional solitons, reductions to the Riccati or Abbel equation, defining off–diagonal deformations of the black hole, wormhole or Taub NUT spacetimes. Those solutions where constructed to be with trivial or nontrivial torsions but if the coefficients of the ansatz (2.4.2) are restricted to satisfy the conditions (2.85) in a compatible form with (2.74) and (2.75), for sure, such metrics will solve the Einstein equations for the Levi–Civita connection.
Finally, we emphasize that the ansatz (2.4.2) defines Einstein spaces with a cosmological constant if we put in (2.85).
3 Anholonomic Transforms and Killing Spacetimes
Anholonomic deformations can be defined for any primary metric and vielbein structures on a spacetime (as a matter of principle, the primary metric can be not a solution of the gravitational field equations). Such deformations always result in a target spacetime possessing one Killing vector symmetry if the last one is constrained to satisfy the Einstein equations for the canonical d–connection, or for the Levi–Civita connection. For such target spacetimes, we can always apply the parametric transform and generate a set of generic off–diagonal solutions labelled by a parameter (2.16). There are possible constructions when the anholonomic frame transforms are applied to a family of metrics generated by the Geroch method as new exact solutions of the vacuum Einstein equations, but such primary metrics have to be parametrized by certain type ansatz admitting anholonomic transforms to other classes of exact solutions.
3.1 Nonholonomic deformations of metrics
Let us consider a –dimensional manifold (spacetime) enabled with a metric structure distinguished in the form
[TABLE]
The local coordinates are parametrized for the indices of type (in brief, horizontal, or h–indices/ components) and (vertical, or v–indices/ components). We suppose that, in general, the metric (3.1) is not a solution of the Einstein equations but can be nonholonomically deformed in order to generate exact solutions. The coefficients from (3.1) state a conventional –splitting in any point and define a class of ’N–adapted’ local bases
[TABLE]
and local dual bases (co–frames) when
[TABLE]
for i.e. where the inner product is denoted by ’’ and the Kronecker symbol is written The vielbeins (3.2) satisfy the nonholonomy (equivalently, anholonomy) relations
[TABLE]
with nontrivial anholonomy coefficients
[TABLE]
A metric parametrized in the form
[TABLE]
is a nonhlonomic transform (deformation), preserving the –splitting, of the metric, if the coefficients of (3.1) and (3.6) are related by formulas
[TABLE]
where the summation rule is not considered for the indices of gravitational ’polarizations’ and in (3.7). For nontrivial values of the nonholonomic frames (3.2) and (3.3) transform correspondingly into
[TABLE]
and
[TABLE]
with the anholonomy coefficients defined by (2.35).
We emphasize that in order to generate exact solutions, the gravitational ’polarizations’ and in (3.7) are not arbitrary functions but restricted in such a form that the values
[TABLE]
define an ansatz of type (2.66), or (2.4.2) (for vacuum configurations) and (2.4.2) for nontrivial matter sources and
Any nonholonomic deformation
[TABLE]
can be described by two vielbein matrices of type (2.27),
[TABLE]
generating the d–metric see formula (2.40), and
[TABLE]
generating the d–metric (3.10).
If the metric and N–connection coefficients (3.7) are stated to be those from an ansatz (2.66) (or (2.4.2)), we should write (or ) and say that the metric (3.1) was nonholonomically deformed in order to generate an exact solution of the Einstein equations for the canonical d–connection (or in a restricted case, for the Levi–Civita connection). In general, such metrics have very different geometrical and (if existing) physical properties. Nevertheless, at least for some classes of ’small’ nonsingular nonholonomic deformations, it is possible to preserve a similar physical interpretation by introducing small polarizations of metric coefficients and deformations of existing horizons, not changing the singular structure of curvature tensors. We shall construct explicit examples and discuss the details in Section 4.
3.2 Superpositions of the parametric transforms and anholonomic
deformations
As a matter of principle, any first type parametric transform can be represented as a generalized anholonomic frame transform labelled by an additional parameter. It should be also noted that there are two possibilities to define superpositions of the parametic transforms and anholonomic frame deformations both resulting in new classes of exact solutions of the vacuum Einstein equations. In the first case, we start with a parametric transform and, in the second case, the anholonomic deformations are considered from the beginning. The aim of this section is to examine such possibilities.
