ArticlePDF Available

# Charged particle motion in Kerr-Newmann space-times

Authors:

## Abstract and Figures

The motion of charged test-particles in the gravitational field of a rotating and electromagnetically charged black hole as described by the Kerr-Newman metric is considered. We completely classify the colatitudinal and radial motion on the extended manifold $-\infty \leq r \leq \infty$, including orbits crossing the horizons or $r=0$. Analytical solutions of the equations of motion in terms of elliptic functions are presented which are valid for all types of orbits.
Content may be subject to copyright.
Charged particle motion in Kerr-Newmann space-times
Hongxiao Xuand Eva Hackmann
ZARM, University of Bremen, Am Fallturm, 28359 Bremen, Germany
The motion of charged test-particles in the gravitational ﬁeld of a rotating and electromagnetically
charged black hole as described by the Kerr-Newman metric is considered. We completely classify
the colatitudinal and radial motion on the extended manifold −∞ ≤ r≤ ∞, including orbits crossing
the horizons or r= 0. Analytical solutions of the equations of motion in terms of elliptic functions
are presented which are valid for all types of orbits.
hxu@uni-bremen.de
eva.hackmann@zarm.uni-bremen.de
2
I. INTRODUCTION
The Kerr-Newman solution to the Einstein-Maxwell equations describes the gravitational ﬁeld of a rotating and
electromagnetically charged stationary black hole [1]. It generalizes both the static and charged Reissner-Nordstr¨om
metric [2, 3] as well as the rotating Kerr metric [4]. The latter is of very high importance not only for general relativity
but also from an astrophysical point of view, as many black hole candidates were found in recent years, which are
expected to rotate. Although it is not very likely that they also carry a net charge, some accretion scenarios were
studied which may create such black holes [5, 6].
On way to explore the gravitational ﬁeld of a Kerr-Newman black hole is to consider the geodesic motion of (charged)
test-particles in this space-time. Already shortly after the discovery of this solution many aspects of the geodesic
motion were studied, among others, timelike equatorial and spherical orbits of uncharged particles [7] and the last
stable orbit of charged particles [8] (see also Sharp [9] and references within). Later the motion of charged particles
was studied by Biˇak et al [10, 11] including, besides some general discussion for the radial motion, a discussion of a
number of special cases like motion along the symmetry axis and circular motion of ultrarelativistic particles. Only
recently, Kov´r et al [12] found oﬀ-equatorial circular orbits of charged particles which are unstable outside the outer
horizon. A comprehensive analysis of photon orbits in Kerr-Newman space-time was presented by Calvani and Turolla
[13] including the extended manifold with negative values of the radial coordinate and naked singularities.
Analogously to the uncharged case, the geodesic equation in Kerr-Newman space-time can be separated by introduc-
ing an additional constant of motion (besides the constants associated to the obvious symmetries of the space-time),
the Carter constant [14], which ensures the integrability of the equations of motions. The resulting structure of the
equations is essentially the same as in Schwarzschild space-time, where the equations of motion can be solved ana-
lytically in terms of elliptic functions as ﬁrst demonstrated by Hagihara in 1931 [15]. However, due to the remaining
coupling of radial and colatitudinal equation, the generalization of his method to Kerr(-Newman) space-time was not
straightforward. This issue was solved by Mino [16] by introducing a new time parameter, often called the Mino time,
which completely decouples the equations of motions and enable a straightforward application of elliptic functions.
This was already used to analytically solve the geodesic equation for bound timelike orbits in Kerr space-time by
Fujita and Hikida [17] and for general timelike and lightlike orbits in Kerr space-time in [18].
In this paper, we will discuss the geodesic motion of charged test-particles in Kerr-Newman black hole space-
times. For the sake of completeness, we will include a magnetic charge of the black hole which was not done in
the cited references but has interesting eﬀects on the colatitudinal motion. After introducing the relevant notations
and equations of motion in the next section, we proceed with a complete classiﬁcation of timelike orbits of (charged)
particles in Kerr-Newman space-time. This includes oﬀ-equatorial orbits, trajectories crossing the horizons, and orbits
with negative values of the radial coordinate. In the fourth section, we will present analytical solutions in terms of
elliptic functions dependent on the Mino time for all coordinates. The paper is closed by a summary and conclusion.
II. GEODESICS IN KERR-NEWMAN SPACE-TIME
A. Kerr-Newman space-time
The Kerr-Newman spacetime is a stationary, axisymmetrical, and asymptotically ﬂat solution of the Einstein-
Maxwell equation
Gµν =2gαβFµα Fνβ 1
4gµν Fαβ Fαβ ,
where Gµν is the Einstein tensor and Fµν the electromagnetic tensor. Throughout, the units are selected such that
c= 1 for the speed of light and G= 1 for the gravitational constant. In Boyer-Lindquist coordinates the metric takes
the form
ds2=ρ2
dr2+ρ22+sin2θ
ρ2asin2(θ)dt2(1)
with
ρ2(r, θ) = r2+a2cos2θ , (2)
∆(r) = r22Mr +a2+Q2+P2,(3)
and M > 0 the mass, athe speciﬁc angular momentum, Qthe electric, and Pthe magnetic charge of the gravitating
source. (The existence of magnetic charges has not been proven yet but it will still be considered for the sake of
3
completeness.) We restrict ourselves here to the case that two horizons exist, given by the coordinate singularities
∆(r) = 0, r±=M±pM2a2Q2P2. The only genuine singularity is for ρ2= 0, where r= 0 and θ=π/2 is
fulﬁlled simultaneously. This means that a test particle approaching r= 0 from above or below the equatorial plane
does not terminate at r= 0 as it would in Schwarzschild space-time but continues to negative values of r. For large
negative values of rthis can be interpreted as a “negative universe”, see [19].
The Kerr-Newman metric reduces to the Kerr metric for Q=P= 0 describing the exterior of rotating non charged
black holes. It reduces to the Reissner-Nordstr¨om metric for a= 0 which describes the exterior of a non rotating but
charged black hole. In the case P=Q=a= 0 the Kerr-Newman metric is reduced to the Schwarzschild metric.
The electromagnetic potential is given by
A=Aνdxν=Qr
ρ2(dt asin2θdφ) + 1
from which the electromagnetic tensor can be calculated by F=1
2(νAµµAν)dxµdxν. By the interchanges
QP,P→ −Qthe electromagnetic potential ˇ
Aof the dual electromagnetic tensor ˇ
Fcan be obtained.
B. Equations of motion
The equations of motion for a test particle of normalized mass ǫ, electric charge e, and magnetic charge hcan be
obtained by the Hamiltonian
H=1
2gµν (πµ+eAµ+hˇ
Aµ)(πν+eAν+hˇ
Aν) (5)
where πµdescribe the general momenta. By introducing
ˆ
Q=eQ +hP
e2+h2,ˆ
P=eP hQ
e2+h2,ˆe=pe2+h2(6)
the Hamiltonian can be reduced to
H=1
2gµν (πµ+eAµ)(πν+eAν),(7)
where all quantities of the this equation refer to ˆ
Q,ˆ
P, and ˆe. Therefore, the discussion of a test particle without
magnetic charge is suﬃcient.
We can obtain three constants of motion directly, since Hdoes not depend on τ,φ, or t. The ﬁrst, ǫ2=gνµ ˙xν˙xµis
the normalization condition with ǫ= 1 for timelike and ǫ= 0 for lightlike trajectories. (The dot denotes diﬀerentiation
with respect to an aﬃne parameter τ.) The second and third equation
E=πt=gtt ˙
tg˙
φ+eAt,(8)
L=πφ=gφt ˙
t+gφφ ˙
φeAφ,(9)
describe the conservation of energy Eand angular momentum in zdirection, respectively. A fourth constant of motion
can be obtained by considering the Hamilton-Jacobi equation
τS=1
2gµν (µS+eAν)(νS+eAµ).(10)
With the ansatz S=1
2ǫτ Et ++S1(r) + S2(θ) it can be shown that the Hamilton-Jacobi equation indeed
separates with the Carter constant Kas separation constant, see [14].
With these four constants the equations of motion become
2
=¯
Kǫ¯a2cos2θT2(θ)
sin2θ=: Θ(θ),(11)
d¯r
2
=R2r)(ǫ¯r2+¯
K)¯
∆(¯r) =: R(¯r),(12)
=¯aR(¯r)
¯
∆(¯r)T(θ)
sin2θ,(13)
dt
=r2+ ¯a2)Rr)
¯
∆(¯r)¯aT(θ),(14)
4
where
Rr) = (¯r2+ ¯a2)E¯a¯
Le¯
Q¯r , (15)
T(θ) = ¯aE sin2θ¯
L+e¯
Pcos θ . (16)
All quantities with a bar are normalized to M, i.e. x= ¯xM for x=r, a, L, Q, P as well as K=¯
KM2and ¯
∆(¯r) =
¯r2r+ ¯a2+¯
Q2+¯
P2. Here γis the normalized Mino time [16] given by =Mρ2with the eigentime τ.
III. CLASSIFICATION OF MOTION
In this section we will classify the types of orbits in terms of colatitude and radial motion. We will analyze which
orbit conﬁguration, i.e. which set of orbit types, may appear for given parameters ¯a, ¯
Q, ¯
P , E , ¯
L, ¯
K, e and which region
in parameter space a given orbit conﬁgurations occupies. Here we assume that ǫ= 1, that is, we restrict ourselves to
test particles with mass, but the discussion for light may be done analogously. The whole analysis will be based on
the conditions Θ(θ)0 and Rr)0 which are necessary for geodesic motion.
For both colatitudinal and radial motion, we will ﬁrst give some general properties as symmetries and notation of
orbit types. We then proceed with the determination of possible orbit conﬁgurations, i.e. sets of orbit types which
are possible for given parameters (for more than one possible orbit type the actual orbit is determined by initial
conditions). Each orbit conﬁguration covers a particular region in the parameter space. Finally, we will analyze how
these regions look like and determine their boundaries in parameter space.
A. Colatitudinal motion
The coordinate θmay only take a speciﬁc value θ0[0, π] if Θ(θ0)0 is valid. We will analyze in the following
whether this is fulﬁlled for a given parameter set.
