This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

Inspiral and Plunging Orbits in Kerr-Newman Spacetimes

Yu-Chung Ko    Da-Shin Lee [email protected]    Chi-Yong Lin [email protected] Department of Physics, National Dong Hwa University, Hualien, Taiwan, Republic of China
Abstract

We present the analytical solutions for the trajectories of particles that spiral and plunge inward the event horizon along the timelike geodesics following general non-equatorial paths within Kerr-Newman spacetimes. Our studies encompass both bound and unbound motions. The solutions can be written in terms of the elliptical integrals and the Jacobian elliptic functions of manifestly real functions of the Mino time. They can respectively reduce to the Kerr, Reissner-Nordstro¨\ddot{o}m, and Schwarzschild black holes in certain limits of the spin and charge of the black holes, and can be compared with the known ones restricted in equatorial motion. These explicit solutions may have some implications for the gravitational wave emission from extreme mass-ratio inspirals.

pacs:
04.70.-s, 04.70.Bw, 04.80.Cc

I Introduction

Recent detections of gravitational waves emitted from the merger of binary systems have confirmed Einstein’s century-old prediction as a consequence of general relativity [1, 2, 3]. The capture of the spectacular images of supermassive black holes M87* at the center of M87 galaxy [4] and Sgr A* at the center of our galaxy [5] leads to another scientific achievement of a direct evidence of the existence of black holes. The black hole is one of the mysterious stellar objects, and it is the solution derived from the Einstein’s field equations [6, 7]. In astrophysics, extreme mass-ratio inspirals (EMRIs), which consist of a stellar mass object orbiting around a massive black hole, have recently received considerable attention. The goal is to analyze gravitational wave signals to accurately test the predictions of general relativity in its strong regime. Gravitational wave signals generated through EMRIs, which are key sources of low frequency gravitational waves and to be observed in the planned space-based Laser Interferometer Space Antenna (LISA), provides an opportunity to measure various fascinating properties of supermassive black holes [8, 9, 10, 11, 12].

The present work is motivated by EMRIs, which can be approximated as a light body travels along the geodesic of the background spacetime of a massive black hole. In particular, the recent studies in [13, 14] have been devoted to inspirals of the particle on the equatorial plane asymptotically from the innermost stable circular orbits (ISCO) of Kerr black holes. They also derive a simple expression for the equatorial radial flow from the ISCO relevant to the dynamics of the accretion disk. These exact solutions may have applications to the generated gravitational waveforms arising from EMRIs [10] as well as constructing theories of black hole accretion [15, 16, 18, 17, 19]. Additionally, the work by [20, 21] broadens the investigation of equatorial motion to encompass generic non-equatorial orbits in Kerr black holes. In the family of the Kerr black holes due to the underlying spacetime symmetry the dynamics possesses two conserved quantities, the energy EmE_{m} and the azimuthal angular momentum LmL_{m} of the particle. Nevertheless, the existence of the third conserved quantity, discovered in the sixties and known nowadays as Carter constant, renders the geodesic equations as a set of first-order differential equations [22]. Later, the introduction of the Mino time [23] further fully decouples the equations of motions with the solutions given in terms of the elliptical functions [24, 25]. In our previous paper [26], we have studied the null and time-like geodesics of the light and the neutral particles respectively in the exterior of Kerr-Newman black holes. We then obtain the solutions of the trajectories written in terms of the elliptical integrals and the Jacobi elliptic functions [27], in which the orbits are manifestly real functions of the Mino time and also the initial conditions can be explicitly specified [28]. In this work, we will mainly focus on the infalling particles into the Kerr-Newman black holes in general nonequatorial motion. Theoretical considerations, together with recent observations of structures near Sgr A* by the GRAVITY experiment [29], indicate possible presence of a small electric charge of central supermassive black hole [30, 31]. Thus, it is of great interest to explore the geodesic dynamics in the Kerr-Newman black hole [32].

Layout of the paper is as follows. In Sec. II, a mini review of the time-like geodesic equations is provided with three conserved quantities of a particle, the energy, azimuthal angular momentum, and Carter constant. The equations of motion can be recast into integral forms involving two effective potentials. In particular, the roots of the radial potential determine the different types of the infalling trajectories of particles to black holes [33]. Sec. III focuses on the parameter regime of the conserved quantities to have triple roots giving the radius of the innermost stable spherical orbits (ISSO) [34]. The analytical solutions of the infalling orbits are derived for the case that the particle starts from the coordinate rr slightly less than that of the ISSO. The two additional infalling orbits of interest are determined by the roots of the radial potential, consisting of a pair of complex roots and two real roots. In Sec. IV, we consider bound motion with one of the real roots inside the horizon and the other outside the horizon. In this case, the particle motion is bound by the turning point from the real root outside the horizon. In Sec. V we will consider unbound motion, in which both real roots are inside the event horizon. The exact solutions for plunging trajectories are given and illustrative examples for such trajectories are plotted. In VI the conclusions are drawn. For the clarity of notation and the completeness of the paper, Appendixes A and B provide some of relevant formulas derived in the earlier publications [26, 36].

II Equation of motion for time-like geodesics

We start from a summary of the equations of motion for the particle in the Kerr-Newman black hole exterior. We work with the Boyer-Lindquist coordinates (t,r,θ,ϕ)(t,r,\theta,\phi). The spacetime of the exterior of the Kerr-Newman black hole with the gravitational mass MM, angular momentum JJ, and angular momentum per unit mass a=J/Ma=J/M is described by the metric

ds2=ΔΣ(dtasin2θdϕ)2+sin2θΣ[(r2+a2)dϕadt]2+ΣΔdr2+Σdθ2,ds^{2}=-\frac{\Delta}{\Sigma}\left(dt-a\sin^{2}\theta d\phi\right)^{2}+\frac{\sin^{2}\theta}{\Sigma}\left[(r^{2}+a^{2})d\phi-adt\right]^{2}+\frac{\Sigma}{\Delta}dr^{2}+\Sigma d\theta^{2}\;, (1)

where Σ=r2+a2cos2θ\Sigma=r^{2}+a^{2}\cos^{2}\theta and Δ=r22Mr+a2+Q2\Delta=r^{2}-2Mr+a^{2}+Q^{2}. The roots of Δ(r)=0\Delta(r)=0 determine outer/inner event horizons r+/rr_{+}/r_{-} as

r±=M±M2(Q2+a2).r_{\pm}=M\pm\sqrt{M^{2}-(Q^{2}+a^{2})}\;. (2)

We assume that 0<a2+Q2M20<a^{2}+Q^{2}\leq M^{2} throughout the paper.

For the asymptotically flat, stationary, and axial-symmetric black holes, the metric is independent of tt and ϕ\phi. Thus, the conserved quantities are energy EmE_{m} and azimuthal angular momentum LmL_{m} of the particle along a geodesic. These can be constructed through the four momentum pμ=muμ=mdxμ/dσmp^{\mu}=mu^{\mu}=m\,dx^{\mu}/d\sigma_{m}, defined in terms of the proper time σm\sigma_{m} and the mass of the particle mm as

Em\displaystyle E_{m} pt,\displaystyle\equiv-p_{t}, (3)
Lm\displaystyle L_{m} pϕ.\displaystyle\equiv p_{\phi}\,. (4)

Additionally, another conserved quantity is the Carter constant explicitly obtained by

Cm=Σ2(uθ)2a2cos2θ(Em)2+Lm2cot2θ+m2a2cos2θ.{C}_{m}=\Sigma^{2}\left(u^{\theta}\right)^{2}-a^{2}\cos^{2}\theta\left({E_{m}}\right)^{2}+L_{m}^{2}\cot^{2}\theta+m^{2}a^{2}\cos^{2}\theta\,. (5)

Together with the time-like geodesics, uμuμ=m2u^{\mu}u_{\mu}=m^{2}, one obtains the equations of motion

Σmdrdσm=±rRm(r),\frac{\Sigma}{m}\frac{d{r}}{d\sigma_{m}}=\pm_{r}\sqrt{R_{m}({r})}\,, (6)
Σmdθdσm=±θΘm(θ),\frac{\Sigma}{m}\frac{d\theta}{d\sigma_{m}}=\pm_{\theta}\sqrt{\Theta_{m}(\theta)}\,, (7)
Σmdϕdσm=aΔ[(r2+a2)γmaλm]1sin2θ(aγmsin2θλm),\frac{\Sigma}{m}\frac{d\phi}{d\sigma_{m}}=\frac{{a}}{{\Delta}}\left[\left({r}^{2}+{a}^{2}\right)\gamma_{m}-{a}\lambda_{m}\right]-\frac{1}{\sin^{2}\theta}\left({a}\gamma_{m}\sin^{2}\theta-\lambda_{m}\right)\,, (8)
Σmdtdσm=r2+a2Δ[(r2+a2)γmaλm]a(aγmsin2θλm),\frac{\Sigma}{m}\frac{d{t}}{d\sigma_{m}}=\frac{{r}^{2}+{a}^{2}}{{\Delta}}\left[\left({r}^{2}+{a}^{2}\right)\gamma_{m}-{a}\lambda_{m}\right]-{a}\left({a}\gamma_{m}\sin^{2}\theta-\lambda_{m}\right)\,, (9)

where we have normalized EmE_{m}, LmL_{m}, and CmC_{m} by the mass of the particle mm

