The principle of equivalence and theories of gravity

We construct classical theories of gravity on the basis of special relativity and the Einstein-Infeld accelerating-elevator thought experiment. The resulting theories share most of the main features of general relativity, namely the nonlinear character of the theory, the metrical significance of the gravitational potentials and the geodesic equation of particle motion. They differ from general relativity in at most nonlinear terms in the gravitational constant G in their equations of particle motion and field equations.     

Violation of the equivalence principle in modified theories of gravity

We show that the metric in f(R) theories of gravity in Palatini formalism can be solved as the product of a rank-two tensor times a scalar function which is very sensitive to the local energy-momentum densities. This local dependence of the metric generates new gravitationally-induced microscopic interactions, which eventually would lead to self-accelerated test body trajectories. These facts make very unlikely the viability of Palatini f(R) models designed to change the late-time cosmic evolution.     

Scalar-tensor theories of gravitation: Foundations and prospects

A generalization of Einstein's theory is discussed in which the gravitation is described by a tensor and a scalar field. The theory is more consistent with Mach's principle and less reliant on absolute properties of space. The modification involves a violation of the “strong principle of equivalence” on which Einstein's theory is based. In the original version of this new theory, the “constant” of gravitationG is varying and particle masses are fixed. Later on another version of the theory was given in whichG is truly a constant and the particle masses vary. The two versions are related by a conformal transformation. The physical and mathematical foundations of this theory have been discussed and the field equations have been derived. The astrophysical and cosmological consequences of the theory have been elaborately reviewed.     

Constraints on unified theories from the experimental tests of the equivalence principle

The predictions of a general unified theory for the gravitational, electromagnetic and scalar field are compared with the results of the experimental tests of the equivalence principle. It is shown that the theoretical predictions do not disagree with experimental data provided that the coupling of the scalar to the electromagnetic field is suppressed by a factork 10^–3, or, alternatively, the scalar field is massive; in this case, a lower limit for its mass is obtained.     

Questioning the equivalence principle

The Equivalence Principle (EP) is not one of the ‘universal’ principles of physics (like the action principle). It is a heuristic hypothesis which was introduced by Einstein in 1907, and used by him to construct his theory of general relativity. In modern language, the (Einsteinian) EP consists in assuming that the only long-range field with gravitational-strength couplings to matter is a massless spin-2 field. Modern unification theories, and notably string theory, suggest the existence of new fields (in particular, scalar fields: ‘dilaton’ and ‘moduli’) with gravitational-strength couplings. In most cases the couplings of these new fields ‘violate’ the EP. If the field is long-ranged, these EP violations lead to many observable consequences (variation of ‘constants’, non-universality of free fall, relative drift of atomic clocks, etc.). The best experimental probe of a possible violation of the EP is to compare the free-fall acceleration of different materials.     

Physical equivalence of theories

A means of distinguishing between equivalent and different (inequivalent) theories is presented. It is applied to distinguish between the Brans-Dicke theory and general relativity. It is shown that an infinite number of Brans-Dicke-type extensions can be constructed for all field theories.This work was performed while on leave of absence from the Mathematics Department, Islamabad University, Islamabad, Pakistan.     

The equivalence principle

The experimental basis of the equivalence principle is reviewed, and the implications for the gravitational interactions of elementary particles are studied within a special relativistic framework. The gravitational red shift is treated in detail and is used to show that antiparticles also obey the equivalence principle. The profound consequences of a violation of the equivalence principle are discussed.     

Gravitational redshift and the equivalence principle

Two problems have long been confused with each other: the gravitational redshift as discussed by the equivalence principle; and the Doppler shift observed by a detector which moves with constant proper acceleration away from a stationary source. We here distinguish these two problems and give for the first time a solution of the former which is exact within the context of the equivalence principle in a sense discussed in the paper. The equivalence principle leads to transformations between flat spacetimes. These are analyzed, and a generalized Lorentz transformation is proposed which covers transformations from inertial to uniformly accelerated frames of reference.     

VariableG and the strong equivalence principle

A possible time variability ofG, implying a violation of the strong equivalence principle, was first proposed by P. A. M. Dirac in 1937. Since such a feature cannot be accommodated within either Newton’s or Einstein’s theories, a new theoretical framework is needed. In this paper we review one such possible scheme, the scale covariant theory, within which the consequences of a variableG on geophysics, astrophysics, and cosmology can be treated consistently. The global verdict is thatG may have varied by as much as a factor of 25 since the time of nucleosynthesis, without any disagreement emerging in any case. In spite of this result, we are not entitled to conclude from our analysis that a variableG has been shown to exist or that it is needed, but only that its variation iscompatible with known data. The proof thatG varies can in fact only come from direct observations. However, since the previous analyses had concluded that aG(t) would entail severe discrepancies with known data, the reversal of the verdict is believed to be significant, since it may hopefully spur new observational interest in this basic problem. Presented at the Dirac Symposium, Loyola University, New Orleans, May 1981.     