1. General properties
First, we notice that Θ does not depend on ¯
Qand that eand ¯
Ponly appear as the product e¯
P. The function Θ
has the following symmetries:
A change of sign of e¯
Phas the same eﬀect as reﬂecting θat the equatorial plane: Θ|e¯
P(θ) = Θ|e¯
P(πθ). In
particular is Θ symmetric with respect to the equatorial plane if e¯
P= 0: Θ|e¯
P=0(θ) = Θ|e¯
P=0(πθ).
A simultaneous change of sign of ¯
Land Eresult in a reﬂection at the equatorial plane: Θ|¯
L,E(θ) = Θ|¯
L,E (πθ)
and Θ¯
L=0=E(θ) = Θ¯
L=0=E(πθ).
Therefore, we assume w.l.o.g. e¯
P0 and E0 in the following. The condition Θ(θ)0 also shows that ¯
K0 is
a necessary condition for geodesic motion as all other (positive) terms are subtracted.
The Carter constant also encodes some geometrical information if considered in its alternative form ¯
C=¯
KaE
¯
L)2. Because of Θ(π/2) = ¯
KaE ¯
L)2=¯
Ca particle may only cross or stay in the equatorial plane if ¯
C0.
For equatorial orbits even ¯
C= 0 is necessary as
π
2= 0 needs to be fulﬁlled. For ¯
K= 0 geodesic motion is only
possible if ¯a=¯
L= 0 or in the equatorial plane with ¯aE =¯
L.
The function Θ(θ) constains a term which diverges for θ= 0, π given by (¯
Le¯
Pcos θ)2
sin2θ. This fact suggests to
distinguish between two cases:
¯
L6=±e¯
P: In this case Θ(θ)→ −∞ for θ0, π, that is, the north or south pole will never be reached.
¯
L=±e¯
P: In this case Θ(θ)¯
K¯a2for θ0 and ¯
L=e¯
Pas well as for θπand ¯
L=e¯
P. Therefore,
a particle with ¯
L=e¯
Pmay reach the north pole and a particle with ¯
L=e¯
Pthe south pole if in addition
¯
K¯a2. For the subcase ¯
L= 0 = e¯
Pboth north and south pole may be reached if ¯
K¯a2.
If the parameters are such that the poles θ= 0, π can not be reached it is convenient to consider ν= cos θinstead of
θ. In terms of νthe diﬀerential equation for colatitudinal motion reads
2
=
4
X
i=0
biνi=: Θν(ν),(17)
5
where b0=¯
K(¯
L¯aE)2=¯
C,b1= 2e¯
P(¯
L¯aE), b2=¯
K¯a2e2¯
P2+ 2¯a2E22¯aE ¯
L,b3= 2¯aEe ¯
P, and
b4= ¯a2(1 E2).
For a given set of parameters of the space-time and the particle diﬀerent types of orbits may be possible. We call
an orbit
northern or N, if it stays in the northern hemisphere θ < π/2,
normal or E, if it crosses or stays in the equatorial plane θ=π/2,
southern or S, if it stays in the southern hemisphere θ > π/2.
Equatorial orbits with θπ/2 are a special case of normal orbits. In addition to the above notions we add an index
Nif the north pole θ= 0 and Sif the south pole θ=πis reached, for example NNfor a northern orbit reaching the
north pole.
2. Orbit conﬁgurations
Let us now analyze which orbit conﬁgurations, i.e. which sets of the above introduced orbit types, are possible
for given parameters. We use the necessary condition for colatitudinal motion Θ(θ)0 for this, which implies to
analyze the occurrence of real zeros of Θ in [0, π] and the behavior of Θ at the boundaries θ= 0, π giving the sign of
Θ between its zeros. (The orbit a test particle with the given parameters actually follows in a space-time with the
given parameters depends on the initial values.)
(A) Case ¯
L6=±e¯
PHere Θ(θ) = −∞ at θ= 0, π which implies that Θνhas an even number of zeros in (1,1)
(counted with multiplicity). If Θνhas no real zeros there, no colatitudinal motion is possible, which gives a restriction
to the permitted sets of parameters for geodesic motion. In the case of 2 real zeros there is a single orbit of type N,
E, or S, which is stable at a constant θif the 2 zeros coincide. If all 4 zeros of Θνlie in (1,1) all combinations of
two orbit types except EE are possible. For two or more coinciding zeros this point is stable if it is a maximum of
Θνand unstable otherwise.
(B) Case ¯
L=±e¯
PHere we have to consider four subcases: (B1) ¯
K < ¯a2: The same orbit types as in (A) are
possible. (B2) ¯
K > ¯a2,¯
L6= 0: In this case Θ has diﬀerent signs at θ= 0, π which implies that Θ has an odd number
of real zeros in (0, π). For 1 real zero there is one orbit of type NNor ENfor ¯
L=e¯
Pand one of type SSor ESfor
¯
L=e¯
P. If Θ has 3 real zeros in (0, π ) there is one additional orbit not reaching a pole. (B3) ¯
K > ¯a2,¯
L= 0 = e¯
P:
Here Θ >0 at θ= 0, π, i.e. Θ has an even number of zeros in (0, π). Also, for e¯
P= 0 the function Θ is symmetric with
respect to the equatorial plane. For no real zeros there is one orbit of type ENS which reaches both poles, and for 2
real zeros an NNand an SSorbit. More zeros in [0, π ] are not possible. (B4) ¯
K= ¯a2: Here an orbit with constant
θ= 0 (θ=π) is possible for ¯
L=e¯
P(¯
L=e¯
P). For ¯
L= 0 no other than the two constant orbits are possible but
for ¯
L6= 0 it is Θ → −∞ at the other boundary. In the latter case, if the orbit is stable, there may be one additional
orbit of type E,N, or S.
For an overview of this diﬀerent orbit conﬁguration see table I.
3. Regions of orbit conﬁgurations in parameter space
It is now of interest for which sets of parameters a given orbit conﬁguration changes. As Θ 0 is necessary for
geodesic motion, this happens if the behavior of Θ at the boundaries changes, which means a switch from one of the
above cases (A),(B1), ..., (B4) to another, or if the number of real zeros of Θ changes. The latter occurs at that
parameters for which Θ has multiple zeros. With these two conditions the diﬀerent regions of orbit conﬁgurations in
parameter space can be completely determined. The ﬁrst condition was already analyzed above.
For θ(0, π) the function Θ has the same zeros as the polynomial Θνand we may use Θνinstead of Θ for all
orbits not reaching θ= 0, π. If ¯
L=±e¯
Pthen ν0=±1 is a zero of Θνbut does not correspond to a turning point of
the colatitudinal motion. If in addition ¯
K= ¯a2then ν0=±1 is a double zero of Θνand θ= 0, π a simple zero of Θ,
which does correspond to a turning point of θ. Keeping this in mind we will also use Θνfor these cases but discuss
the occurence of multiple zeros at θ= 0, π separatly using Θ.
The condition for a double zero ν0is dΘν
(ν0) = 0 = Θν(ν0). This can be read as 2 conditions on 2 of the 5
parameters E,¯
L,¯
K,e¯
P, and a. Solving these two conditions for Eand ¯
Ldependent on the position of the double
6
zeros range of θ[0, π] types of orbits
2N
2E
2S
4N,N
4N,E
4N,S
4E,S
4S,S
(A) ¯
L6=±e¯
P; (B1) ¯
L=±e¯
P,¯
K < ¯a2
zeros range of θ[0, π] types of orbits
1NN
1EN
3NN,N
3NN,E
3NN,S
3EN,S
(B2) ¯
L=e¯
P,¯
L6= 0, ¯
K > ¯a2
zeros range of θ[0, π] types of orbits
0EN,S
2NN,SS
(B3) ¯
L= 0 = e¯
P,¯
K > ¯a2
zeros range of θ[0, π] types of orbits
2NN
3NN
3EN
4NN,N
4NN,E
4NN,S
(B4) ¯
L=e¯
P,¯
L6= 0, ¯
K= ¯a2
zeros range of θ[0, π] types of orbits
4NN,SS
(B4) ¯
L= 0 = e¯
P,¯
K= ¯a2
TABLE I: Overview of diﬀerent orbit conﬁgurations for colatitudinal motion. The vertical bar of the second column
denotes θ=π/2 and the thick lines Θ 0, i.e. regions where a motion is possible. Dots represent single zeros and
circles double zeros. If zeros merge, the resulting orbits are stable if a line is reduced to a point and unstable if lines
merge. The conﬁgurations (B2) and (B4) with ¯
L=e¯
Pare obtained by a reﬂection at the equatorial plane.
zero ν0and the other parameters yields
E1,2=e¯
P
0±1
2pν2
0(1 ν2
0)( ¯
K¯a2ν2
0)( ¯
K+ ¯a2a2ν2
0)2
(ν2
01)( ¯
K¯a2ν2
00
,(18)
¯
L1,2=e¯
P(ν2
0+ 1)
2ν0±1
2
ν0(1 ν2
0)( ¯
K¯a2)( ¯
K+ ¯a2a2ν2
0)
pν2
0(1 ν2
0)( ¯
K¯a2ν2
0)( ¯
K+ ¯a2a2ν2
0)2.(19)
The expressions for Eand Ldiverge at ν0= 0 for e¯
P6= 0, at ν0=p¯
K/¯a2, and at ν0=±1 for E, what suggests to
consider the 2 conditions for double zeros directly for these points:
Equatorial orbits For ν0= 0 the 2 conditions imply that either ¯
K= 0 and ¯
L= ¯aE or ¯
K= (¯aE ¯
L)2and e¯
P= 0
are necessary and suﬃcient for the existence of equatorial orbits. The asymptotic behavior of Eand ¯
Lat ν0= 0 also
displays these conditions,
E1,2=e¯
P
0¯
K+ ¯a2
a¯
K+O(ν2
0),(20)
¯
L1,2=e¯
P
2ν0±¯
K¯a2
2¯
K+O(ν0) = ¯aE1,2±p¯
K+O(ν0).(21)
In the case ¯
K= 0, ¯
L= ¯aE the equatorial orbit is the only possible geodesic orbit (for Θ 6≡ 0) and, thus, stable,
whereas for ¯
K= (¯aE ¯
L)2,e¯
P= 0 the sign of A:= ¯a2(E21) ¯
L2has to be considered: The orbit is stable if A0
(for Θ 6≡ 0) and unstable if A > 0. In the case A= 0 the zero θ=π/2 is even fourfold. For ¯
K= ¯a2,e¯
P= 0 = ¯
L, and
E2= 1 the function Θ is identical to zero.