γmEmm,λmLmm,ηmCmm2.\displaystyle\gamma_{m}\equiv\frac{E_{m}}{m},\hskip 5.69054pt\lambda_{m}\equiv\frac{L_{m}}{m},\hskip 5.69054pt\eta_{m}\equiv\frac{{C}_{m}}{m^{2}}. (10)

The symbols ±r=sign(ur)\pm_{r}={\rm sign}\left(u^{r}\right) and ±θ=sign(uθ)\pm_{\theta}={\rm sign}\left(u^{\theta}\right) are defined by the 4-velocity of the particle. Moreover, the radial potential Rm(r)R_{m}({r}) in (6) and and angular potential Θm(θ)\Theta_{m}(\theta) in (7) are obtained as

Rm(r)=[(r2+a2)γmaλm]2Δ[ηm+(aγmλm)2+r2],R_{m}({r})=\left[\left({r}^{2}+{a}^{2}\right)\gamma_{m}-{a}\lambda_{m}\right]^{2}-{\Delta}\left[\eta_{m}+\left({a}\gamma_{m}-\lambda_{m}\right)^{2}+{r}^{2}\right]\,, (11)
Θm(θ)=ηm+a2γm2cos2θλm2cot2θa2cos2θ.\Theta_{m}(\theta)=\eta_{m}+{a}^{2}\gamma_{m}^{2}\cos^{2}\theta-\lambda_{m}^{2}\cot^{2}\theta-{a}^{2}\cos^{2}\theta\,. (12)

As well known [23], the set of equations of motion (6)-(9) can be fully decoupled by introducing the so-called Mino time τm\tau_{m} defined as

dxμdτmΣmdxμdσm.\frac{dx^{\mu}}{d\tau_{m}}\equiv\frac{\Sigma}{m}\frac{dx^{\mu}}{d\sigma_{m}}\,\,. (13)

For the source point xiμx_{i}^{\mu} and observer point xμx^{\mu}, the integral forms of the equations above can be rewritten as [28]

τmτmi=Imr=Gmθ,\tau_{m}-\tau_{mi}=I_{mr}=G_{m\theta}\,, (14)
ϕmϕmi=Imϕ+λmGmϕ,\phi_{m}-\phi_{mi}=I_{m\phi}+{\lambda_{m}}G_{m\phi}\,, (15)
tmtmi=Imt+a2γmGmt,t_{m}-t_{mi}=I_{mt}+a^{2}{\gamma_{m}}G_{mt}\,, (16)

where the integrals ImrI_{mr}, ImϕI_{m\phi}, and ImtI_{mt} involve the radial potential Rm(r)R_{m}(r)

Imrrir1±rRm(r)𝑑r,,I_{mr}\equiv\int_{r_{i}}^{r}\frac{1}{\pm_{r}\sqrt{R_{m}(r)}}dr,\,, (17)
Imϕrira[(2MrQ2)γmaλm]±rΔRm(r)𝑑r,I_{m\phi}\equiv\int_{r_{i}}^{r}\frac{{a\left[\left(2Mr-Q^{2}\right)\gamma_{m}-a\lambda_{m}\right]}}{\pm_{r}\Delta\sqrt{R_{m}(r)}}dr, (18)
Imtrirr2γmΔ+(2MrQ2)[(r2+a2)γmaλm]±rΔRm(r)𝑑r.I_{mt}\equiv\int_{r_{i}}^{r}\frac{{r^{2}\gamma_{m}\Delta+(2Mr-Q^{2})\left[\left(r^{2}+a^{2}\right)\gamma_{m}-a{\lambda_{m}}\right]}}{\pm_{r}\Delta\sqrt{R_{m}(r)}}dr\,. (19)

The angular integrals are

Gmθθiθ1±θΘm(θ)𝑑θ,G_{m\theta}\equiv\int_{\theta_{i}}^{\theta}\frac{1}{\pm_{\theta}\sqrt{\Theta_{m}(\theta)}}d\theta\,, (20)
Gmϕθiθcsc2θ±θΘm(θ)𝑑θ,G_{m\phi}\equiv\int_{\theta_{i}}^{\theta}\frac{\csc^{2}\theta}{\pm_{\theta}\sqrt{\Theta_{m}(\theta)}}d\theta\,, (21)
Gmtθiθcos2θ±θΘm(θ)𝑑θ.G_{mt}\equiv\int_{\theta_{i}}^{\theta}\frac{\cos^{2}\theta}{\pm_{\theta}\sqrt{\Theta_{m}(\theta)}}d\theta\,. (22)

The radial potential Rm(r)R_{m}(r) is given by a quartic polynomial and the types of its roots play essential roles in the classification of different orbits. In the previous work [26, 36], we have shown the exact solutions to some of the cases of both null and time-like geodesics. For the present work, we will mainly focus on the infalling orbits of the bound and the unbound motion at the black hole exterior. In this context, we will consider three types of such orbits in the subsequent sections. The discussion of the angular potential Θ(θ)\Theta(\theta) and the integrals involved has been presented in Ref. [26]. For the sake of completeness, we will provide a short summary in Appendix A.

Before ending this section, let us introduce a few notations that will be used in the subsequent sections. Related to Rm(r)R_{m}(r) we define the integrals

Inrirrn1γm2Rm(r)𝑑riInU,n=1,2,I_{n}\equiv\int_{r_{i}}^{r}r^{n}\sqrt{\frac{1-\gamma_{m}^{2}}{R_{m}(r)}}\,dr\equiv iI^{U}_{n}\;,\;n=1,2\,, (23)
I±rir1(rr±)1γm2Rm(r)𝑑riI±U.I_{\pm}\equiv\int_{r_{i}}^{r}\frac{1}{\left(r-r_{\pm}\right)}\sqrt{\frac{1-\gamma_{m}^{2}}{R_{m}(r)}}\,dr\equiv iI^{U}_{\pm}\,. (24)

In terms of I1I_{1}, I2I_{2}, and I±I_{\pm} we can rewrite (18) and (19) as follows

Imϕ(τm)=γm1γm22Mar+r[(r+a(λmγm)+Q22M)I+(τm)(ra(λmγm)+Q22M)I(τm)],I_{m\phi}(\tau_{m})=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\frac{2Ma}{r_{+}-r_{-}}\left[\left(r_{+}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{+}(\tau_{m})-\left(r_{-}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{-}(\tau_{m})\right]\;, (25)
Imt(τm)=γm1γm2{4M2r+r[(r+Q22M)(r+a(λmγm)+Q22M)I+(τm)\displaystyle I_{mt}(\tau_{m})=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\left\{\frac{4M^{2}}{r_{+}-r_{-}}\left[\left(r_{+}-\frac{Q^{2}}{2M}\right)\left(r_{+}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{+}(\tau_{m})\right.\right.
(rQ22M)(ra(λmγm)+Q22M)I(τm)]+2MI1(τm)+I2(τm)}\displaystyle\quad\quad\quad\quad\quad\quad\quad\left.\left.-\left(r_{-}-\frac{Q^{2}}{2M}\right)\left(r_{-}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{-}(\tau_{m})\right]+2MI_{1}(\tau_{m})+I_{2}(\tau_{m})\right\}
+(4M2Q2)γmτm.\displaystyle\quad\quad\quad\quad\quad\quad\quad+\left(4M^{2}-Q^{2}\right)\gamma_{m}\tau_{m}\;. (26)

III INSPIRAL ORBITS IN BOUND MOTION

According to previous studies of radial potential [26], by examining different ranges of the parameters λm\lambda_{m} and ηm\eta_{m} for bound motion (γm<1\gamma_{m}<1), it is clear that there are two distinct categories of infalling motion that traverse the horizon and enter the black hole. One is that the particle starts from ririssor_{i}\leq r_{\rm isso}, where the radius of the ISSO orbit rissor_{\rm isso} is within the parameters located at A and B in Fig. 1, spirals and then plunge into the horizon of the black holes [37, 38]. The other one is starting from ri<rm4r_{i}<r_{m4}, with the parameters of C and D in Fig. 5, and plunge through the horizon of the black holes.