3.2.1 The parametric transforms as generalized anholonomic
deformations
We note that any metric defining an exact solution of the vacuum Einstein equations can be represented in the form (3.1). Then, any metric (2.16) from a family of new solutions generated by the first type parametric transform can be written as (3.6) and related via certain polarization functions of type (3.7), in the parametric case depending on parameter i.e. and Roughly speaking, the parametric transform can be represented as a generalized class of anholonomic frame transforms additionally parametrized by and adapted to preserve the –splitting structure.151515It should be emphasized that such constructions are not trivial, for usual coordinate transforms, if at least one of the primary or target metrics is generic off-diagonal. The corresponding vielbein transforms (2.21) and (2.23) are parametrized, respectively, by matrices of type (3.12) and (3.13), also ”labelled” by Such nonholonomic parametric deformations
[TABLE]
are defined by the vielbein matrices,
[TABLE]
generating the d–metric and
[TABLE]
generating the d–metric Using the matrices (3.15) and (3.16), we can compute the matrix of parametric transforms
[TABLE]
like in (2.24), but for ”boldfaced’ objects, where is inverse to 161616we use a ”boldface” symbol because in this case the constructions are adapted to a –splitting and define the target set of metrics in the form
[TABLE]
At first site, there are two substantial differences from the case of usual anholonomic frame transforms (3.11) and the case of parametric deformations (3.14). The first one is that the metric was not constrained to be an exact solution of the Einstein equations like it was required for The second one is that even can be restricted to be an exact vacuum solution, generated by a special type of deformations (3.10), in order to get an ansatz of type (2.4.2), an arbitrary metric from a family of solutions will not be parametrized in a form that the coefficients will satisfy the conditions (2.83). Nevertheless, even in such cases, we can consider additional nonholonomic frame transforms when is transformed into an exact solution and any particular metric from the set will be deformed into an exact solution defined by an ansatz (2.4.2) with additional dependence on
The first result of this section is that, by superpositions of nonholonomic deformations, we can always parametrize a solution formally constructed following the Geroch method (from a primary solution depending on variables in the form
[TABLE]
with the coefficients restricted to solve the equations (2.83) but depending additionally on parameter
[TABLE]
for and where, for instance, and
The second result of this section is that if even, in general, any primary solution can not be parametrized as an ansatz (2.4.2), it is possible to define nonholonomic deformations to such a generic off–diagonal ansatz or any defined by an ansatz (3.1), which in its turn can be transformed into a metric of type (3.19) without dependence on 171717in our formulas we shall not point dependencies on coordinate variables if that will not result in ambiguities
Finally, we emphasize that in spite of the fact that both the parametric and anholonomic frame transforms can be parametrized in very similar forms by using vielbein transforms there is a criteria distinguishing them one from another: For a ”pure” parametric transform, the matrix and related \widetilde{\mathbf{A}}_{\alpha}^{\ \underline{\alpha}}\andare generated by a solution of the Geroch equations (2.19). If the ”pure” nonholonomic deformations, or their superposition with a parametric transform, are introduced into consideration, the matrix (3.13), or its generalization to a matrix (3.16), can be not derived only from solutions of (2.19). Such transforms define certain, in general, nonintegrable distributions related to new classes of Einstein equations.
3.2.2 Parametric transforms of anholonomically generated solutions
and two parameter transforms
First, let us consider an exact vacuum solution (2.4.2) in Einstein gravity generated following the anholonomic frame method. Even through it is generic off–diagonal and depends on various types of integration functions and constants, it is obvious that it possesses at least a Killing vector symmetry because the metric does not depend on variable We can apply the first type parametric transform to a metric generated by anholonomic deforms (3.11). If we work in a coordinate base with the coefficients of defined in the form we generate a set of exact solutions
[TABLE]
see (2.16), were the transforms (2.24), labelled by a parameter are not adapted to a nonholonomic –splitting. We can elaborate N–adapted constructions starting with an exact solution parametrized in the form (3.6), for instance, like , with being of type (3.13) with coefficients satisfying the conditions (3.10). The target ’boldface’ solutions are generated as transforms
[TABLE]
where
[TABLE]
like in (2.24), but for ”boldfaced’ objects, the matrix is inverse to
[TABLE]
and the matric is considered
[TABLE]
where and are gravitational polarizations of type (3.7).181818we do not summarize on repeating two indices if they both are of lower/ upper type Here it should be emphasized that even are exact solutions of the vacuum Einstein equations they can not be represented by ansatz of type (3.19), with because the mentioned polarizations were not constrained to be of type (3.10) and satisfy any conditions of type (3.20).191919As a matter of principle, we can deform nonholonomically any solution from the family to an ansatz of type (3.19).