Orbits with θ0, π Geodesic motion along the axis θ= 0, π is possible only if θ= 0, π is a double zero of Θ. The
2 conditions on double zeros show that ¯
L=e¯
Pand ¯
K= ¯a2are necessary and suﬃcient for θ0, and ¯
L=e¯
P,
¯
K= ¯a2for θπ. This can also be seen by considering the asymptotic behaviour of ¯
L1,2and E1,2as ν0approaches
±1: it is given by limν01¯
L1,2e¯
Pand limν0→−1¯
L1,2=e¯
Pwhereas E1,2diverges for ¯
K6= ¯a2.
7
Let us discuss the stability of the orbits θ0, π : The orbits are unstable if ¯a2aE ¯
L/2)2>0 and stable if
¯a2aE ¯
L/2)2<0. For ¯a2aE ¯
L/2)2= 0 the poles θ= 0, π are fourfold zeros and the orbit θ0 is stable
if E=¯
L
2a1, and the orbit θπif E=¯
L
2a+ 1. In the special case of ¯
L= 0 = e¯
Pthe two orbits are unstable if
E2<1 and stable if E2>1. For ¯
L= 0 = e¯
P,¯
K= ¯a2, and E2= 1 again Θ 0.
Orbits with θ≡ ±p¯
K/¯a2The singularity ν0=±p¯
K/¯a2is located in (1,1) and is not equal to zero only if
0<¯
K < ¯a2. Assuming this, orbits with constant cos θ≡ ±p¯
K/¯a2can exist only if ¯a¯
La2¯
K)E=±e¯
P¯
Kis
fulﬁlled. This can be infered from the asymptotes of ¯
L1,2in terms of E1,2around ν0=±p¯
K/¯a2,
¯
L1,2=¯a2¯
K
¯aE1,2±e¯
P¯
K
¯a+O
sνr¯
K
¯a2
.(22)
The expressions (18) and (19) depend linearly on e¯
Pand a rescaling of parameters ( ¯
L/¯a,¯
K/¯a2,e¯
P /¯a) removes
the rotation parameter ¯acompletely from the equations. Only the dependence on ¯
Kis not obvious. If we solve the 2
conditions dΘν
(ν0) = 0 = Θν(ν0) for Eand ¯
Lthis yields
E1,2=e¯
P
0±1
2p(e¯
P(ν2
0+ 1) 2¯
0)2+ 4ν2
0¯a2(ν2
01)2
¯0(ν2
01) (23)
¯
K1,2= ¯a2e¯
P(ν2
0+ 1) 2ν0¯
L
2ν2
0(ν2
01) (e¯
P(ν2
0+ 1) 2ν0¯
L)±q(e¯
P(ν2
0+ 1) 2¯
0)2+ 4ν2
0¯a2(ν2
01)2(24)
Note that E1and ¯
K1behave regular at ν0= 0 and ¯aE1=¯
L+O(ν0), ¯
K1=O(ν2) whereas E2and ¯
K2are ﬁnite
at ν0= 0 only if e¯
P= 0. Then ¯
KaE ¯
L)2as expected from the analysis of equatorial orbits above. Also, for
ν0→ ±1 the conditions ¯
L=±e¯
P,¯
K¯a2are recovered.
Triple zeros are also of interest as they correspond to parameters where a stable orbit with constant θbecomes
unstable and vice versa. Therefore, will will study them here. The 3 conditions for such points are 0 = d2Θν
2(ν0) =
dΘν
(ν0) = Θν(ν0) which we read as 3 conditions on E,¯
L, and e¯
Pyielding
E1,2=±1
2
2ν6
0¯a4+ 6ν2
0¯
K¯a2(3ν4
0¯a2+¯
K)( ¯
K+ ¯a2)
p(1 ν2
0)( ¯
K¯a2ν2
0)(1 ν2
0)( ¯
K¯a2ν2
0a,(25)
¯
L1,2=1
2p(1 ν2
0)( ¯
K¯a2ν2
0)( ¯
K¯a2)(¯a2ν6
03ν4
0¯a2+ 3ν2
0¯
K¯
K)
(ν2
01)2(¯
K¯a2ν2
0)2,(26)
e¯
P1,2=±ν3
0(¯
K¯a2)2
p(1 ν2
0)( ¯
K¯a2ν2
0)(ν2
01)( ¯
K¯a2ν2
0).(27)
In particular, triple zeros are also double zeros and the asymptotic behavior of ¯
Las a function of Eand ¯
Kat the
singularities ν0=±p¯
K/¯a2is determined by (22). The points ν0=±1 were studied above but we note here that
θ= 0, π is a double zero of Θ if ν=±1 is a triple zeros of Θν.
Besides the equatorial orbits and orbits with constant θ= 0, π discussed above, fourfold or even higher order zeros
are only possible for parameters corresponding to Θ 0, which are ¯a=e¯
P=¯
L=¯
K= 0 with arbitrary Eor E=±1,
¯
K= ¯a2, and ¯
L=e¯
P= 0.
The results of this section are visualized in ﬁgures 1 to 4. As the ﬁve dimensional parameter space can not be
completely pictured we have to ﬁx at least two parameters. For this we choose ¯a, which can be removed from equations
(18), (19), (23), and (24) by a rescaling of parameters, and e¯
P, which enters only linearly in (18) an (19). As a three
dimensional plot of E,¯
L, and ¯
Kis still confusing we present two dimensional plots of ¯
Lover Eand ¯
Kover E. Special
cases to be discussed are then e¯
P= 0, ¯
L=±e¯
P,¯
K= ¯a2, and ¯
K= 0. For the latter geodesic motion is only possible
on a stable orbit of constant θ(if not Θ 0), more precise on the equator for ¯
L= ¯aE or on e¯
Pcos θ=¯
Lfor ¯a= 0.
In each plot, we use the following conventions:
Solid lines indicate double zeros of Θ which correspond to stable or unstable orbits of constant θ. We use red
lines for orbits with constant θ < π
2, blue for constant θ > π
2, and green for equatorial orbits.
Dashed lines denote orbits with turning points at the equatorial plane and are given by ¯
L= ¯aE ±¯
K. They
also mark the transition from Nor Sto Eorbits and are asymptotically approached by solid lines corresponding
to orbits with constant θnear the equatorial plane.
8
¯
K= 3, e¯
P= 11.25 ¯
K= 0.5, e¯
P= 0.3
FIG. 1: Orbit conﬁgurations for the colatitudinal motion with ¯a= 0.5, ¯
K > ¯a2, and e¯
P > 0. For a general
description see the text. The orbit conﬁgurations on the solid lines contain an orbit with constant θ. They are
unstable if marked by the red solid line starting at the dot and approaching the dashed line, and stable
otherwise.The dash dotted line at ¯
L=e¯
P(¯
L=e¯
P) denotes an orbit crossing the north (south) pole and
corresponds to the orbit conﬁgurations (B2). The other regions correspond to the conﬁgurations (A). Small plots on
the top are enlarged details of the lower plot.
Dash dotted lines correspond to orbits which cross a pole. They only appear in ¯
Lover Eplots for ¯
K¯a2
and are located at ¯
L=±e¯
P. In this case they are asymptotes to the solid lines corresponding to orbits with
constant θnear a pole.
Dotted lines are asymptotes to solid lines. They do NOT separate diﬀerent orbit conﬁgurations. For ¯
Lover E
plots they only appear for ¯
K < ¯a2and are approached for ν0→ ±¯
K/¯a. If ¯
Kis plotted versus Ethey are
approached for ν0→ ±1.
Single dots mark triple zeros which separate stable from unstable orbits.
The labels N,E, and Sindicate the orbit conﬁgurations summarized in table I. Solid, dashed, and dash dotted
lines indicate transitions from one orbit conﬁguration to another. Here dash dotted lines are special in the sense
that on both sides there are always the same orbit conﬁgurations but only the line itself corresponds to another
conﬁguration. Regions where a colatitudinal motion is forbidden are marked with ‘none‘.
Note that we plot usually only positive values of Eas negative values can be obtained by E→ −E,¯
L→ −¯
Land a
reﬂection at the equatorial plane (e.g. an Norbit becomes an Sorbit).
9
(a) ¯
K= ¯a2(b) ¯
K= 0.2
FIG. 2: Orbit conﬁgurations for the colatitudinal motion with ¯a= 0.5, ¯
K¯a2, and e¯
P= 0.3. For a general
description see the text. The orange solid lines in (a) correspond to orbits of constant θ= 0, π, which are unstable
for energies less than the one marked by the dots. These orbits correspond to the conﬁgurations (B4), other regions
(in both plots) to the conﬁgurations (A) or (B1). In both plots, the orbits labeled by the red solid line starting at
the dot and approaching the dashed line are unstable. All other orbits on solid lines are stable. Small plots on the
top are enlarged details of the lower plot.
The discussion of the radial motion will be analogous to that of the θ-motion. An orbit can only take a speciﬁc
value in [−∞,] if the radicand of the right hand side of eq. (12) is larger than or equal to zero at that point, i.e. if
Rr) = ((¯r2+ ¯a2)E¯a¯
Le¯
Q¯r)2(ǫ¯r2+¯
K)(¯r22¯r+ ¯a2+¯
Q2+¯
P2)0.
We will analyze for which values of the parameters E,¯
L,¯
K,e,¯
Pand ¯
Qthis inequality is satisﬁed.