Refer to caption
Figure 1: The main graphics shows the parametric plot of λm(risso)\lambda_{m}(r_{\rm isso}) versus ηm(risso)\eta_{m}(r_{\rm isso}). The triple roots rissor_{\rm isso} are the solutions of the equations Rm′′(r)=Rm=Rm(r)=0R^{\prime\prime}_{m}(r)=R^{\prime}_{m}=R_{m}(r)=0. The inset illustrates the behavior of the radial potential RmR_{m} with the parameters located at A and B. The case of B has ηm=0\eta_{m}=0, which is an example of equatorial motion.

We first consider the particle starts from ririssor_{i}\leq r_{\rm isso}. The solutions along the rr direction can be obtained from the inversion of (14) with the integral ImrI_{mr} in (17), where the radial potential (11) is given in the case of the triple root located at the ISSO radius, namely rm2=rm3=rm4=rissor_{m2}=r_{m3}=r_{m4}=r_{\rm isso}, and the initial rr is set at ririssor_{i}\leq r_{\rm isso}. So, from the integral (14) and (17), we can have the Mino time τm\tau_{m} as the function of rr

τmI(r)=2(rissorm1)1γm2[rrm1rissorrirm1rissori].\tau^{I}_{m}(r)=\frac{-2}{(r_{\rm isso}-r_{m1})\sqrt{1-\gamma_{m}^{2}}}\Bigg{[}\sqrt{\frac{r-r_{m1}}{r_{\rm isso}-r}}-\sqrt{\frac{r_{i}-r_{m1}}{r_{\rm isso}-r_{i}}}\Bigg{]}\,. (27)

The particle moves toward the horizon with νri=1\nu_{r_{i}}=-1. Thus, the inverse of (27) leads to

rI(τm)=rm1+risso[XI(τm)]21+[XI(τm)]2,r^{I}(\tau_{m})=\frac{r_{m1}+r_{\rm isso}\left[X^{I}(\tau_{m})]^{2}\right.}{1+{\left[X^{I}(\tau_{m})]^{2}\right.}}\;, (28)

where

XI(τm)=1γm2(rissorm1)2τmrirm1rissori.X^{I}(\tau_{m})=\frac{\sqrt{1-\gamma_{m}^{2}}(r_{\rm isso}-r_{m1})}{2}\tau_{m}-\sqrt{\frac{r_{i}-r_{m1}}{r_{\rm isso}-r_{i}}}\,. (29)

The solution (28) of the coordinate rr involves the triple root rissor_{\rm isso} of the radial potential, which can be determined as follows.

From the double root solutions R(r)=R(r)=0R(r)=R^{\prime}(r)=0 [26], two constants of motion in the case of spherical orbits are given by

λmss=[rmss(MrmssQ2)a2M]γmΔ(rmss)rmss2(γm21)+Mrmssa(rmssM),\lambda_{\rm mss}=\frac{\left[r_{\rm mss}\left(Mr_{\rm mss}-Q^{2}\right)-a^{2}M\right]\gamma_{m}-\Delta\left(r_{\rm mss}\right)\sqrt{r_{\rm mss}^{2}\left(\gamma_{m}^{2}-1\right)+Mr_{\rm mss}}}{a\left(r_{\rm mss}-M\right)}\;, (30)
ηmss=rmssa2(rmssM)2{rmss(MrmssQ2)(a2+Q2Mrmss)γm2.\displaystyle{\eta}_{\rm mss}=\frac{r_{\rm mss}}{a^{2}\left(r_{\rm mss}-M\right)^{2}}\Big{\{}r_{\rm mss}\left(Mr_{\rm mss}-Q^{2}\right)\left(a^{2}+Q^{2}-Mr_{\rm mss}\right)\gamma_{m}^{2}\Big{.}
+2(MrmssQ2)Δ(rmss)γmrmss2(γm21)+Mrmss\displaystyle\quad\quad\quad\quad\quad+2\left(Mr_{\rm mss}-Q^{2}\right)\Delta\left(r_{\rm mss}\right)\gamma_{m}\sqrt{r_{\rm mss}^{2}\left(\gamma_{m}^{2}-1\right)+Mr_{\rm mss}}
.+[a2(MrmssQ2)(Δ(rmss)a2)2][rmss(γm21)+M]}.\displaystyle\Big{.}\quad\quad\quad\quad\quad\left.+\left[a^{2}\left(Mr_{\rm mss}-Q^{2}\right)-\left(\Delta\left(r_{\rm mss}\right)-a^{2}\right)^{2}\right]\left[r_{\rm mss}\left(\gamma_{m}^{2}-1\right)+M\right]\Big{\}}\right.\;. (31)

The subscript ”ss” means the spherical orbits with s=±s=\pm, which denotes the two types of motion with respect to the relative sign between the black hole’s spin and the azimuthal angular of the particle (see Section III C of [26]). Together with the two relations above, an additional equation from R′′(r)=0R^{\prime\prime}(r)=0 determines the triple roots. We have found the radius of rissor_{\rm isso} satisfying the following equation

Mrisso5Δ(risso)+4(Mrisso3Q2risso2+a2ηissoasΓms)2=0,-Mr_{\rm isso}^{5}\Delta\left(r_{\rm isso}\right)+4\left(Mr_{\rm isso}^{3}-Q^{2}r_{\rm isso}^{2}+a^{2}\eta_{\rm isso}-as\sqrt{\Gamma_{\rm ms}}\right)^{2}=0\,, (32)

where

Γms=risso4(MrissoQ2)ηisso[risso(risso3M)+2Q2]risso2+a2ηisso2.\Gamma_{\rm ms}=r_{\rm isso}^{4}\left(Mr_{\rm isso}-Q^{2}\right)-\eta_{\rm isso}\left[r_{\rm isso}\left(r_{\rm isso}-3M\right)+2Q^{2}\right]r_{\rm isso}^{2}+a^{2}\eta_{\rm isso}^{2}\,. (33)

We proceed by evaluating the coordinates ϕm(τm)\phi_{m}(\tau_{m}) and tm(τm)t_{m}(\tau_{m}) using (15) and (16), which involve not only the angular integrals GmϕG_{m\phi} and GmtG_{mt}, but also the radial integrals (18) and (19). With the help of (25) and (26), we first rewrite (18) and (19) as

ImϕI(τm)=γm1γm22Mar+r[(r+a(λmγm)+Q22M)I+I(τm)(ra(λmγm)+Q22M)II(τm)],I^{I}_{m\phi}(\tau_{m})=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\frac{2Ma}{r_{+}-r_{-}}\left[\left(r_{+}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{+}^{I}(\tau_{m})-\left(r_{-}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{-}^{I}(\tau_{m})\right]\,,\\ (34)
ImtI(τm)=γm1γm2{4M2r+r[(r+Q22M)(r+a(λmγm)+Q22M)I+I(τm)\displaystyle I^{I}_{mt}(\tau_{m})=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\left\{\frac{4M^{2}}{r_{+}-r_{-}}\left[\left(r_{+}-\frac{Q^{2}}{2M}\right)\left(r_{+}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{+}^{I}(\tau_{m})\right.\right.
(rQ22M)(ra(λmγm)+Q22M)II(τm)]+2MI1I(τm)+I2I(τm)}\displaystyle\quad\quad\quad\quad\quad\quad\left.\left.-\left(r_{-}-\frac{Q^{2}}{2M}\right)\left(r_{-}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)I_{-}^{I}(\tau_{m})\right]+2MI_{1}^{I}(\tau_{m})+I_{2}^{I}(\tau_{m})\right\}
+(4M2Q2)γmτm.\displaystyle\quad\quad\quad\quad\quad\quad+\left(4M^{2}-Q^{2}\right)\gamma_{m}\tau_{m}\;. (35)

For the present case of the triple roots, the calculation of the integrals is straightforward and one can express InII_{n}^{I} and I±II_{\pm}^{I} in terms of elementary functions,

