¡Descarga PointParticle y más Apuntes en PDF de Matemáticas solo en Docsity!
The Newtonian Point Particle
Bartolom´e Coll†
Gravitation et Cosmologie Relativistes CNRS - Universit´e Paris VI 4, place Jussieu. F-75252 Paris Cedex 05, France E-mail: [email protected]
and
Joan Josep Ferrando
Departament d’Astronomia i Astrof´ısica Universitat de Valencia E-46100 Burjassot (Valencia), Spain E-mail: [email protected]
Abstract Local inverse questions for Newtonian gravitation remain unanswered. Restricted to one particle, these questions are: i) What are the local neces- sary and sufficient conditions for an acceleration field to be the force field of a sole point particle?, and ii) What are the mass and the position of the particle corresponding to such an acceleration field? Their answer is given for both, inertial and accelerated observers. For the last ones, this is made through a characterization of inertial acceleration fields. The results are covariant, intrinsic and constructive, i.e., they are coordinate- free, expressed in terms of the sole acceleration field, and may be checked by direct substitution of the field and its derivatives.
1 Introduction
The questions considered here concern also the domains of electromagnetism and general relativity. The extreme dificulties to analyse them for Einstein’s and for Maxwell’s equations has led us to first consider these questions in Newtonian gravity. For simplicity,we shall here limit their setting to Newton’s gravity. Our starting point is the following standard situation: an arbitrary and unknown distribution of masses creates a gravitational field,which is measured
∗Proc. Spanish Relativistic Meeting, ERE 1997, Sept. 1997, Mallorca, Spain †Now at DANOF, Observatoire de Paris, [email protected]
a(x,t)
m 3
m 1
m 2
Figure 1: An unknown distribution of masses creates a gravitational field, which is measured only in a local domain (the interior of the rocket). What information about the masses may be constructively extracted from this local knowledge?
only in a local domain or laboratory (the interior of the rocket in Figure 1). We are here concerned with the informations that can be extracted from this local knowledge. All what Newton’s theory says about the gravitational fields measured in this local domain,is that they verify the gravitational equations:
∆V = 0 ⇔
da = 0 δa = 0
a being the acceleration field , a = dV ,and V the gravitational potential. But these field equations are incomplete: although they are supposed to contain all physical fields [1],it is well known that,conversely, almost all their solutions are unphysical fields [2]. We are thus led to ask our first question:
Question 1 Is it possible to find a complete set of local field equations for gravity, that is to say, a set of local field equations such that their solutions be only the physical ones?
In general,it would be interesting to know how to find,for any local solution of equations (1),the position and “charges” of its singularities,but we shall here restrinct our question to the sole physical solutions. The second question is thus:
Question 2 Does there exist a method allowing to find the masses and positions corresponding to a local physical gravitational field?
In theoretical physics,questions may be answered in many different,non equivalent forms. The above two questions have been analysed in Newtonian
γ (x,t)
Galilean Observer
m
r
Figure 2: A Galilean observer measures a local gravitational field γ(x, t). How to deduce that it is created by a spherically symmetric mass, and how to know and locate this mass?
Expresion (2) is a well known consequence of a inverse-square field; it is its sufficient character which is a less trivial result and,we believe,a new one. Once known that the acceleration field γ(x, t) is due to a sole massive par- ticle,question 2 naturaly arises: where it is located and what is its mass? The answer is as follows:
Theorem 2 Let γ be the (local)acceleration field of a gravitational point par- ticle. Then, its mass m and its position r are given by
m = 4 | γ |^5 (L(γ) | γ |)^2
, r = − 2 | γ | L(γ) | γ |
γ , (3)
where L(γ) stands for the Lie derivative along γ.
Of course,the integrability conditions of equation (2) implies that the mass given by equation (3) is constant. Its possitive character is insured by the condition f < 0 of theorem 1.
3 Inertial Acceleration Field
for Accelerated Observers
Suppose now that a local accelerated observer measures,in the domain of its laboratory,an acceleration field α(x, t). This acceleration field is,in general,the superposition of its proper accelerated motion,and of the exterior gravitational fields. Can he know when exterior gravitational fields are absent,as Figure 3 shows? or,in other words,can he be sure that he is not submitted but to inertial forces?