1. General properties
The three parameters e,¯
Q, and ¯
Pappear only in the two combinations e¯
Q=: Qand ¯
Q2+¯
P2=: P2which may be
considered instead. The function Rhas the following symmetries:
P(and P): R|¯
P=R|¯
P(R|−P =R|P).
A simultaneous change of sign of ¯
Qand eresults in the same motion: R|¯
Q,e=R|¯
Q,e.
Also, a simultaneous change of sign of ¯
L,Eand e(or Q) results in the same motion: R|E,¯
L,e=R|E, ¯
L,e
(R|E,¯
L,−Q =R|E, ¯
L,Q).
10
(a) ¯
K= 0.5 (b) ¯
K= 0.2
FIG. 3: Orbit conﬁgurations for the colatitudinal motion with ¯a= 0.5 and e¯
P= 0. For a general description see the
text. The green solid lines correspond to equatorial orbits which are unstable from dot to larger ¯
Land else stable.
The black solid line marks two stable orbits of constant θ6=π
2which are symmetric with respect to the equatorial
plane. The dash dotted line in (a) marks the case (B3) with an EN S orbit between the green solid lines and an
NNSSconﬁguration else.
Thus it suﬃces to consider ¯
P0, ¯
Q0, and e0 (or, equivalently, P ≥ 0 and Q ≥ 0). Note that ¯
K0 was a
necessary condition for the colatitudinal motion to be possible at all and, therefore, this condition remains valid.
The sign of ¯
C=¯
KaE ¯
L)2again encodes some geometrical information. At ¯r= 0 the polynomial R(0) =
(¯
Q2+¯
P2)(¯
L¯aE)2a2+¯
Q2+¯
P2)¯
Ccan only be positive if ¯
C0. Since ¯
C0 needs to be satisﬁed for an
orbit to reach the equatorial plane this implies that ¯
C= 0 is necessary for an orbit to hit the ring singularity. For
¯
Q2+¯
P26= 0 also ¯
L¯aE = 0 is neccessary to hit the singularity and, thus, also ¯
K= 0.
The zeros of the parabola ¯
∆(¯r) are the horizons ¯r±= 1 ±p1¯a2+¯
Q2+¯
P2with ¯
∆(¯r)<0 in between. Therefore,
the polynomial R(¯r) is always postive for ¯rr,¯r+]. This implies that there can not be any turning points or
spherical orbits between the horizons. There is a turning point at a horizon if Rr±) = 0 (see eq. (15)).
For a given set of parameters of the space-time and the particle diﬀerent types of orbits may be possible, for which
we use the terminologies
transit or T, if ¯rstarts at ±∞ and ends at ,
ﬂyby or F, if ¯rstarts and ends at +or −∞,
bound or B, if ¯rremains in a ﬁnite interval [¯rmin,¯rmax].
We add an index +, 0, or to a ﬂyby or bound orbit if it stays at ¯r > 0, crosses ¯r= 0, or stays at ¯r < 0. Also, a
superscript will be added if the orbit crosses the horizons, i.e. contains the interval [¯r,¯r+]. For example, the orbit
F
+comes from inﬁnity, crosses the two horizons, turns at some 0 <¯r < ¯r, and goes back to inﬁnity. If more than
one orbit is possible for a given set of parameters, the actual orbit of the test particle is determined by the initial
conditions.
2. Orbit conﬁgurations
We will now analyze which sets of the above introduced orbit types are possible for given parameters. Geodesic
motion is only possible in regions of Rr)0 and, therefore, the possible orbit conﬁgurations are fully determined
by the number of real zeros of the polynomial Rand its sign at ±∞. The latter is determined by the sign of E21
(or the lower order coeﬃcients if E2= 1, what we will consider separately). This suggest to introduce the following
classes of orbit conﬁgurations.
11
(a) ¯
L= 0.5 (b) ¯
L=e¯
P
FIG. 4: Orbit conﬁgurations for the colatitudinal motion with ¯a= 0.5 and e¯
P= 0.3. For a general description see
the text. For both plots, solid lines with dashed asymptotes correspond to unstable and all other to stable orbits. In
(b) the regions above the solid orange line at ¯
K= ¯a2correspond to the conﬁgurations (B2) and below to (B1). The
line itself corresponds to (B4) and covers all possible conﬁgurations (leftmost corresponds to topmost in table I)
with the exception of the NNNconﬁguration which is only possible for large values of e¯
P.
(I) Case E2>1 Here R(¯r)→ ∞ for ¯r→ ±∞ and Rmay have none, two, or four real zeros. For no zeros there
is a transit orbit, for two zeros there are two ﬂyby orbits, and for four zeros there are two ﬂyby and a bound orbit.
(II) Case E2<1 For such energies a test particle can not reach ±∞,R(¯r)→ −∞ for ¯r→ ±. As Rr)>0
between the horizons it has at least 2 real zeros and there is always one bound orbit crossing the horizons. If Rhas
four real zeros there is an additional bound orbit.
(III) Case E2= 1 Here the behaviour of Rat inﬁnity depends on the sign of 1 e¯
Q. For 1 e¯
Q > 0 it is
Rr)→ ±∞ for ¯r→ ±∞ and the other way around for 1 e¯
Q < 0. In both cases Rhas one or three real zeros. For
one real zero there is a ﬂyby orbit which crosses the horizons and for three real zeros there is a ﬂyby and a bound
orbit. The ﬂyby orbit reaches +for 1 e¯
Q > 0 and −∞ for 1 e¯
Q < 0. If also 1 e¯
Q= 0 one has to consider the
sign of the second order coeﬃcient and so on.
For an overview of this diﬀerent orbit conﬁguration for the radial motion see table II.
3. Regions of orbit conﬁgurations in parameter space
After considering the possible orbit conﬁgurations for the radial motion we will now ﬁnd the sets of parameters for
which a given orbit conﬁguration changes to another. The orbit conﬁgurations are fully determined by the signs of
R(±∞) as categorized above and the number of real zeros of R, which changes if two zeros of Rmerge. The latter
occurs if the two conditions on double zeros Rr0) = 0 and dR
d¯rr0) = 0 are fulﬁlled. Read as two conditions on Eand
12
real zeros range of ¯rtypes of orbits
0 T
2F
0,F+
2F0,F
+
2F,F
+
2F,F
0
4F
0,B+,F+
4F0,B
+,F+
4F,B
+,F+
4F,B
0,F+
4F0,B+,F
+
4F,B+,F
+
4F,B0,F
+
4F,B,F
+
4F,B,F
0
(I) E2>1
real zeros range of ¯rtypes of orbits
1F
0
3F,B
0
3F,B
+
3F0,B
+
3F
0,B+
(III) E2= 1, e¯
Q > 1
real zeros range of ¯rtypes of orbits
2B
+
2B
0
4B
+,B+
4B
0,B+
4B+,B
+
4B0,B
+
4B,B
+
4B,B
0
(II) E2<1
real zeros range of ¯rtypes of orbits
1F
+
1F
0
3B
+,F+
3B
0,F+
3B+,F
+
3B0,F
+
3B,F
+
3B,F
0
(III) E2= 1, e¯
Q < 1
TABLE II: Overview of diﬀerent orbit conﬁgurations for radial motion. The vertical bars of the second column mark
¯r= 0, ¯r= ¯r, and ¯r= ¯r+(from left to right). The dots represent the real zeros of R(turning points) and the thick
lines R0, i.e. regions where a motion is possible. If zeros merge, the resulting orbits are stable if a line is reduced
to a point and unstable if lines merge.
¯
Lthis implies
E1,2=e¯
Q
r0±1
2p¯
∆(¯r0)(¯r2
0+¯
K)(¯r0¯
∆(¯r0) + (¯r01)( ¯
K+ ¯r2
0))2
¯
∆(¯r0)(¯r2
0+¯
Kr0
,(28)
¯
L1,2=e¯
Qa2¯r2
0)
a¯r0±1
2p¯
∆(¯r0)(¯r2
0+¯
K)(¯r0¯
∆(¯r0) + (¯r01)( ¯
K+ ¯r2
0))2
¯
∆(¯r0)(¯r2
0+¯
K)(¯r0¯
∆(¯r0) + (¯r01)( ¯
K+ ¯r2
0))¯r0¯a
×r0¯
∆(¯r0)(¯a2¯
K) + ( ¯
K+ ¯r2
0)(¯r2
0¯r0(¯
P2+¯
Q2)¯a2)) .(29)
These expressions diverge at ¯r0= 0 and at the horizons ¯r0= ¯r±. We consider these points separately, see below. In
Ldiverges also at ¯r0=±∞,¯
L1,2±sign(¯r0)e¯
Q
a¯r0+O(1) for ¯r0→ ±∞. This implies that for e¯
Q > 1
other orbit conﬁgurations than for e¯
Q < 1 may appear. At the limits ¯r0 ±∞ the energy Eremains ﬁnite,
E1,2→ ±sign(¯r0) there.
Orbits at ¯r= 0 As the ring singularity is located at ¯r= 0, θ=π
2only those orbits which are not equatorial do
not terminate at ¯r= 0. From the discussion of the colatitudinal motion equatorial orbits occur for ¯
K= 0, ¯aE =¯
Lor
¯
K= (¯aE ¯
L)2,e¯
P= 0. If we exclude these parameters ¯r= 0 is a multiple zero if and only if e¯
Q6= 0, ¯
K=e2¯
Q2¯
∆(0),
and ¯aE ¯
L=e¯
Q¯
∆(0)
¯a. This can also be infered from the asymptotic behaviour of E1,2and ¯
L1,2at ¯r0= 0, which is
13
given by
E1,2=e¯
Q¯
∆(0) ±p¯
K¯
∆(0)
2¯
∆(0)¯r0¯
∆(0)(1 + e2¯
Q2)e2¯
Q2
2e¯
Q¯
∆(0) +Or0),(30)
¯
L1,2= ¯aE1,2±pK¯
∆(0)
¯a+O(¯r0).(31)
Here the equation for E2implies two facts: First, at ¯
K=e2¯
Q2¯
∆(0) the regions of orbit conﬁgurations essentially
change because the sign of the r1
0term changes and E2switches between ±∞. Second, for ¯
K=e2¯
Q2¯
∆(0) the
expressions for E2,L2, and e¯
Qare the same as (35)-(37) with ¯r0= 0, which correspond to triple zeros.