I±I(τm)=1γm2rissor±τm+1(r±rm1)(rissor±)3tanh1(r±rm1)(rissorI(τm))(rissor±)(rI(τm)rm1)±iI,I_{\pm}^{I}(\tau_{m})=\frac{\sqrt{1-\gamma_{m}^{2}}}{r_{\rm isso}-r_{\pm}}\tau_{m}+\frac{1}{\sqrt{(r_{\pm}-r_{m1})\left(r_{\rm isso}-r_{\pm}\right)^{3}}}\tanh^{-1}\sqrt{\frac{(r_{\pm}-r_{m1})(r_{\rm isso}-r^{I}(\tau_{m}))}{(r_{\rm isso}-r_{\pm})(r^{I}(\tau_{m})-r_{m1})}}-\mathcal{I}_{\pm_{i}}^{I}\,, (36)
I1I(τm)=1γm2rissoτm+2tan1rissorI(τm)rI(τm)rm11iI,I_{1}^{I}(\tau_{m})=\sqrt{1-\gamma_{m}^{2}}r_{\rm isso}\tau_{m}+2\tan^{-1}\sqrt{\frac{r_{\rm isso}-r^{I}(\tau_{m})}{r^{I}(\tau_{m})-r_{m1}}}-\mathcal{I}_{1_{i}}^{I}\,, (37)
I2I(τm)=rI(τm)(rm1risso)+risso(3rissorm1)2τm+(rm1+3risso)tan1rissorI(τm)rI(τm)rm12iI.I_{2}^{I}(\tau_{m})=\frac{r_{I}(\tau_{m})(r_{m1}-r_{\rm isso})+r_{\rm isso}(3r_{\rm isso}-r_{m1})}{2}\tau_{m}+(r_{m1}+3r_{\rm isso})\tan^{-1}\sqrt{\frac{r_{\rm isso}-r^{I}(\tau_{m})}{r^{I}(\tau_{m})-r_{m1}}}-\mathcal{I}_{2_{i}}^{I}\,. (38)

It is worthwhile to mention that ±iI\mathcal{I}_{\pm_{i}}^{I}, 1iI\mathcal{I}_{1_{i}}^{I}, 2iI\mathcal{I}_{2_{i}}^{I} are obtained by evaluating I±I{I}_{\pm}^{I}, I1I{I}_{1}^{I}, I2I{I}_{2}^{I} at r=rir=r_{i} of the initial condition, that is, I±I(0)=I1I(0)=I2I(0)=0{{I}_{\pm}^{I}(0)={I}_{1}^{I}(0)={I}_{2}^{I}(0)=0}. The solutions of ϕI(τm)\phi^{I}(\tau_{m}) and tI(τm)t^{I}(\tau_{m}) can be constructed from ImϕI_{m\phi} (18), GmϕG_{m\phi} (21) and ImtI_{mt} (19) and GmtG_{mt} (22) through (15) and (16). Together with the solutions along rr and θ\theta directions in (28) and (90), they are infalling motions of the general nonequatorial orbits in the Kerr-Newman exterior. An illustrative example is shown in Fig. 2 with the parameters of the case A in Fig. 1. The particle starts by inspiraling around rissor_{\rm isso} and then plunges into the black hole’s horizon.

Refer to caption
Figure 2: An illustrative example of nonequatorial orbit with parameters A of Fig. 1. In this case, the particle starts from ririssor_{i}\leq r_{\rm isso} and inspirals into the black hole after many azimuthal and longitudinal revolutions. From the top view one notices the very different time scales of the spiralling and plunging phases

For the particle initially at ri=rissor_{i}=r_{\rm isso}, the solution (28) gives r(τ)=rissor(\tau)=r_{\rm isso} obtained from XIX^{I}\rightarrow-\infty in (29) when ririssor_{i}\rightarrow r_{\rm isso}, meaning that the particle will stay in spherical motion with the rissor_{\rm isso} radius. However, for ri<rissor_{i}<r_{\rm isso} of our interest, as rr reaches the outer horizon r+r_{+}, it takes finite Mino time τm\tau_{m}. Nevertheless, because of the tanh1\tanh^{-1} function in (36), I±issoI_{\pm}^{\rm isso}\rightarrow\infty as rr+r\rightarrow r_{+}, giving the coordinate time tt\rightarrow\infty and the azimuthal angle ϕ\phi\rightarrow\infty observed in the asymptotical flat regime. The above expressions can be further reduced to the Kerr black hole case by sending Q0Q\rightarrow 0 in [20]. As for the contributions to the evolution of the angle ϕ\phi in (15) from the integrals involving the radial potential Rm(r)R_{m}(r), one can write the tanh1\tanh^{-1} function in I±I(τm)I_{\pm}^{I}(\tau_{m}) by log\log function through tanh1(x)=12log(1+x)log(1x)\tanh^{-1}(x)=\frac{1}{2}\frac{\log(1+x)}{\log(1-x)}. Then, the remaining terms in (34) directly proportional to τm\tau_{m} are the same in [20]. Together with (93) of the integrals involving the θ\theta potential Θm(θ)\Theta_{m}(\theta), the variables z1z_{1} and z2z_{2} defined in [20] can be related to the roots of Θm(θ)\Theta_{m}(\theta) by z12=um+z_{1}^{2}=u_{m+} and z22=uma2(1γm2)z_{2}^{2}=u_{m-}a^{2}(1-\gamma_{m}^{2}), leading to kz2=um+umk_{z}^{2}=\frac{u_{m+}}{u_{m-}}, which gives the same expression in [20] from (93).

One of the interesting cases is considering the equatorial motion by taking θ=π2\theta=\frac{\pi}{2} and ηm0\eta_{m}\rightarrow 0 limits in the results above. The particle starts from the coordinate rr slightly less than the radius of innermost circular motion riscor_{\rm isco}. In particular, Gmϕ=τmG_{m\phi}=\tau_{m} and the solution of ϕmI\phi^{I}_{m} in (15) simplifies to [26, 36]

ϕmI(r)=ImϕI(τm(r))+λmτmI(r)+ϕmiI,\phi^{I}_{m}\left(r\right)=I^{I}_{m\phi}\left(\tau_{m}\left(r\right)\right)+\lambda_{m}\tau^{I}_{m}\left(r\right)+\phi^{I}_{mi}\,, (39)

where ImϕII^{I}_{m\phi} is given by (34). In addition, one can replace Mino time τm\tau_{m} by the coordinate rr through Eq. (27). Then the infalling solution of the angle ϕm\phi_{m} on the equatorial plane can be expressed as a function of rr,

ϕmI(r)=2rrm1(1γm2)(riscor)risco2λm+(2MriscoQ2)(aγmλm)(riscor+)(riscor)(riscorm1)\displaystyle\phi^{I}_{m}(r)=-2\sqrt{\frac{r-r_{m1}}{\left(1-\gamma_{m}^{2}\right)\left(r_{\rm isco}-r\right)}}\frac{r_{\rm isco}^{2}\lambda_{m}+\left(2Mr_{\rm isco}-Q^{2}\right)\left(a\gamma_{m}-\lambda_{m}\right)}{\left(r_{\rm isco}-r_{+}\right)\left(r_{\rm isco}-r_{-}\right)\left(r_{\rm isco}-r_{m1}\right)}
2r+r(2Maγmrλm)r+Q2(aγmλm)(riscor+)(1γm2)(r+rm1)(riscor+)tanh1(r+rm1)(riscor)(riscor+)(rrm1)\displaystyle-\frac{2}{r_{+}-r_{-}}\frac{\left(2Ma\gamma_{m}-r_{-}\lambda_{m}\right)r_{+}-Q^{2}\left(a\gamma_{m}-\lambda_{m}\right)}{\left(r_{\rm isco}-r_{+}\right)\sqrt{\left(1-\gamma_{m}^{2}\right)\left(r_{+}-r_{m1}\right)\left(r_{\rm isco}-r_{+}\right)}}\tanh^{-1}\sqrt{\frac{\left(r_{+}-r_{m1}\right)\left(r_{\rm isco}-r\right)}{\left(r_{\rm isco}-r_{+}\right)\left(r-r_{m1}\right)}}
+2r+r(2Maγmr+λm)rQ2(aγmλm)(riscor)(1γm2)(rrm1)(riscor)tanh1(rrm1)(riscor)(riscor)(rrm1).\displaystyle+\frac{2}{r_{+}-r_{-}}\frac{\left(2Ma\gamma_{m}-r_{+}\lambda_{m}\right)r_{-}-Q^{2}\left(a\gamma_{m}-\lambda_{m}\right)}{\left(r_{\rm isco}-r_{-}\right)\sqrt{\left(1-\gamma_{m}^{2}\right)\left(r_{-}-r_{m1}\right)\left(r_{\rm isco}-r_{-}\right)}}\tanh^{-1}\sqrt{\frac{\left(r_{-}-r_{m1}\right)\left(r_{\rm isco}-r\right)}{\left(r_{\rm isco}-r_{-}\right)\left(r-r_{m1}\right)}}\,. (40)