Now, we prove that by using superpositions of nonholonomic and parametric transforms we can generate two parameter families of solutions. This is possible, for instance, if the metric form (3.22), in its turn, was generated as an ansatz of type (3.19) from another exact solution We write
[TABLE]
and define the superposition of transforms
[TABLE]
It can be considered an iteration procedure of nonholonomic parametric transforms of type (3.27) when an exact vacuum solution of the Einstein equations is related via a multi –parameters vielbein map with another prescribed vacuum solution. Using anholonomic deformations, one introduces (into chains of such transforms) certain classes of metrics which are not exact solutions but nonholonomically deformed from, or to, some exact solutions.
Finally, we briefly discuss the symmetry properties of such anholonomic multi –parameter solutions of the Einstein equations. In the parameter space, they possess symmetries with infinite dimensional parametric group structures [4, 5] but with respect to anholonomic deforms one can be considered various types of prescribed Lie algebroid, solitonic, pp–wave and/or nonholonomic noncommutative symmetries [9, 26, 3]. In general, many of such way generated solutions do not have obvious physical interpretation. Nevertheless, if certain small (non–coordinate) parameters of nonholonomic deformations are introduced into consideration, it is possible to prescribe various interesting physical situations for a subset of metrics generated by maps of type (3.27), preserving certain similarities with a primary solution. We construct and analyze some examples of such solutions in the next section.
4 Examples of Off–Diagonal Exact Solutions
The purpose of this section is to present explicit examples of how superpositions of nonholonomic deformations and parametric transforms can be applied in order to generate new classes of solutions and how physically valuable configurations can be selected. Some constructions will be performed for 5D spacetimes with torsion, for instance, related to the so–called (antisymmetric) ”H–fields” in string gravity but the bulk of them will be restricted to define usual 4D Einstein spacetimes with generic off–diagonal metrics.
4.1 Five classes of primary metrics
We begin with a list of 5D quadratic elements (defined by certain primary metrics) which will be used for generating new classes of exact solutions following superpositions of nonholonomic deformations and parametric transforms:
The fist quadratic element, defined as a particular case of metric (3.1), is
[TABLE]
where the local coordinates and nontrivial metric coefficients are parametrized in the form
[TABLE]
for
[TABLE]
For the constants and being a point mass, the element (4.1) defines just a trivial embedding into 5D (with extra dimension coordinate of the Schwarzschild solution written in spacetime spherical coordinates 202020For simplicity, we consider only the case of vacuum solutions, not analyzing a more general possibility when is related to the electric charge for the Reissner–Nordström metric (see, for example, [19]). In our further considerations we shall treat as a small parameter, for instance, defining a small deformation of a circle into an ellipse (eccentricity).
The second quadratic element
[TABLE]
where the local coordinates are
[TABLE]
for
[TABLE]
and the Schwarzschild radius of a point mass is defined where is the 4D Newton constant and is the light velocity. The nontrivial metric coefficients in (4.4) are parametrized
[TABLE]
The quadratic element defined by (4.4) and (4.7) is a trivial embedding into 5D of the Schwarzschild quadratic element multiplied to the conformal factor We emphasize that this metric is not a solution of the Einstein equations but it will be used in order to construct nonholonomic deformations and parametric transforms to such solutions.
We shall use a quadratic element when the time coordinate is considered to be ”anisotropic”,
[TABLE]
where the local coordinates are
[TABLE]
and the nontrivial metric coefficients are parametrized
[TABLE]
The formulas (4.8) and (4.10) are respective reparametrizations of (4.4) and (4.7) when the forth and fifth coordinates are inverted. Such metrics will be used for constructing new classes of exact solutions in 5D with explicit dependence on time like coordinate.