Orbits at the horizons It was noted above that a horizon is a turning point of the radial motion if Rr±) = 0. This
implies that the horizons can not be multiple zeros because of dR
d¯rr±) = 2(¯r±1)(¯r2
±+¯
K)6= 0 for R(r±) = 0. (This
is valid for massive test particles only. For light rays a horizon is a multiple zero if ¯
K= 0 and ¯a¯
L=Er2
±+ ¯a2)e¯
Q¯r±
but the corresponding orbit is always unstable.) The asymptotic behaviour of E1,2and ¯
L1,2near the horizons is given
by
lim
¯r0¯r+
E1,2→ ±2
4
¯r+1q¯r2
++K
¯r+¯r0¯r+
+O(1) ,(32)
lim
¯r0¯r
E1,2→ ∓2
4
¯r1q¯r2
+K
¯r¯r0¯r
+O(1) ,(33)
lim
¯r0¯r±
¯
L1,2E1,2
¯a(2¯r±¯
P2¯
Q2) + O(1) .(34)
Now let us turn back to the general case. The expressions (28) and (29) depend linearly on Q=eQ and the
parameter ¯acan be removed by a rescaling of the other parameters ( ¯
L
¯a,Q
¯a,P
¯a2,¯
K
¯a2) and the radial coordinate ( ¯r
¯a).
(Note that also distances measured in units of Mhave to be rescaled). The dependence on ¯
Kand P2=¯
Q2+¯
P2is less
obvious. This can be studied by solving (28) and (29) for these parameters but the expressions are quite cumbersome
and we do not give them here. However, in Figure 9 orbit conﬁgurations for varying ¯
Kare shown and all possible
orbit types already appear in Figures 5 to 8.
Let us now analyze where triple zeros occur as they mark transitions from stable to unstable orbits. Solving the
three conditions d2R
d¯r2r0) = dR
d¯rr0) = R(¯r0) = 0 for E,¯
L, and Q=e¯
Qyields
E1,2=±1
2
¯
K¯
2r0) + (¯r2
0+¯
K)(3¯r2
0r0+¯
K)¯
∆(¯r0)(¯r2
0+¯
K)2r01)2
r2
0+¯
K)¯
∆(¯r0)p(¯r2
0+¯
K)¯
∆(¯r0),(35)
L1,2=1
2pr2
0+¯
K)¯
∆(¯r0)
r2
0+¯
K)2¯
2r0a¯
K(3¯r2
0¯a2+ 2 ¯
K)¯
2r0) + (¯r2
0+¯
K)(¯r4
0¯
Kr2
0r0+ ¯a2)¯a2(3¯r2
0r0)) ¯
∆(¯r0)
+ (¯r2
0+¯
K)2a2¯r2
0)(¯r01)2,(36)
Q1,2=¯r3
0¯
2r0)(¯r2
0+¯
K)(2¯r3
0¯r2
0+¯
K)¯
∆(¯r0) + (¯r2
0+¯
K)2¯r0r01)2
r2
0+¯
K)¯
∆(¯r0)p(¯r2
0+¯
K)¯
∆(¯r0).(37)
Here again, the horizons ¯r±are singularities and in between triple zeros are not possible. Fourfold zeros may only
occur for ¯r < 0, given by E1,2=±r1
2¯rr1) ,¯
L1,2=±¯r2rP2¯a2
a¯r( ¯r1) ,Q1,2=±¯r
¯rr1) , and ¯
K=¯r(¯
∆(0)¯r)
¯r1, or at the
ring singularity if E=±1
¯
∆(0) ,¯
K= 0, Q=1
E, and ¯
L= ¯aE.
The regions of diﬀerent orbit conﬁgurations are visualized in Figures 5-8. As the parameter space is six dimensional,
we have to ﬁx at least three parameters for plotting. We always choose to ﬁx ¯aand Q=e¯
Qfor the reasons outlined
above. Here Q=±1 and Q=±p¯
K/ ¯
∆(0) will separate quite diﬀerent plot structures: at Q=±1 the behaviour of
¯
L1,2at inﬁnity changes and at Q=±p¯
K/ ¯
∆(0) the behaviour of Eat ¯r= 0. We will restrict here to positive values
of Qbut allow all values of ¯
Land E. The values e¯
Q < 0 are recovered by (E, ¯
L)(E, ¯
L) as noted in the section
on general properties. Therefore, we distinguish between four diﬀerent regions: (i) Q<min{1,q¯
K
¯
∆(0) },(ii) q¯
K
¯
∆(0) <
Q<1, (iii) 1<Q<q¯
K
¯
∆(0) , and (iv) max{1,q¯
K
¯
∆(0) }<Q. Note that for e¯
Q= 0 the polynomial Ris unchanged
by the transformation (E, ¯
L)(E, ¯
L) and, therefore, regions of orbit conﬁgurations diﬀer only slightly from this
14
(a) e¯
Q= 0.3, e¯
Q < 1<q¯
K
¯
∆(0)
(b) e¯
Q= 0.9, e¯
Q < 1<q¯
K
¯
∆(0)
FIG. 5: Orbit conﬁgurations for the radial motion with ¯a= 0.6, ¯
K= 1, and P2= 0.4 for case (i). For a general
description of colours and linestyles see the text. The spherical orbits marked by the dark blue (green) solid lines
starting at the dots and approaching the light blue dashed (the black dash dotted) lines are stable. In (b) also the
orbits on the red solid line between the two dots are stable. All other spherical orbits are unstable. Note that for the
small e¯
Qof (a) regions of orbit conﬁgurations are only slightly deformed by the transformation E→ −E,¯
L→ −¯
L
but remain unchanged otherwise. Small plots on the top are enlarged details of the lower plot.
symmetry for small Qin region (i). However, if Qis larger than (37) with ¯r0=1
2(¯
∆(0) ¯
Kp(¯
∆(0) ¯
K)2+ 4 ¯
K)
(local minimum) the structure changes as an additional pair of triple zeros appears, see Figure 5b.
For better comparison, we always plot ¯
Lover Eand indicate the dependence on ¯
Kand P=¯
P2+¯
Q2by slowly
varying them in a plot series. In each plot, we use the following conventions:
Solid lines indicate double zeros of Rwhich correspond to stable or unstable spherical orbits of constant ¯r. We
use red lines for orbits with constant ¯r < 0, blue for constant 0 <¯r < ¯r, and green for constant ¯r > ¯r+.
Dashed lines denote orbits with turning points at ¯r= 0 and are given by ¯
L±= ¯aE ±¯
K¯
∆(0)
¯a. They mark
15
FIG. 6: Orbit conﬁgurations for the radial motion with ¯a= 0.6, ¯
K= 0.5, P2= 0.4, and e¯
Q= 0.9, q¯
K
¯
∆(0) < e ¯
Q < 1.
For a general description of colours and linestyles see the text. The spherical orbits on solid lines starting at a dot
and approaching the black dash dotted lines or the light blue dashed lines are stable (two green lines, a dark blue
line, and a red line). All other spherical orbits are unstable. In the detailed plot on the upper right the solid blue
line approaches the light blue dashed line so close from below that they are hard to distinguish; the regions
indicated there are meant to be between them. The same holds for the lower right plot with the red solid line
approaching from above.
transitions between orbits with diﬀerent indices. Red or blue solid lines corresponding to orbits with constant
¯rnear ¯r= 0 asymptotically approach ¯
L±.
The dash dotted line marks E2= 1. Between E=±1 no orbit can reach inﬁnity. In addition, E=±1 is
asymptotically approached by red or green solid lines with ¯r→ ±∞
Dotted lines are asymptotes to solid lines for ¯r¯r±. They do NOT separate diﬀerent orbit conﬁgurations.
Single dots mark triple zeros which separate stable from unstable orbits.
The labels T,F, and Bindicated the orbit conﬁgurations summarized in table II. Solid, dashed, and dash
dotted lines indicate transitions from one orbit conﬁguration to another.
IV. ANALYTICAL SOLUTIONS
We will now solve the equations of motion (11) - (14) with the initial conditions
θ(γ0) = θ0,¯r(γ0) = ¯r0, φ(γ0) = φ0, t(γ0) = t0,(38)
In addition the initial direction, i.e. the sign of dx
(γ0) for x=θ, ¯r, has to be speciﬁed. We denote this by σx=
sgn dx
(γ0).
16
FIG. 7: Orbit conﬁgurations for the radial motion with ¯a= 0.6, P2= 0.4, ¯
K= 1, and e¯
Q= 1.1, 1 < e ¯
Q < q¯
K
¯
∆(0) .
For a general description of colours and linestyles see the text. The spherical orbits on solid lines starting at a dot
and approaching the black dash dotted lines or the light blue dashed lines are stable (two dark blue lines, a green
line, and a red line). All other spherical orbits are unstable. In the detailed plot on the upper right one of the solid
red lines approaches the light blue dashed line so close from below that they are hard to distinguish; the regions
indicated there are meant to be between them.