Likewise, for the equatorial orbits, Eq. (16) with Gmt=0G_{mt}=0 gives

tmI(r)=ImtI(τm)+tmiI,t^{I}_{m}\left(r\right)=I^{\rm I}_{mt}\left(\tau_{m}\right)+t^{I}_{mi}\,, (41)

where ImtII^{I}_{mt} has been calculated in (35). Substituting τmI\tau^{I}_{m} to replace rr, we find

tmI(r)=γm(rrm1)(riscor)1γm2+γm(rm1+3risco+4M)1γm2tan1riscorrrm1\displaystyle t^{I}_{m}\left(r\right)=-\gamma_{m}\sqrt{\frac{\left(r-r_{m1}\right)\left(r_{\rm isco}-r\right)}{1-\gamma_{m}^{2}}}+\frac{\gamma_{m}\left(r_{m1}+3r_{\rm isco}+4M\right)}{\sqrt{1-\gamma_{m}^{2}}}\tan^{-1}\sqrt{\frac{r_{\rm isco}-r}{r-r_{m1}}}
2rrm1(1γm2)(riscor)risco2(risco2+a2)γm+(2MriscoQ2)a(aγmλm)(riscor+)(riscor)(riscorm1)\displaystyle-2\sqrt{\frac{r-r_{m1}}{\left(1-\gamma_{m}^{2}\right)\left(r_{\rm isco}-r\right)}}\frac{r_{\rm isco}^{2}\left(r_{\rm isco}^{2}+a^{2}\right)\gamma_{m}+\left(2Mr_{\rm isco}-Q^{2}\right)a\left(a\gamma_{m}-\lambda_{m}\right)}{\left(r_{\rm isco}-r_{+}\right)\left(r_{\rm isco}-r_{-}\right)\left(r_{\rm isco}-r_{m1}\right)}
2(2Mr+Q2)r+r2Mγmr+(aλm+Q2γm)(riscor+)(1γm2)(r+rm1)(riscor+)tanh1(r+rm1)(riscor)(riscor+)(rrm1)\displaystyle-\frac{2\left(2Mr_{+}-Q^{2}\right)}{r_{+}-r_{-}}\frac{2M\gamma_{m}r_{+}-\left(a\lambda_{m}+Q^{2}\gamma_{m}\right)}{\left(r_{\rm isco}-r_{+}\right)\sqrt{\left(1-\gamma_{m}^{2}\right)\left(r_{+}-r_{m1}\right)\left(r_{\rm isco}-r_{+}\right)}}\tanh^{-1}\sqrt{\frac{\left(r_{+}-r_{m1}\right)\left(r_{\rm isco}-r\right)}{\left(r_{\rm isco}-r_{+}\right)\left(r-r_{m1}\right)}}
+2(2MrQ2)r+r2Mγmr(aλm+Q2γm)(riscor)(1γm2)(rrm1)(riscor)tanh1(rrm1)(riscor)(riscor)(rrm1).\displaystyle+\frac{2\left(2Mr_{-}-Q^{2}\right)}{r_{+}-r_{-}}\frac{2M\gamma_{m}r_{-}-\left(a\lambda_{m}+Q^{2}\gamma_{m}\right)}{\left(r_{\rm isco}-r_{-}\right)\sqrt{\left(1-\gamma_{m}^{2}\right)\left(r_{-}-r_{m1}\right)\left(r_{\rm isco}-r_{-}\right)}}\tanh^{-1}\sqrt{\frac{\left(r_{-}-r_{m1}\right)\left(r_{\rm isco}-r\right)}{\left(r_{\rm isco}-r_{-}\right)\left(r-r_{m1}\right)}}\,. (42)

As for the initial conditions one can determine ϕmiI\phi^{I}_{mi} and tmiIt^{I}_{mi} by ImϕI(τmI(r))+λmτmi(r)I^{I}_{m\phi}\left(\tau^{I}_{m}\left(r\right)\right)+\lambda_{m}\tau^{i}_{m}\left(r\right) and ImtI(τm)I^{I}_{mt}\left(\tau_{m}\right) vanishing at the initial rir_{i}. The corresponding trajectories are shown in Fig. 3, with the additional parameter QQ apart from aa of the black holes. This generalizes the solution in [13] for the Kerr black holes, where the particle starts from rriscor\lesssim r_{\rm isco} at tm(r)=t_{m}(r)=-\infty and inspirals to the event horizon. In the limit of Q0Q\rightarrow 0 where rm10r_{m1}\rightarrow 0, 1γm22M/3risco1-\gamma_{m}^{2}\rightarrow{2M}/{3r_{\rm isco}}, and (riscor+)(riscor)risco22Mrisco+a2(r_{\rm isco}-r_{+})(r_{\rm isco}-r_{-})\rightarrow r^{2}_{\rm isco}-2Mr_{\rm isco}+a^{2}, the first term of Eq.(40) reduces to the corresponding term rriscor\sqrt{\frac{r}{r_{\rm isco}-r}} of ϕmI(r)\phi^{I}_{m}(r) in [13]. Together with t±=r±riscor±t_{\pm}=\sqrt{\frac{r_{\pm}}{r_{\rm isco}-r_{\pm}}} defined in [13], one can make the replacement

1r+r1t+2t2riscorisco22Mrisco+a2,\frac{1}{r_{+}-r_{-}}\rightarrow\frac{1}{t^{2}_{+}-t^{2}_{-}}\frac{r_{\rm isco}}{r^{2}_{\rm isco}-2Mr_{\rm isco}+a^{2}}\,, (43)

to reproduce the tanh1\tanh^{-1} terms.

One of the limiting cases that can significantly simplify the above expressions is to consider the extremal limit of the Kerr black hole, aMa\rightarrow M. For Q0Q\rightarrow 0 giving rm1=0r_{m1}=0, and for the extremal black holes, the ISCO radius for direct orbits is on the event horizon. Here we focus on the extremal retrograde motion with risco=9Mr_{\rm isco}=9M, λm=223M/9\lambda_{m}=-22\sqrt{3}M/9 and γm=53/9\gamma_{m}=5\sqrt{3}/9. Notice that in the extremal black holes, r+r_{+} and rr_{-} collapse into the same value. Then the solutions can be reduced into the known ones [13, 14]

ϕmI(r)=223r32(rM)9Mr,\phi^{I}_{m}\left(r\right)=-\frac{2\sqrt{2}}{3}\frac{r^{\frac{3}{2}}}{(r-M)\sqrt{9M-r}}\,, (44)
tmI(r)=(9Mr)r2(4M5rrM)11722Mr9Mr\displaystyle t^{I}_{m}\left(r\right)=\sqrt{\frac{(9M-r)r}{2}}\left(\frac{4M-5r}{r-M}\right)-\frac{117\sqrt{2}}{2}M\sqrt{\frac{r}{9M-r}}
+15522Mtan19Mrr4Mtanh19Mr8r.\displaystyle\quad\quad\quad\quad+\frac{155\sqrt{2}}{2}M\tan^{-1}\sqrt{\frac{9M-r}{r}}-4M\tanh^{-1}\sqrt{\frac{9M-r}{8r}}\,. (45)
Refer to caption
Figure 3: Illustration of the orbit on the equatorial plane with the parameters of B in Fig. 1. The particle starts from ri<rissor_{i}<r_{\rm isso} and inspirals into the black hole horizon.

Another limiting case is considering the Reissner Nordstro¨\ddot{o}m (RN) black hole. Since a0a\rightarrow 0 leading to the spherically symmetric metric, the general motion can be studied by considering equatorial motions. The coefficients of the tanh1\tanh^{-1} terms of the above expressions (40) all vanish. The expressions of ϕmI(r)\phi^{I}_{m}\left(r\right) and tmI(r)t^{I}_{m}\left(r\right) can be simplified as

ϕmI(r)=2λmriscorm1rrm1(1γm2)(riscor),\phi^{I}_{m}\left(r\right)=-\frac{2\lambda_{m}}{r_{\rm isco}-r_{m1}}\sqrt{\frac{r-r_{m1}}{(1-\gamma_{m}^{2})(r_{\rm isco}-r)}}\,, (46)
tmI(r)=\displaystyle t^{I}_{m}\left(r\right)= γm(rrm1)(riscor)1γm2+γm(rm1+3risco+4M)1γm2tan1riscorrrm1\displaystyle-\gamma_{m}\sqrt{\frac{\left(r-r_{m1}\right)\left(r_{\rm isco}-r\right)}{1-\gamma_{m}^{2}}}+\frac{\gamma_{m}\left(r_{m1}+3r_{\rm isco}+4M\right)}{\sqrt{1-\gamma_{m}^{2}}}\tan^{-1}\sqrt{\frac{r_{\rm isco}-r}{r-r_{m1}}}
2rrm1(1γm2)(riscor)risco4γm(riscor+)(riscor)(riscorm1)\displaystyle-2\sqrt{\frac{r-r_{m1}}{\left(1-\gamma_{m}^{2}\right)\left(r_{\rm isco}-r\right)}}\frac{r_{\rm isco}^{4}\gamma_{m}}{\left(r_{\rm isco}-r_{+}\right)\left(r_{\rm isco}-r_{-}\right)\left(r_{\rm isco}-r_{m1}\right)}
2r+r(2Mr+Q2)2γm(riscor+)(1γm2)(r+rm1)(riscor+)tanh1(r+rm1)(riscor)(riscor+)(rrm1)\displaystyle-\frac{2}{r_{+}-r_{-}}\frac{\left(2Mr_{+}-Q^{2}\right)^{2}\gamma_{m}}{\left(r_{\rm isco}-r_{+}\right)\sqrt{\left(1-\gamma_{m}^{2}\right)\left(r_{+}-r_{m1}\right)\left(r_{\rm isco}-r_{+}\right)}}\tanh^{-1}\sqrt{\frac{\left(r_{+}-r_{m1}\right)\left(r_{\rm isco}-r\right)}{\left(r_{\rm isco}-r_{+}\right)\left(r-r_{m1}\right)}}
+2r+r(2MrQ2)2γm(riscor)(1γm2)(rrm1)(riscor)tanh1(rrm1)(riscor)(riscor)(rrm1).\displaystyle+\frac{2}{r_{+}-r_{-}}\frac{\left(2Mr_{-}-Q^{2}\right)^{2}\gamma_{m}}{\left(r_{\rm isco}-r_{-}\right)\sqrt{\left(1-\gamma_{m}^{2}\right)\left(r_{-}-r_{m1}\right)\left(r_{\rm isco}-r_{-}\right)}}\tanh^{-1}\sqrt{\frac{\left(r_{-}-r_{m1}\right)\left(r_{\rm isco}-r\right)}{\left(r_{\rm isco}-r_{-}\right)\left(r-r_{m1}\right)}}\,. (47)