Galilean Observer
α (x,t)
Accelerated Observer
Figure 3: An accelerated observer measures locally an acceleration field α(x, t). How to know that he is not submitted but to inertial forces?
According to question 3,we are looking for constructive answers. In order to obtain them,we need two results concerning square roots of tensors. More precisely,we need to know when a symmetric tensor L admits an antisymmetric square root A, A^2 = L,and,in that case, what is its expression in terms of L. The corresponding results are given by propositions 1 and 2:
Proposition 1 A second order symmetric tensor L admits an antisymmetric square root if, and only if, it verifies
L^2 −
(tr L) L = 0. (4)
Proposition 2 For such a second order symmetric tensor L, its antisymmetric square root ∧
L is given by √ ∧L = √ 1
i^2 (x)L
∗ i(x)L , (5)
where
L ≡ L −
(tr L) g , (6)
x is an arbitrary regular vectorfield for L, i.e. such that i(x) L = 0 , i(.) denotes the interior product and ∗ the Hodge dual operator.
Note that,in spite of the arbitrary character of the vector x appearing in equation (5), ∧
L is unique and independent of the chosen regular x. We shall not consider here the operator ∧
· on symmetric tensors that would be
In order to present the answers to our above three basic questions,it is convenient to previously introduce some differential concomitants of an arbitrary vectorfield v. An adequate measure of the modulus of the hessian of the vectorfield v,is the scalar Φ(v) given by
Φ(v) ≡
| ∇∇v |
whose gradient allows to define the 1-form Γ(v) by
Γ(v) ≡
| dΦ(v) |^3
dΦ(v). (9)
On the other hand,theorem 1 suggest us to introduce for any v the gravita- tional differential concomitant G(v) associated to equation (2):
G(v) ≡ ∇v + λ^2
g − 3 v ⊗ v | v |^2
Similarly,theorem 3 suggests to introduce the doble inertial differential con- comitant I(v) associated to equations (7):
I(v) ≡
∇L(v)g
dv −
√∧ 2 L(v) g
With the aid of these concomitants,the complete characterization of the acceleration fields in question is the following one:
Theorem 4 A local acceleration field a(x, t) is the total acceleration field of an accelerated observer inmersed in the gravitational field of one point particle if, and only if, it verifies the equations
G (Γ(a)) = 0 , I (a − Γ(a)) = 0. (12)
Taking into account the meaning of equations (2) and (7),explained re- spectively by theorems 1 and 3,from equations (12) and theorem 2 one easily obtains:
Corollary 1 For a total acceleration field a(x, t), the gravitational acceleration γ of the particle and the inertial acceleration α of the observer are given by
γ = Γ(a) , α = a − Γ(a) , (13)
and the mass m and position r of the particle by
m =
4 | Γ |^5
(L(Γ)| Γ |)^2
, r = −
L(Γ)| Γ |
5 Conclusions
We would like to conclude by noting briefly a few points (a detailed analysis will be given elsewhere [3]): i) The above results show that,at least for one particle, a complete gravita- tional theory including location of masses is possible,and has been constructed. ii) Even for one particle,this theory is far from being trivial. iii) The method may,in principle,be extended to a finite number of particles (and perhaps,with a limiting process,to an infinite number). iv) It seems to show that a complete theory of fields is necessarily related to a hierarchy of equations,and not to a unic,universal set of equations,as we used to think up to now. v) The present results are first elements of a non usual way of thinking field theory. For some applications they may be too hard to use,but for some others they (and their plausible extensions) constitute the shortest and clear answers. vi) Although Newtonian,these results are also heuristically interesting for some relativistic problems,for example the physical interpretation of static space-times. We shall consider these problems elsewhere.
References
[1] By physical gravitational fields we understand here fields that may be created by means of point particles of positive mass.
[2] It is a direct consequence of the linear structure of the space of solutions and the positivity of the mass for physical gravitational fields.
[3] B. Coll and J.J. Ferrando,Newtonian Gravitation of One Point Particle,to be submitted to the Journal of Mathematical Physics.