A. θ-motion
The equation of motion (11) needs to be solved. We ﬁrst concentrate on the case in which the poles are not reached,
i.e. θ(γ)(0, π). Then it is convenient to substitute ν= cos(θ), compare (17), and to solve the equivalent equation
of motion
2
=
4
X
i=0
biνi= Θν(ν) (39)
with ν(γ0) = ν0:= cos(θ0) and sgn
(γ0)=σν:= σθ. If Θν(ν) has a zero of multiplicity two or more the
solution can be solved by elementary functions. In general Θν(ν) is a polynomial of order four with simple zeroes
only and can be solved with the following procedure that uses the Weierstrass elliptic function (see [18, 20]. We
transform the equation to the Weierstrass form by the substitution ν= ( 4
a3ξa2
3a3)1+νΘ, where νΘis an arbitrary
zero of Θνand ai=1
(4i)!
d(4i)Θν
2
= 4ξ3gθ2ξgθ3(40)
17
(a) e¯
Q= 2 (b) e¯
Q= 10
FIG. 8: Orbit conﬁgurations for the radial motion with ¯a= 0.6, ¯
K= 1, P2= 0.4, and 1 <q¯
K
¯
∆(0) < e ¯
Q. For a
general description of colours and linestyles see the text. The spherical orbits on the blue (green) solid lines starting
at a dot and approaching the light blue dashed (black dash dotted) lines are stable. In addition, the spherical orbits
on the lower red solid line are stable but all others are unstable. Small plots on the top are enlarged details of the
lower plot.
with
gθ2=1
12a2
21
4a1a3, gθ3=1
216a3
2+1
48a1a2a31
16a0a2
3.(41)
The initial conditions are ξ(γ0) = ξ0:= 1
4(a3
ν0νΘ+a2
3) and sgn
(γ0)=σξ:= sgn(a3)σν. The solution of eq. (40)
can now be expressed in terms of the Weierstrass elliptic function,
ξ(γ) = (γγθ,in;gθ2;gθ3),(42)
where γθ,in is a constant such that (γ0γθ,in) = ξ0and sgn((γ0γθ ,in)) = σξ. This is e.g. fulﬁlled by γθ,in =
γ0σξRξ0
4τ3gθ2τgθ3
(with the principal branch of the square root). The solution for θis then given by
θ(γ) = arccos a3
4(γγθ,in;gθ2;gθ3)a2
3
+νΘ.(43)
Let us now consider orbits which reach the poles. First, let θ(γ) be in the open interval (0, π) for γ(γ1, γ2) but on the
endpoints γ1,2the orbit may reach the poles, θ(γ1,2)∈ {0, π}. Then the solution on (γ1, γ2), given by eq. (43), is also
valid on the complete interval [γ1, γ2] because the right hand side of eq. (43) is continuous on the whole closed interval
18
¯
K= 0 ¯
K= 0.02 ¯
K= 0.0684
¯
K= 0.08 ¯
K= 0.6¯
K= 1.0
FIG. 9: Orbit conﬁgurations for the radial motion with ¯a= 0.6, P2= 0.4, e¯
Q= 0.3 and varying ¯
K. For a general
description of colours and linestyles see the text. Note that at ¯
K= 0.0684 the plots change from case (ii) to (i).
P2= 0.1P2= 0.195 P2= 0.2
FIG. 10: Orbit conﬁgurations for the radial motion with ¯a= 0.6, ¯
K= 0.8, e¯
Q= 1.2 and varying P2. For a general
description of colours and linestyles see the text. Note that at P20.1956 the plots change from case (iii) to (iv).
19
P2= 0.5P2= 0.627 P2= 0.63
FIG. 11: Orbit conﬁgurations for the radial motion with ¯a= 0.6, ¯
K= 0.8, e¯
Q= 0.9 and varying P2. For a general
description of colours and linestyles see the text. Note that at P20.6277 the plots change from case (i) to (ii).
with limits θ(γ1,2) as γapproaches γ1,2. In general let γi,i1, be the parameters with θ(γi)∈ {0, π}and γi< γi+1 .
Deﬁne θi:= θ|[γi1i]and solve the diﬀerential equation in each interval with the condition θi(γi1) = θi1(γi1),
sgn i
(γi1)=sgn i1
(γi1). The switch in sign of i
in γicanonically identiﬁes θon [0, π].
B. ¯r-motion
The procedure to solve the equation of motion for the ¯r, see (12)
d¯r
2
=R2r)(ǫ¯r2+¯
K)¯
∆(¯r) = R(¯r) (44)
is analogous to that in the previous section. Again, the right hand side is a polynomial of fourth order. If it has a
zero of multiplicity two or more the diﬀerential equation can be solved in terms of elementary functions. The general
case can be solved with the substitution ¯r=c34ξc2
31+ ¯rR, where ¯rRis a zero of Rand ci=1
(4i)!
d(4i)R
d¯r(4i)rR).
¯r(γ) = c3
4(γγ¯r,in;g¯r2;g¯r3)c2
3
+ ¯rR(45)
with g¯r2,g¯r3given as in (41) with ai=ciand g¯ri=gθi. The parameter γ¯r,in only depends on the initial
conditions, (γ0γ¯r,in ;g¯r2;g¯r3) = 1
4c3
¯r0¯rR+c2
3and sgn((γ0γ¯r,in;g¯r2;g¯r3)) = sgn(c3)σ¯r, e.g. γ¯r,in =
γ0+ sgn(c3)σ¯rRξ0
4τ3g¯r2τg¯r3
.
C. φ-motion
The equation describing the φ-motion, see (13), can be rewritten as
φ(γ) = φ0+Zγ
γ0
¯aR(¯r)
¯
∆(¯r)Zγ
γ0
T(θ)
sin2θdγ , (46)
where the right hand side is separated in a part which only depends on ¯rand one that only depends on θ. We will
now treat both integrals separately.
20
The θ-dependent integral
Iφθ(γ) = Zγ
γ0
T(θ)
sin2θdγ. (47)
If we insert the expression for θgiven by eq. (43), which we write symbolically as θ=θ((γγθ,in)), we get
Iφθ(γ) = Zγ
γ0
T(θ((γγθ,in))
sin2θ((γγθ,in)) =Zγ
γ0
Rφθ((γγθ,in))(48)
with Rφ(θ((γγθ,in)) a rational function of (γγθ,in). A partial fraction decomposition then yields
Iφθ(γ) = Zγ
γ0
αθ+αθ1
(γγθ,in)βθ1
+αθ2
(γγθ,in)βθ2
(49)
with
αθ= ¯aE +e¯
P νΘ¯
L
1ν2
Θ
, αθ1,2=1
8
(e¯
P±¯
L)a4
(νΘ±1)2, βθ1,2=1
12
3a4±a3νΘa3
νΘ±1.(50)
The integral over each summand in eq. (49) can be expressed in terms of the Weierstrass ζ=ζ(·;gθ2;gθ3) and
σ=σ(·;gθ2;gθ3) function (see [20, 21])
Iφθ(γ) =αθ(γγ0) + αθ1
(v1)ln σ(γv1)
σ(γ0v1)ln σ(γ+v1)
σ(γ0+v1)+ 2(γγ0)ζ(v1)
+αθ2
(v2)ln σ(γv2)
σ(γ0v2)ln σ(γ+v2)
σ(γ0+v2)+ 2(γγ0)ζ(v2)(51)
where v1, v2need to chosen such that (v1+γθ ,in) = βθ1, and (v2+γθ,in) = βθ2.
The ¯r-dependent integral
The procedure to solve the ¯r-dependent integral
Iφ¯r(γ) = Zγ
γ0
¯aR(¯r)
¯
∆(¯r)
is analogous to the previous section. We substitute ¯r= ¯r((γγ¯r,in)) of eq. (45) and obtain a rational function as
integrand, Rφ¯rof (γγ¯r,in), which we decompose into partial fractions
Iφ¯r(γ) = Zγ
γ0
Rφ¯r((γγ¯r,in))(52)
=Zγ
γ0
α¯r+α¯r1
(γ)β¯r1
+α¯r2
(γ)β¯r2
dγ . (53)
Here the constants are given by
α¯r= ¯aR(¯rR)
¯
∆(¯rR), β¯r1,2=c3
12 +±pc2
4(1 + ¯a2+¯
Q2+¯
P2)c4¯r+c4
4¯
∆(¯r)(54)
and α¯r1,2the coeﬃcients of the partial fraction decomposition. The solution has the the form of eq. (51) with
αθ1,2=α¯r1,2,βθ1,2=β¯r1,2and v1, v2such that (v1+γ¯r,in) = β¯r1and (v2+γ¯r,in ) = β¯r2. Also the ,ζ, and σ
functions refer to the parameters g¯r2,g¯r3.
21
D. t-motion
The solution to the diﬀerential equation describing the t-motion, see (14), can be found in a similar way as for the
φmotion. An equivalent formulation of the diﬀerential equation is
t(γ) = t0+Zγ
γ0
r2+ ¯a2)Rr)
¯
∆(¯r)Zγ
γ0
¯aT(θ)dγ , (55)
where the right hand side is a again separated in an ¯rand a θdependent part. We treat these integrals separately.
The θ-dependent integral
We solve the integral
Itθ(γ) := Zγ
γ0
¯aT(θ)
with the same ansatz as in the foregoing section: First, we substitute θ=θ((γγθ,in )) of eq. (43) and then decompose
the integrand Rtφwhich is a rational function of (γγθ,in ) in partial fractions,
Itθ(γ) = Zγ
γ0
Rtθ((γγθ,in))
=Zγ
γ0
αθ+αθ1
(γ)βθ
+αθ2
((γ)βθ)2dγ , (56)
where
αθ=¯a¯
L¯a2Eν2
Θ+ ¯a2E+ ¯ae ¯
P νΘ, βθ=1
12a3,
αθ1=1
4¯aa4(2¯aEνΘe¯
P), αθ2=1
16¯a2Ea2
4.
(57)
The solution written in terms of ζ=ζ(·;gθ2;gθ3) and σ=σ(·;gθ2;gθ3) is given by
Itθ(γ) = αθ(γγ0) + αθ1
(v)ln σ(γv)
σ(γ0v)ln σ(γ+v)
σ(γ0+v)+ 2(γγ0)ζ(v)
+αθ2
2(v)ζ(γv) + ζ(γ0v)ζ(γ+v) + ζ(γ+v)(v)(γγ0)
′′(v)
2(v)ln σ(γv)
σ(γ0v)ln σ(γ+v)
σ(γ0+v)+ 2(γγ0)ζ(v) (58)
where vsuch that (v+γθ,in) = βθ.