Further simplification occurs in the extremal limit. For M=±QM=\pm Q in the RN black holes, r±=Mr_{\pm}=M, and with risco=4Mr_{\rm isco}=4M, rm1=4M/5r_{m1}=4M/5, λm=22M\lambda_{m}=2\sqrt{2}M and γm=36/8\gamma_{m}=3\sqrt{6}/8, (46) and (47) can have such a simple form

ϕmI(r)=25r4M4Mr,\phi^{I}_{m}\left(r\right)=-2\sqrt{\frac{5r-4M}{4M-r}}\,, (48)
tmI(r)=\displaystyle t^{I}_{m}\left(r\right)= 33(4Mr)(r4M/5)5+2521525Mtan14Mrr4M/5\displaystyle-3\sqrt{\frac{3(4M-r)(r-4M/5)}{5}}+\frac{252\sqrt{15}}{25}M\tan^{-1}\sqrt{\frac{4M-r}{r-4M/5}}
32M5r4M12M3r(2M21)2(Mr)M3(4Mr)(5r4M)3\displaystyle-32M\sqrt{\frac{5r-4M}{12M-3r}}-\frac{(2M^{2}-1)^{2}}{(M-r)M^{3}}\sqrt{\frac{(4M-r)(5r-4M)}{3}}
4(2M21)M3tanh14Mr15r12M.\displaystyle-\frac{4(2M^{2}-1)}{M^{3}}\tanh^{-1}\sqrt{\frac{4M-r}{15r-12M}}\,. (49)

Finally, in the limits of Q0Q\rightarrow 0 and a0a\rightarrow 0 we have the case of Schwarzschild black hole, with the spherical symmetric metric. The horizons become r+2Mr_{+}\rightarrow 2M and r0r_{-}\rightarrow 0, and the general motion can be considered in the equatorial plane. Thus, with the further ISCO inputs in this case, risco=6Mr_{\rm isco}=6M, λm=23M\lambda_{m}=2\sqrt{3}M, and γm=22/3\gamma_{m}=2\sqrt{2}/3, the solutions become as simple as

ϕmI(r)=23r6Mr,\phi^{I}_{m}\left(r\right)=-2\sqrt{3}\sqrt{\frac{r}{6M-r}}\;, (50)
tmI(r)\displaystyle t^{I}_{m}\left(r\right) =8642M25r6Mr22(6Mr)r\displaystyle=\frac{864\sqrt{2}M}{25}\sqrt{\frac{r}{6M-r}}-2\sqrt{2}\sqrt{(6M-r)r}
+442Mtan16Mrr4Mtanh16Mr2r.\displaystyle+44\sqrt{2}M\tan^{-1}\sqrt{\frac{6M-r}{r}}-4M\tanh^{-1}\sqrt{\frac{6M-r}{2r}}\;. (51)

We then recover the results of two recent publications [13, 14].

Refer to caption
Figure 4: The plots show the variation of azimuthal angle ΔϕmIϕmIϕmiI\Delta\phi_{m}^{I}\equiv\phi_{m}^{I}-\phi^{I}_{mi} as a function of rr for the inspire bound motion for various black-hole models, including Kerr-Newman (Q=0.7M,a=0.7M,λm=3.84M,ηm=0MQ=0.7M,a=0.7M,\lambda_{m}=-3.84M,\eta_{m}=0M), Kerr-extremal (Q=0M,a=M,λm=2239M,ηm=0MQ=0M,a=M,\lambda_{m}=-\frac{22\sqrt{3}}{9}M,\eta_{m}=0M) and RN (Q=0.7M,a=0M,λm=3.2M,ηm=0MQ=0.7M,a=0M,\lambda_{m}=3.2M,\eta_{m}=0M). In the figure, the blue curve and red curves illustrate the direct and retrograde orbits, respectively. See the text for more discussion.

It is then of great interest to plot and observe how the azimuthal angle ϕ\phi depends on the evaluation of coordinate rr, as illustrated for a few exemplary cases in Fig. 4. In particular, when the motion approaches the horizon, the angle ϕ\phi diverges for the spinning black holes whereas it remains finite for non-spinning black holes. From Eq. (46) for RN black holes and Eq. (50) for Schwarzschild black holes, ϕ\phi smoothly changes across the horizon. Thus, another usefulness of the obtained result in (40) is to examine the behavior of ϕ\phi across the horizon. The divergence in the azimuthal angle arises from the tanh1\tanh^{-1} terms and the straightforward calculations shows that ϕmIln(rr+)\phi_{m}^{I}\sim\ln(r-r_{+}). However, for the extremal Kerr-Newman black holes where r+=r=Mr_{+}=r_{-}=M, the additional divergence in the coefficients of the tanh1\tanh^{-1} terms in (40) shifts the leading order divergence into that of ϕmI1/(rM)\phi_{m}^{I}\sim 1/(r-M) in (44), apart from the ln(rM)\ln(r-M) divergence. However, for the extremal Kerr black holes, ϕmI1/(rM)\phi_{m}^{I}\sim 1/(r-M) from (44), where the ln(rM)\ln(r-M) divergence disappears. The dramatic difference in the behavior of the azimuthal angle ϕ\phi across the horizon has been found on equatorial infalling trajectories in Kerr black holes [13, 14]. The same types of phenomena are also seen on infalling trajectories of general nonequatorial orbits in Kerr-Newman black holes. Notice that in all cases, the coordinate time tt\rightarrow\infty as rr+r\rightarrow r_{+}. This finding may have some implications for the gravitational wave emission measured by an observer far away from the black holes [10] as in the study of another interesting trajectories of homoclinic orbits in [35, 36].

IV Plunging ORBITS IN BOUND MOTION

Another bound orbit, in which particles eventually fall into the black hole, is the motion with the parameters of C and D in Fig. 5. In this case, there are two real roots, being rm1r_{m1} inside the inner horizon, rm4r_{m4} outside the outer horizon, and a pair of the complex-conjugated roots rm2=rm3r_{m2}=r_{m3}^{*}. Assuming that the particle starts from rirm4r_{i}\leq r_{m4}, it will either plunge directly into the black hole or travel toward the root rm4r_{m4}, return back and plunge into the black hole, in the absence of any other real root along its trajectory. This section is devoted to finding the analytical solution for the orbit in this case of rm2=rm3r_{m2}={r}^{*}_{m3} and rm4>ri>r+>r>rm1r_{m4}>r_{i}>r_{+}>r_{-}>r_{m1}.

Refer to caption
Figure 5: The graphics shows the portion of parameter space bound by the double root solution, rm2=rm3r_{m2}=r_{m3}. The equation Rm(r)R_{m}(r)=0 with parameters in the blue zone have complex roots, rm2=rm3r_{m2}=r_{m3}^{*}, so that, a particle in this region, say C or D, and starts from ri<r4r_{i}<r_{4} will plunge directly into the black hole horizon. The inset shows the behavior of the radial potential Rm(r)R_{m}(r) for the case of the parameters located in C and D.