The forth quadratic element is introduced by inverting the forth and fifth coordinates in (4.1) (having the same definitions as in that case)
[TABLE]
where the local coordinates and nontrivial metric coefficients are parametrized in the form
[TABLE]
Such metrics can be used for constructing exact solutions in 4D gravity with anisotropic dependence on time coordinate.
Finally, we consider
[TABLE]
where the local coordinates are
[TABLE]
and the nontrivial metric coefficients are parametrized
[TABLE]
The metric (4.13) is a trivial embedding into 5D of the vacuum solution of the Einstein equation defining pp–waves [30] for any solving
[TABLE]
with and where are usual Cartesian coordinates and is the time like coordinates. The simplest explicit examples of such solutions are
[TABLE]
defining a plane monochromatic wave, or
[TABLE]
defining a wave packet travelling with unit velocity in the negative direction.
4.2 Solitonic pp–waves and string torsion
Pp–wave solutions are intensively exploited for elaborating string models with nontrivial backgrounds [31, 32, 33]. A special interest for pp–waves in general relativity is the fact that any solution in this theory can be approximated by a pp–wave in vicinity of horizons. Such solutions can be generalized by introducing nonlinear interactions with solitonic waves [34, 35, 36, 6, 10] and nonzero sources with nonhomogeneous cosmological constant induced by an ansatz for the antisymmetric tensor fields of third rank. A very important property of such nonlinear wave solutions is that they possess nontrivial limits defining new classes of generic off–diagonal vacuum Einstein spacetimes.
4.2.1 Pp–waves and nonholonomic solitonic interactions
Let us consider the ansatz
[TABLE]
where the local coordinates are
[TABLE]
and the nontrivial metric coefficients and polarizations are parametrized
[TABLE]
For trivial polarizations and the metric (4.18) is just the pp–wave solution (4.13).
Exact solitonic pp–wave solutions in string gravity:
Our aim is to define such nontrivial values of polarization functions when is defined by a 3D soliton , for instance, a solution of
[TABLE]
see formula (B.1) in Appendix, and is a solution of (2.60),
[TABLE]
The solitonic deformations of the pp–wave metric will define exact solutions in string gravity with –fields, see in Appendix the equations (A.10) and (A.11) for the string torsion ansatz (A.12).212121as a matter of principle we can consider that is a solution of any 3D solitonic, or other, nonlinear wave equation.
Introducing the above stated data for the ansatz (4.18) into the equation (2.61), we get an equation relating and Such solutions can be constructed in general form, respectively, following formulas (B.4) and (B.5) (in this section, we take We obtain
[TABLE]
and
[TABLE]
where is an integration function. Having defined the coefficients we can solve the equations (2.62) and (2.63) in a form similar to (2.68) and (2.69) but expressing the solutions through and defined by pp– and solitonic waves as in (4.23) and (4.22). The corresponding solutions are
[TABLE]
for and
[TABLE]
where and are integration functions.
We conclude that the ansatz (4.18) with the coefficients computed following the equations and formulas (4.21), (4.23), (4.22), (4.24) and (4.25) define a class of exact solutions (depending on integration functions) of gravitational field equations in string gravity with –field. In a more explicit form, depending on above stated functions and and respective integration functions, the class of such solutions is parametrized as
[TABLE]
where some constants and multiples depending on and included into It should be noted that such spacetimes possess nontrivial nonholonomically induced torsion (we omit explicit formulas for the nontrivial components which can be computed by introducing the coefficients of our ansatz into formulas (2.49) and (2.47)). This is a very general class of solutions describing nonlinear interactions of pp–waves and 3D solutions in string gravity. The term can be eliminated in order to describe only 4D configurations. Nevertheless, in this case, there is not a smooth limit of such 4D solutions for to those in general relativity.
Pp–waves and solitonic interactions in vacuum Einstein gravity:
We prove that the anholonomic frame method can be used in a different form in order to define 4D metrics induced by nonlinear pp–waves and solitonic interactions for vanishing sources and the Levi–Civita connection. We can apply the formulas (2.77), (2.78), (2.79) and (2.80), for simplicity, considering that and being a function generating solitonic and pp–wave interactions. For an ansatz of type (4.18), we write
[TABLE]
A 3D solitonic solution can be generated if is subjected to the condition to solve a solitonic equation, like in (4.20). It is not possible to satisfy the integrability conditions (2.79) for any