The ¯r-dependent integral
Now the Integral
It¯r:= Zγ
γ0
r2+ ¯a2)Rr)
¯
∆(¯r)(59)
will be solved. The substitution ¯r= ¯r((γγ¯r,in)) of eq. (45) again leads to a rational function Rt¯rof (γγ¯r,in ),
which becomes after a partial fraction decomposition
It¯r=Zγ
γ0
Rφ¯r((γγ¯r,in))
=Zγ
γ0
α¯r+α¯r1
(γ)β¯r1
+α¯r2
(γ)β¯r2
dγ . (60)
Here α¯r,β¯r1,2are deﬁned as in eq. (54), and α¯r1,2are the coeﬃcients of the partial fraction decomposition. The
solution has the the form of eq. (51) with αθ1,2=α¯r1,2,βθ1,2=β¯r1,2and v1, v2such that (v1+γ¯r ,in) = β¯r1and
(v2+γ¯r,in) = β¯r2. The ,ζ, and σfunctions here refer to g¯r2and g¯r3.
22
FIG. 12: Charged particle orbit in Kerr-Newman space-time. Here a B+orbit in the northern hemisphere (Norbit)
with ¯a= 0.6, ¯
P= 0.47, ¯
Q= 0.16, e= 158.11, E= 1.98, ¯
L=62.62, ¯
K= 33.33 is shown. Left: three-dimensional
plot. The gray spheres correspond to the horizons. Right: projection on (x, z) plane. The dotted lines correspond to
extremal θvalues.
-1 012345
-1
1
2
3
4
5
6
FIG. 13: Charged particle orbit in Kerr-Newman space-time. Here a F
+orbit in the northern hemisphere (Norbit)
with ¯a= 0.6, ¯
P= 0.47, ¯
Q= 0.16, e= 158.11, E=11.83, ¯
L=5.18, ¯
K= 33.33 is shown. Left: three-dimensional
plot. The gray spheres correspond to the horizons. Middle: projection on (x, z) plane. The dotted lines correspond
to extremal θvalues. Right: projection on (x, y) plane. Dotted lines correspond to the horizons. Note that φ(γ)
diverges at the horizons at some ﬁnite γ0. In this plot we stopped at some γ0γ(∆γsmall) and continued with
γ0+ ∆γon the other side of the horizon.
E. Examples
The analytical solutions to the equations of motions are given by (43), (45), (51), and (58), respectively with the
appropriate constants (54). Here, we use these results to exemplify the orbit structure in Kerr-Newman space-time,
see Figs. 12 and 13.
23
V. SUMMARY AND CONCLUSIONS
In this paper we discussed the motion of charged particles in the gravitational ﬁeld of Kerr-Newman space-times
describing stationary rotating black holes with electric and magnetic charge. We demonstrated that it is suﬃcient to
consider test-particles with electric charge only as an additional magnetic charge would only lead to reparametrization.
After that we classiﬁed the orbits in radial and colatitudinal direction. For both a large variety of orbit conﬁgurations
was identiﬁed as summarized in tables I and II. In particular, we also identiﬁed orbits crossing the horizons or r= 0.
These conﬁgurations were then assigned to regions in the parameter space pictured in several ﬁgures. The boundaries
of these regions are amongst others given by parameter combinations which represent orbits of constant ror θ, which
were discussed in detail. For all orbit conﬁgurations analytical solutions to the equations of motion were presented in
terms of elliptic functions dependent on the Mino time.
For the sake of completeness we considered here a black hole endowed with magnetic charge although such was not
found until now. This has a big impact on the motion in θdirection in contrast to electric charge, which does not
inﬂuence the colatitudinal motion at all. Not only deviates the motion from the symmetry to the equatorial plane for
a nonvanishing magnetic charge but also additional types of orbits appear. For example, stable oﬀ-equatorial circular
orbits outside the horizon do exist in this case, which are not possible else [12]. They are given by the intersection
of a red solid line with E2<1 as in Fig. 1 and a green solid line with E2<1 as in Fig. 5. (E.g. ¯a= 0.5, Q= 0.3,
P= 0.6, e¯
P2.1, ¯
K= 1, E0.9, and ¯
L2.4 results in an orbit with constant ¯r= 3 and co s θ0.8.) On the
contrary, the magnetic charge does not inﬂuence the radial motion as is appears only in the combination ¯
P2+¯
Q2.
Still, to our knowledge the discussion of the radial motion in this paper is the most complete so far for Kerr-Newman
space-times.
The analytical solution presented here are largely based on the 19th century mathematics of elliptic functions
already used by Hagihara [15] to solve the geodesic equation in Schwarzschild space-time. However, a key ingredient
here is the introduction of the Mino time which decouples the radial and colatitudinal equations of motion. We
presented the results here in terms of Weierstrass elliptic functions which may be rewritten in terms of Jacobian
elliptic functions. The advantage of our presentation is that one formula is valid for all orbit conﬁgurations.
ACKNOWLEDGMENT
We would like to thank Claus L¨ammerzahl for suggesting this research topic and for helpful discussions. E.H.
acknowledges ﬁnancial support from the German Research Foundation DFG and support from the DFG Research
Training Group 1620 Models of Gravity.
[1] Newman, E. T. et a.: Metric of a Rotating, Charged Mass. In: J. Math. Phys. 6 (1965), S. 918
[2] Reissner, H.: ¨
Uber die Eigengravitation des elektrischen Feldes nach der Einsteinschen Theorie. In: Annalen der Physik
50 (1916), S. 106
[3] Nordstr¨
om, G.: On the Energy of the Gravitational Field in Einstein’s Theory. In: Verhandl. Koninkl. Ned. Akad.
Wetenschap., Afdel. Natuurk 26, S. 1201
[4] Kerr, R.P.: Gravitational ﬁeld of a spinning mass as an example of algebraically special metrics. In: Phys. Rev. Let. 11
(1963), S. 237
[5] Damour, T et a.: Regions of magnetic support of a plasma around a black-hole. In: Phys. Rev. D 17 (1978), Nr. 6, S.
1518
[6] Ruffini, R. et a.: Electron-positron pairs in physics and astrophysics: From heavy nuclei to black holes. In: Physics
Reports 487 (2010), S. 1
[7] Johnston, M. ; Ruffini, R.: Generalized Wilkins eﬀect and selected orbits in a Kerr-Newman geometry. In: Phys. Rev.
D10 (1974), S. 2324
[8] Young, P.J.: Capture of particles from plunge orbits by a black hole. In: Phys. Rev. D 14 (1976), S. 3281
[9] Sharp, N. A.: Geodesics in Black Hole Space-Times. In: Gen. Rel. Grav.
[10] Biˇ
c´
ak, J. et a.: The motion of charged particles in the ﬁeld of rotating charged black holes and naked singularities - I.
The general features of the radial motion and the motion along the axis of symmetry. In: Bull. Astron. Inst. Czechosl. 40
(1989), S. 65
[11] Biˇ
c´
ak, J. et a.: The motion of charged particles in the ﬁeld of rotating charged black holes and naked singularities - II.
The motion in the equatorial plane. In: Bull. Astron. Inst. Czechosl. 40 (1989), S. 133
[12] Kov´
aˇ
r, J. et a.: Oﬀ-equatorial orbits in strong gravitational ﬁelds near compact objects. In: Class. Quant. Grav. 25
(2008), S. 095011
24
[13] Calvanie, M. ; Turolla, R.: Complete description of photon trajectories in the Kerr-Newman space-time. In: J. Phys.
A: Math. Gen. 14 (1981), S. 1931
[14] Carter, B.: Global Structure of the Kerr Family of Gravitational Fields. In: Physical Review 174 (1968), October, Nr.
5, S. 1559–1571
[15] Hagihara, Y: Theory of the Relativistic Trajectories in a Gravitational Field of Schwarzschild. In: Japanese Journal of
Astronomy and Geophysics 8 (1931), S. 67–176
[16] Mino, Y.: Perturbative approach to an orbital evolution around a supermassive black hole. In: Phys. Rev. D 67 (2003),
S. 084027
[17] Fujita, R. ; Hikida, W.: Analytical solutions of bound timelike geodesic orbits in Kerr spacetime. In: Class. Quantum
Grav. 26 (2009), S. 135002
[18] Hackmann, E.: Geodesic equations in black hole space-times with cosmological constant, Universit¨at Bremen, Dissertation,
2010
[19] Hawking, S. W. ; Ellis, G.F.R.: The large scale structure of space-time. Cambridge : Cambridge Univ. P., 1973
[20] Markushevich, A. I.: Theory of functions of a complex variable. Vol. I, II, III. English. New York : Chelsea Publishing
Co., 1977. – ISBN 0–8284–0296–5. – Translated and edited by Richard A. Silverman
[21] Hackmann, E. ; L¨
ammerzahl, C. ; Kagramanova, K. ; Kunz, J.: Analytical solution of the geodesic equation in
Kerr-(anti-) de Sitter space-times. In: Phys. Rev. D 81 (2010), S. 044020
... Substituting the contravariant components g µν , (13), (14), (15), (16), (17) into (37), we have: ...
... that has been thoroughly studied by Hackmann in [17]. The motion is allowed almost in the whole θ range, with the exception of the pole at θ = 0; varyingẼ changes the value of dθ/dλ at the equatorial plane (θ = π/2), such that in-creasingẼ decreases the angular velocity of the test particle, dθ/dλ in the equatorial plane and its vicinity. ...
... One of the positive roots is in the interior of the BH, and the larger radius of the turning points defines the region of the bounded orbits. Geodesics of the KN spacetime are analyzed in [17], see also [18]. ...
Preprint
Full-text available
In this paper the motion of charged and uncharged test particles in the rotating nonlinearly charged black hole is examined. Its asymptotics can be de Sitter or anti-de Sitter, depending on the value of the nonlinear parameter; consequently this BH can present one, two or three horizons, the third one being the cosmological horizon in the de Sitter case. Angular and radial test particle motions are analyzed and compared with its linear electromagnetic counterpart, the Kerr-Newman black hole (KN-BH). Several differences arise with the KN-BH, namely, the equatorial asymmetry is enhanced by the NLE field and for charged particles the access to one of the poles is forbidden; besides, a second circular orbit in the neighborhood of the external horizon appears; the presence of the nonlinear electromagnetic field increses the curvature producing bounded orbits closer to the horizon.
... Their solutions are expressed in terms of explicit quadratures. Although such a covariant vector potential is replaced with a more complicated form, the integrability of charged particle motions in Kerr-Newmann spacetimes was shown by Hackmann and Xu [49]. In other words, the fourth constant of motion is still existent. ...