The solutions in the present cases are expressed in a similar form as in the previous section. The integration of (14) is straightforward, but elliptical integrals and the Jacobi elliptic functions are involved for the representation of the solutions [27]. We find after some algebra

τmB(r)=1(1γm2)AmBm(F(φ(r)|kB)F(φ(ri)|kB))\tau_{m}^{B}(r)=-\frac{1}{\sqrt{(1-\gamma_{m}^{2})A_{m}B_{m}}}\left(F\left(\varphi(r)|k^{B}\right)-F\left(\varphi(r_{i})|k^{B}\right)\right) (52)

where F(φ|k)F(\varphi|k) is the incomplete elliptic integral of the first kind. The two parameters of the elliptic integrals are

φ(r)=cos1(Bm(rm4r)Am(rrm1)Bm(rm4r)+Am(rrm1))\varphi(r)=\cos^{-1}\left(\frac{B_{m}(r_{m4}-r)-A_{m}(r-r_{m1})}{B_{m}(r_{m4}-r)+A_{m}(r-r_{m1})}\right) (53)

and

kB=(rm4rm1)2(AmBm)24AmBm,k^{B}=\frac{(r_{m4}-r_{m1})^{2}-(A_{m}-B_{m})^{2}}{4A_{m}B_{m}}\;, (54)

where we have used the short notations

Am=(rm4rm2)(rm4rm3),Bm=(rm3rm1)(rm2rm1).A_{m}=\sqrt{(r_{m4}-r_{m2})(r_{m4}-r_{m3})}\hskip 2.84526pt,\hskip 2.84526ptB_{m}=\sqrt{(r_{m3}-r_{m1})(r_{m2}-r_{m1})}\,. (55)

With the help of the Jacobian elliptic cosine function [27] one finds the inversion of (52) as

rB(τm)=(Bmrm4+Amrm1)(Bmrm4Amrm1)cn(XB(τm)|kB)(Bm+Am)(BmAm)cn(XB(τm)|kB),r^{B}(\tau_{m})=\frac{(B_{m}r_{m4}+A_{m}r_{m1})-(B_{m}r_{m4}-A_{m}r_{m1}){\rm cn}\left(X^{B}(\tau_{m})\left|k^{B}\right)\right.}{(B_{m}+A_{m})-(B_{m}-A_{m}){\rm cn}\left(X^{B}(\tau_{m})\left|k^{B}\right)\right.}\;, (56)

where

XB(τm)=(1γm2)AmBmτmF(cos1(Bm(rm4ri)Am(rirm1)Bm(rm4ri)+Am(rirm1))|kB).\displaystyle X^{B}(\tau_{m})=\sqrt{\left(1-\gamma_{m}^{2}\right)A_{m}B_{m}}\tau_{m}-F\Bigg{(}\cos^{-1}\left(\frac{B_{m}(r_{m4}-r_{i})-A_{m}(r_{i}-r_{m1})}{B_{m}(r_{m4}-r_{i})+A_{m}(r_{i}-r_{m1})}\right)\left|k^{B}\Bigg{)}\right.\,. (57)

Notice that Am>Bm>0A_{m}>B_{m}>0, 0<kB<10<k^{B}<1, and for r<rm4r<r_{m4}, 1<Bm(rm4ri)Am(rirm1)Bm(rm4ri)+Am(rirm1)<1-1<\frac{B_{m}(r_{m4}-r_{i})-A_{m}(r_{i}-r_{m1})}{B_{m}(r_{m4}-r_{i})+A_{m}(r_{i}-r_{m1})}<1. The Jacobian elliptic cosine function is the real-valued function.

The solutions of the coordinates ϕmB(τm)\phi_{m}^{B}(\tau_{m}) and tmB(τm)t^{B}_{m}(\tau_{m}) involve the integrals ImϕBI_{m\phi}^{B} and ImtBI_{mt}^{B} given in (25) and (26). In the present case, the integration of I1BI_{1}^{B}, I2BI_{2}^{B}, and I±BI_{\pm}^{B} is direct, but the results have cumbersome representations:

I±B(τm)=1Bm(rm4r±)+Am(r±rm1)[BmAmAmBmXB(τm)\displaystyle I_{\pm}^{B}(\tau_{m})=\frac{1}{B_{m}\left(r_{m4}-r_{\pm}\right)+A_{m}\left(r_{\pm}-r_{m1}\right)}\left[\frac{B_{m}-A_{m}}{\sqrt{A_{m}B_{m}}}X^{B}(\tau_{m})\right.
+2(rm4rm1)AmBmBm(rm4r±)Am(r±rm1)R1(β±B;ΥτmB|kB)]±iB\displaystyle\quad\quad\quad\quad\quad\quad\quad\left.+\frac{2(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{B_{m}\left(r_{m4}-r_{\pm}\right)-A_{m}\left(r_{\pm}-r_{m1}\right)}R_{1}(\beta_{\pm}^{B};\Upsilon_{\tau_{m}}^{B}|k^{B})\right]-\mathcal{I}_{\pm_{i}}^{B} (58)
I1B(τm)=(Bmrm4Amrm1BmAm)XB(τm)AmBm+2(rm4rm1)AmBmAm2Bm2R1(βB;ΥτmB|kB)1iB\displaystyle I_{1}^{B}(\tau_{m})=\left(\frac{B_{m}r_{m4}-A_{m}r_{m1}}{B_{m}-A_{m}}\right)\frac{X^{B}(\tau_{m})}{\sqrt{A_{m}B_{m}}}+\frac{2(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{A_{m}^{2}-B_{m}^{2}}R_{1}(\beta^{B};\Upsilon_{\tau_{m}}^{B}|k^{B})-\mathcal{I}_{1_{i}}^{B} (59)
I2B(τm)=(Bmrm4Amrm1BmAm)2XB(τm)AmBm\displaystyle I_{2}^{B}(\tau_{m})=\left(\frac{B_{m}r_{m4}-A_{m}r_{m1}}{B_{m}-A_{m}}\right)^{2}\frac{X^{B}(\tau_{m})}{\sqrt{A_{m}B_{m}}}
+4(Amrm1Bmrm4AmBm)(rm4rm1)AmBmAm2Bm2R1(βB;ΥτmB|kB)\displaystyle\quad\quad\quad\quad\quad\quad\left.+4\left(\frac{A_{m}r_{m1}-B_{m}r_{m4}}{A_{m}-B_{m}}\right)\frac{(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{A_{m}^{2}-B_{m}^{2}}R_{1}(\beta^{B};\Upsilon_{\tau_{m}}^{B}|k^{B})\right.
+AmBm(2(rm4rm1)AmBmAm2Bm2)2R2(βB;ΥτmB|kB)2iB\displaystyle\quad\quad\quad\quad\quad\quad\quad+\sqrt{A_{m}B_{m}}\left(\frac{2(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{A_{m}^{2}-B_{m}^{2}}\right)^{2}R_{2}(\beta^{B};\Upsilon_{\tau_{m}}^{B}|k^{B})-\mathcal{I}_{2_{i}}^{B} (60)

In the formulas above, the parameters of the functions R1R_{1} and R2R_{2} are related with the roots of Rm(r)R_{m}(r) as follows

β±B=Bm(rm4r±)+Am(r±rm1)Bm(rm4r±)Am(r±rm1),βB=AmBmAm+Bm\displaystyle\hskip 22.76219pt\beta_{\pm}^{B}=-\frac{B_{m}(r_{m4}-r_{\pm})+A_{m}(r_{\pm}-r_{m1})}{B_{m}(r_{m4}-r_{\pm})-A_{m}(r_{\pm}-r_{m1})}\;,\hskip 11.38109pt\beta^{B}=\frac{A_{m}-B_{m}}{A_{m}+B_{m}}\, (61)
ΥrB=cos1(Bm(rm4r)Am(rrm1)Bm(rm4r)+Am(rrm1)),ΥτmB=am(XB(τm)|kB)\displaystyle\Upsilon_{r}^{B}=\cos^{-1}\left(\frac{B_{m}(r_{m4}-r)-A_{m}(r-r_{m1})}{B_{m}(r_{m4}-r)+A_{m}(r-r_{m1})}\right),\hskip 11.38109pt\Upsilon_{\tau_{m}}^{B}={\rm am}\left(X_{B}(\tau_{m})\left|k_{B}\right)\right. (62)

where am\rm am is the Jacobi amplitude function. The quantities ±iB\mathcal{I}_{\pm_{i}}^{B}, 1iB\mathcal{I}_{1_{i}}^{B}, and 2iB\mathcal{I}_{2_{i}}^{B} are obtained by evaluating I±B{I}_{\pm}^{B}, I1B{I}_{1}^{B}, and I2B{I}_{2}^{B} at r=rir=r_{i} of the initial condition, I±B(0)=I1B(0)=I2B(0)=0{{I}_{\pm}^{B}(0)={I}_{1}^{B}(0)={I}_{2}^{B}(0)=0}. Finally, R1R_{1} and R2R_{2} are the integral of Jacobian elliptic cosine function,