ADM electrons and the equivalence principle

Noting that the general relativistic ADM equation for the mass of a sphere of charged dust (with no angular momentum) reveals that the masses of point-like particles are determined solely by their electrical charge, electron models based on extended spheres of such purely electrical dust are examined. It is shown that for all realistic electron models of this type (where the observed electron mass is positive and many orders of magnitude smaller than either the Planck or ADM mass) the electron's bare active gravitational mass must be taken to be negative. Because of the negativity of the bare active gravitational mass, one of the two realistic models leads to a violation of the weak equivalence principle, but the other does not. A means of testing whether negative mass obeys the equivalence principle is mentioned.     

Energy conservation and the principle of equivalence

If the equivalence principle is violated, then observers performing local experiments can detect effects due to their position in an external gravitational environment (preferred-location effects) or can detect effects due to their velocity through some preferred frame (preferred-frame effects). We show that the principle of energy conservation implies a quantitative connection between such effects and structure-dependence of the gravitational acceleration of test bodies (violation of the Weak Equivalence Principle). We analyze this connection within a general theoretical framework that encompasses both non-gravitational local experiments and test bodies as well as gravitational experiments and test bodies, and we use it to discuss specific experimental tests of the equivalence principle, including non-gravitational tests such as gravitational redshift experiments, Eötvös experiments, the Hughes-Drever experiment, and the Turner-Hill experiment, and gravitational tests such as the lunar-laser-ranging “Eötvös” experiment, and measurements of anisotropies and variations in the gravitational constant. This framework is illustrated by analyses within two theoretical formalisms for studying gravitational theories: the PPN formalism, which deals with the motion of gravitating bodies within metric theories of gravity, and the TH?μ formalism that deals with the motion of charged particles within all metric theories and a broad class of non-metric theories of gravity.     

Quantum aspects of the equivalence principle

Two thought experiments are discussed which suggest, first, a geometric interpretation of the concept of a (say, vector) potential (i.e., as a kinematic quantity associated with a transformation between moving frames of reference suitably related to the problem) and, second, that, in a quantum treatment one should extend the notion of the equivalence principle to include not only the equivalence of inertial forces with suitable real forces, but also the equivalence of potentials of such inertial forces and the potentials of suitable real forces. The two types of cancellation are physically independent of each other, because of the Aharonov-Bohm effect. Finally, we show that the latter effect itself can be understood geometrically as a kinematic effect arising upon the transformation between the two reference frames.On leave of absence from the Department of Physics, Tel-Aviv University, Israel, and the Department of Physics, Yeshiva University, New York.Supported by the NSF under Contract GP-14911.     

Physics with and without the equivalence principle

A differential manifold (d-manifold, for short) can be defined as a pair (M, C), where M is any set and C is a family of real functions on M which is (i) closed with respect to localization and (ii) closed with respect to superposition with smooth Euclidean functions; one also assumes that (iii) M is locally diffeomorphic to Rⁿ. These axioms have a straightforward physical interpretation. Axioms (i) and (ii) formalize certain compatibility conditions which usually are supposed to be assumed tacitly by physicists. Axiom (iii) may be though of as a (nonmetric) version of Einstein's equivalence principle. By dropping axiom (iii), one obtains a more general structure called a differential space (d-space). Every subset of Rⁿ turns out to be a d-space. Nevertheless it is mathematically a workable structure. It might be expected that somewhere in the neighborhood of the Big Bang there is a domain in which space-time is not a d-manifold but still continues to be a d-space. In such a domain we would have a physics without the (usual form of the) equivalence principle. Simple examples of d-spaces which are not d-manifolds elucidate the principal characteristics the resulting physics would manifest.on leave of absence from the Institute of Nuclear Physics, Department of Theoretical Physics, ul. Radzikowskiego 152, 31–342 Cracow, Poland.     

Neutrino oscillations in non-inertial frames and the violation of the equivalence principle Neutrino mixing induced by the equivalence principle violation

Neutrino oscillations are analyzed in an accelerating and rotating reference frame, assuming that the gravitational coupling of neutrinos is flavor dependent, which implies a violation of the equivalence principle. Unlike the usual studies in which a constant gravitational field is considered, such frames could represent a more suitable framework for testing if a breakdown of the equivalence principle occurs, due to the possibility to modulate the (simulated) gravitational field. The violation of the equivalence principle implies, for the case of a maximal gravitational mixing angle, the presence of an off-diagonal term in the mass matrix. The consequences on the evolution of flavor (mass) eigenstates of such a term are analyzed for solar (oscillations in the vacuum) and atmospheric neutrinos. We calculate the flavor oscillation probability in the non-inertial frame, which does depend on its angular velocity and linear acceleration, as well as on the energy of neutrinos, the mass-squared difference between two mass eigenstates, and on the measure of the degree of violation of the equivalence principle (). In particular, we find that the energy dependence disappears for vanishing mass-squared difference, unlike the result obtained by Gasperini, Halprin, Leung, and other physical mechanisms proposed as a viable explanation of neutrino oscillations. Estimations on the upper values of are inferred for a rotating observer (with vanishing linear acceleration) comoving with the earth, hence rad/sec, and all other alternative mechanisms generating the oscillation phenomena have been neglected. In this case we find that the constraints on are given by for solar neutrinos and for atmospheric neutrinos. Received: 14 December 2000 / Published online: 15 March 2001