... Apart from the two Kerr type black holes with external magnetic fields mentioned in [19,49], the dynamics of charged particles moving around the Kerr-Newman black hole surrounded by cloud strings, quintessence, and electromagnetic field [18] is integrable. Providing such an integrable example is the main motivation of the present paper. ...
... In this case, E AE are the standard radial effective potentials at the equatorial plane θ ¼ π=2 in many references (e.g., [31][32][33][34][35][36]). Solving Eqs. (49) and (55) [or Eq. (42) with θ ¼ π=2], we have the energies ...
... Apart from the study of photons and massive particles, researchers have shown considerable interest in the study of charged particles around black hole spacetime due to the presence of the Maxwell fields leading to additional interacting terms. The study of charged particles around Kerr-Newman black hole has been carried out in [30][31][32][33]. Also the motion of the charged particles in case of black holes immersed in external electromagnetic fields has been studied in [34][35][36][37][38]. ...
... After studying the null-geodesics and geodesics of the chargeless massive particles, we move on to study the geodesic motion of the massive charged particles. Incorporating the interactions of the gauge fields, the Hamiltonian of the particle in this case gets modified to [33] 2H = g μν p μ + qA μ (p ν + qA ν ) = = −1 (60) with the electromagnetic potential for charged spinning black hole coupled to PFDM given by [38] ...
Article
Full-text available
In this work we have obtained a charged black hole solution in the presence of perfect fluid dark matter (PFDM) and discuss its energy conditions. The metric corresponding to the rotating avatar of this black hole solution is obtained by incorporating the Newman–Janis algorithm. We then compute two types of circular geodesics, namely, the null geodesics and time-like geodesics for this rotating spacetime geometry. For the case of time-like geodesics, we consider both neutral as well as charged massive particles. The effective potentials of the corresponding circular geodesics has also been studied. We then present our results by graphically representing the collective effects of the black hole parameters, namely, the charge of the black hole (Q), spin parameter (a) and the PFDM parameter (α) on the energy (E), angular momentum (L) and effective potential (V eff) of the concerned particle. Finally, we discuss the Penrose process in order to study the negative energy particles having possible existence within the ergosphere, and which in turn leads to the energy gain of the emitted particle.
... For many well known spacetimes like Schwarzschild and Kerr the equations of motion are of elliptic type (see, e.g., [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] and references therein). These can easily be solved analytically in terms of the elliptic and therefore doubly-periodic Weierstraß ℘-function. ...
Preprint
Analysis of black hole spacetimes requires study of the motion of particles and light in these spacetimes. Here exact solutions of the geodesic equations are the means of choice. Numerous interesting black hole spacetimes have been analyzed in terms of elliptic functions. However, the presence of a cosmological constant, higher dimensions or alternative gravity theories often necessitate an analysis in terms of hyperelliptic functions. Here we review the method and current status for solving the geodesic equations for the general hyperelliptic case, illustrating it with set a of examples of genus $g=2$.
... If the external magnetic fields do not destroy the symmetry, no chaotic dynamics is possible. For example, charged particle motions in the Kerr-Newman black hole spacetime are regular and nonchaotic because of the existence of four integrals leading to the integrability of the system [36]. Chaos describes a dynamical system sensitive dependence on initial conditions. ...
Preprint
Full-text available
We mainly focus on the effects of small changes of parameters on the dynamics of charged particles around the Kerr black hole surrounded by an external magnetic field, which can be considered as a tidal environment. The radial motions of charged particles on the equatorial plane are studied via an effective potential. It is found that the particle energies at the local maxima values of the effective potentials increase with an increase of the black hole spin and the particle angular momenta, but decrease with an increase of one of the inductive charge parameter and magnetic field parameter. The radii of stable circular orbits on the equatorial plane also increase, whereas those of the innermost stable circular orbits decrease. On the other hand, the effects of small variations of the parameters on the orbital regular and chaotic dynamics of charged particles on the non-equatorial plane are traced by means of a time-transformed explicit symplectic integrator, Poincar\'{e} sections and fast Lyapunov indicators. It is shown that the dynamics sensitively depends on small variations of the inductive charge parameter, magnetic field parameter, energy and angular momentum. Chaos occurs easily as each of the dynamical parameters increases. When the dragging effects of the spacetime increase, the chaotic properties are not always weakened under some circumstances.
Article
Full-text available
Analysis of black hole spacetimes requires study of the motion of particles and light in these spacetimes. Here exact solutions of the geodesic equations are the means of choice. Numerous interesting black hole spacetimes have been analyzed in terms of elliptic functions. However, the presence of a cosmological constant, higher dimensions or alternative gravity theories often necessitate an analysis in terms of hyperelliptic functions. Here we review the method and current status for solving the geodesic equations for the general hyperelliptic case, illustrating it with a set of examples of genus g=2: higher dimensional Schwarzschild black holes, rotating dyonic U(1)2 black holes, and black rings.
Article
Full-text available
Considering electrically charged test particles, we continue our study of the exterior dynamics of a charged Weyl black hole which has been previously investigated regarding the motion of mass-less and (neutral) massive particles. In this paper, the deflecting trajectories of charged particles are designated as being gravitationally Rutherford-scattered and detailed discussion of angular and radial particle motions is presented. © 2021, The Author(s), under exclusive licence to Società Italiana di Fisica and Springer-Verlag GmbH Germany, part of Springer Nature.
Article
In this article we study the innermost stable circular orbit (ISCO) of electrically charged particles in the electrically charged Reissner-Nordström spacetime, the Kerr-Newman spacetime and the Kerr-Sen spacetime. We find that the radius of the ISCO increases with an increasing particle-black hole charge product |qQ| in the case of attractive Coulomb interaction qQ<0. For repulsive Coulomb interaction, the ISCO radius first decreases to a minimum and then increases again, until it diverges as the charge product approaches one. If the charge Q of the black hole is very small, the minimum of the ISCO radius lies at qQ=0. Repulsive and attractive Coulomb interactions will always increase the ISCO radius in this limit. Stable bound orbits of charged particles cease to exist in the Reissner-Nordström and Kerr-Newman spacetime for qQ≥1. In the Kerr-Sen spacetime the limiting case depends on the charge of the black hole and if dilaton coupling is applied to the test particle. We find qQ≥1+Q2 without dilaton-coupling and qQ≥1+3/2 Q2 with dilaton coupling α=1.
Article
Full-text available
Algebraically special solutions of Einstein's empty-space field ; equations that are characterized by the existence of a geodesic and shear-free ; ray congruence are considered. A class of solutions is presented for which the ; congruence is diverging and is not necessarily hypersurface orthogonal. (C.E.S.);
Article
The Kerr family of solutions of the Einstein and Einstein-Maxwell equations is the most general class of solutions known at present which could represent the field of a rotating neutral or electrically charged body in asymptotically flat space. When the charge and specific angular momentum are small compared with the mass, the part of the manifold which is stationary in the strict sense is incomplete at a Killing horizon. Analytically extended manifolds are constructed in order to remove this incompleteness. Some general methods for the analysis of causal behavior are described and applied. It is shown that in all except the spherically symmetric cases there is nontrivial causality violation, i.e., there are closed timelike lines which are not removable by taking a covering space; moreover, when the charge or angular momentum is so large that there are no Killing horizons, this causality violation is of the most flagrant possible kind in that it is possible to connect any event to any other by a future-directed timelike line. Although the symmetries provide only three constants of the motion, a fourth one turns out to be obtainable from the unexpected separability of the Hamilton-Jacobi equation, with the result that the equations, not only of geodesics but also of charged-particle orbits, can be integrated completely in terms of explicit quadratures. This makes it possible to prove that in the extended manifolds all geodesics which do not reach the central ring singularities are complete, and also that those timelike or null geodesics which do reach the singularities are entirely confined to the equator, with the further restriction, in the charged case, that they be null with a certain uniquely determined direction. The physical significance of these results is briefly discussed.
Article
Wilkins has pointed out that, in addition to the periodicities associated with circular orbits in the equatorial plane of a Kerr geometry, there exist periodicities connected with the longitudinal motion of particles. We extend this result to spherical orbits of charged particles in a Kerr-Newman geometry. We give explicit examples of nonspherical orbits, illustrating dragging of inertial frames and these longitudinal periodicities. The relevance of these results for the physics of collapsed objects is discussed.
Article
Photon and "parabolic" particle orbits around a Kerr-Newman black hole are considered, both for uncharged particles moving along geodesics and for charged particles under the influence of the Lorentz force (neglecting radiative processes). Investigation of the radial equation of motion gives conditions under which a photon or particle is captured from a "plunge" orbit when incident from a large distance; the cross sections and accreted angular momenta are calculated for various fluxes of particles incident on the black hole. For example, the cross section of an extreme Kerr (a=1) black hole to an isotropic flux of particles is 0.90 of that for a Schwarzschild hole, and the accreted angular momentum per unit mass of swallowed flux is -0.828 leading to rapid spin-down of the hole. The periastra of "escape" orbits are considered, and also the minimum periastron corresponding to the unstable spherical orbit that divides plunge and escape orbits. The evolution of the spin and charge of a black hole accreting particles and photons is discussed, including that of primordial black holes. It is shown that such black holes may be of the Schwarzschild type having spun-down due to consumption of radiation and particles at early times and subsequent neutralization in an ionized intergalactic medium. The statistics of proton and electron capture by the smallest surviving primordial black holes, M=1015 g, suggests that they are most likely to be found possessing a single quantum of positive charge.
Article
The motion in the equatorial plane is discussed in three cases which exhibit well the interaction of charged particles and Kerr-Newman black holes. These are: (1) the circular orbits of ultrarelativistic particles with high specific charge; (2) the motion of charged particles with zero angular momentum (where the interplay between the gravitational and magnetic dragging can be seen in the simplest form); and (3) the motion of charged particles with l = a x gamma (including all particles that can hit the ring singularity and also the cases when the trajectories can be given in terms of elementary functions).
Article
A new solution of the Einstein-Maxwell equations is presented. This solution has certain characteristics that correspond to a rotating ring of mass and charge.