R1(α;ϕ|k)0F(ϕ|k)du1+αcn(u|k)=11α2[Π(α2α21;ϕ|k)αf(pα,ϕ,k)]R_{1}(\alpha;\phi|k)\equiv\int_{0}^{F(\phi|k)}\frac{du}{1+\alpha{\rm cn}(u|k)}=\frac{1}{1-\alpha^{2}}\left[\Pi\Bigg{(}\frac{\alpha^{2}}{\alpha^{2}-1};\phi\left|k\Bigg{)}\right.-\alpha f(p_{\alpha},\phi,k)\right] (63)
R2(α;ϕ|k)0F(ϕ|k)du[1+αcn(u|k)]2\displaystyle R_{2}(\alpha;\phi|k)\equiv\int_{0}^{F(\phi|k)}\frac{du}{[1+\alpha{\rm cn}(u|k)]^{2}}
=1α21[F(ϕ|k)α2k+(1k)α2(E(ϕ|k)αsin(ϕ)1ksin2(ϕ)1+αcos(ϕ))]\displaystyle\quad\quad=\frac{1}{\alpha^{2}-1}\left[F\left(\phi|k\right)-\frac{\alpha^{2}}{k+(1-k)\alpha^{2}}\left(E(\phi|k)-\frac{\alpha\sin(\phi)\sqrt{1-k\sin^{2}(\phi)}}{1+\alpha\cos(\phi)}\right)\right]
+1k+(1k)α2(2kα2α21)R1(α;ϕ|k)\displaystyle\quad\quad\quad+\frac{1}{k+(1-k)\alpha^{2}}\left(2k-\frac{\alpha^{2}}{\alpha^{2}-1}\right)R_{1}(\alpha;\phi|k) (64)

in which

f(pα,ϕ,k)=pα2ln(pα1ksin2(ϕ)+sin(ϕ)pα1ksin2(ϕ)sin(ϕ)),pα=α21k+(1k)α2.\displaystyle f(p_{\alpha},\phi,k)=\frac{p_{\alpha}}{2}\ln\left(\frac{p_{\alpha}\sqrt{1-k\sin^{2}(\phi)}+\sin(\phi)}{p_{\alpha}\sqrt{1-k\sin^{2}(\phi)}-\sin(\phi)}\right)\,,\quad p_{\alpha}=\sqrt{\frac{\alpha^{2}-1}{k+(1-k)\alpha^{2}}}\,. (65)

In particular, for α=βB,β±B\alpha=\beta^{B},\;\beta^{B}_{\pm} where 1<α<1-1<\alpha<1, the solutions are the real-valued functions.

We then apply the exact solution obtained above to the parameters set C of Fig. 5. In this case, λm=1\lambda_{m}=1, ηm=7\eta_{m}=7, and γm=0.98\gamma_{m}=0.98, with the black hole parameters a=0.7a=0.7 and Q=0.7Q=0.7. Fig. 6 shows that the particle stars from the initial position ri=7.4Mr_{i}=7.4M, θi=π/2\theta_{i}=\pi/2, and ϕi=0\phi_{i}=0 and it falls almost directly into the black hole.

Refer to caption
Figure 6: Illustration of an orbit off the equatorial plane with the parameters of C in Fig. 5. In this case the particle starts from ri<r4r_{i}<r_{4} and plunges directly into the black hole horizon.

From the above general formulas one obtains the case of the equatorial motion, in which θ=π2\theta=\frac{\pi}{2} and ηm0\eta_{m}\rightarrow 0. The bound plunge solution of the coordinates ϕmB\phi_{m}^{B} and tmBt_{m}^{B} can be rewritten as the function of rr as follows

ϕmB(r)=ImϕB(τm(r))+λmτmB(r)+ϕmiB\displaystyle\phi_{m}^{B}\left(r\right)=I^{B}_{m\phi}\left(\tau_{m}\left(r\right)\right)+\lambda_{m}\tau^{B}_{m}\left(r\right)+\phi^{B}_{mi}
=γm1γm2[2Mar+r(𝒥m+𝒥m)λmγmf(r)],\displaystyle=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\left[\frac{2Ma}{r_{+}-r_{-}}\left(\mathcal{J}_{m+}-\mathcal{J}_{m-}\right)-\frac{\lambda_{m}}{\gamma_{m}}f\left(r\right)\right]\;, (66)
tmB(r)=ImtB(τm)+tmiB\displaystyle t_{m}^{B}\left(r\right)=I^{B}_{mt}\left(\tau_{m}\right)+t^{B}_{mi}
=γm1γm2{4M2r+r(𝒯m+𝒯m)+Bmrm4Amrm1BmAm(Bmrm4Amrm1BmAm+M)f(r)\displaystyle=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\left\{\frac{4M^{2}}{r_{+}-r_{-}}\left(\mathcal{T}_{m+}-\mathcal{T}_{m-}\right)+\frac{B_{m}r_{m4}-A_{m}r_{m1}}{B_{m}-A_{m}}\left(\frac{B_{m}r_{m4}-A_{m}r_{m1}}{B_{m}-A_{m}}+M\right)f\left(r\right)\right.
+2(rm4rm1)AmBmAm2Bm2[2(Bmrm4Amrm1BmAm)+M]R1(βB;φ(r)|kB)\displaystyle+\frac{2(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{A_{m}^{2}-B_{m}^{2}}\left[2\left(\frac{B_{m}r_{m4}-A_{m}r_{m1}}{B_{m}-A_{m}}\right)+M\right]R_{1}\left(\beta^{B};\varphi(r)|k^{B}\right)
+4AmBm[(rm4rm1)AmBmAm2Bm2]2R2(βB;φ(r)|kB)+(4M2Q2)f(r)},\displaystyle\left.+4\sqrt{A_{m}B_{m}}\left[\frac{(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{A_{m}^{2}-B_{m}^{2}}\right]^{2}R_{2}\left(\beta^{B};\varphi(r)|k^{B}\right)+\left(4M^{2}-Q^{2}\right)f\left(r\right)\right\}\;, (67)

where

𝒯m±=(r±Q22M)𝒥m±,\displaystyle\mathcal{T}_{m\pm}=\left(r_{\pm}-\frac{Q^{2}}{2M}\right)\mathcal{J}_{m\pm}\,, (68)
𝒥m±=(r±a(λmγm)+Q22M){(BmAm)f(r)Bm(rm4r±)+Am(r+r±)\displaystyle\mathcal{J}_{m\pm}=\left(r_{\pm}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)\left\{\frac{(B_{m}-A_{m})f\left(r\right)}{B_{m}(r_{m4}-r_{\pm})+A_{m}(r_{+}-r_{\pm})}\right.
+2(rm4rm1)AmBm[Bm(rm4r±)]2[Am(r±rm1)]2R1(β±B;φ(r)|kB)},\displaystyle\quad\quad\left.+\frac{2(r_{m4}-r_{m1})\sqrt{A_{m}B_{m}}}{\left[B_{m}(r_{m4}-r_{\pm})\right]^{2}-\left[A_{m}(r_{\pm}-r_{m1})\right]^{2}}R_{1}\left(\beta_{\pm}^{B};\varphi(r)|k^{B}\right)\right\}\;, (69)
f(r)=1AmBmF(φ(r)|kB).\displaystyle f\left(r\right)=\frac{1}{\sqrt{A_{m}B_{m}}}F\left(\varphi(r)|k^{B}\right)\,. (70)

Fig. 7 shows an exemplary orbit of this type using the these solutions.

Refer to caption
Figure 7: Illustration of an orbit on the equatorial plane with the parameters of D in Fig. 5. The particle initiates its journey at point rir_{i}, moves outward, reaches the turning point at rm4r_{m4} , and then reverses its course, plunging back into the black hole.

The expression can also be converted into the solutions of the Kerr and RN black holes by taking the respective a0a\rightarrow 0 and Q0Q\rightarrow 0 limit. For the Kerr black hole, one can substitute straightforwardly Q=0Q=0 and the root rm1=0r_{m1}=0 into the definition of kBk^{B} and BmB_{m} in (54) and (55), as well as the solutions (66) and (67). Nevertheless, in the RN black hole the limits of a0a\rightarrow 0 but rm10r_{m1}\neq 0 give huge simplification. The formula (66) becomes

ϕmB(r)=λm(1γm2)AmBmF(cos1(Bm(rm4r)Am(rrm1)Bm(rm4r)+Am(rrm1))|kB),\displaystyle\phi_{m}^{B}\left(r\right)=-\frac{\lambda_{m}}{\sqrt{(1-\gamma_{m}^{2})A_{m}B_{m}}}F\Bigg{(}\cos^{-1}\left(\frac{B_{m}(r_{m4}-r)-A_{m}(r-r_{m1})}{B_{m}(r_{m4}-r)+A_{m}(r-r_{m1})}\right)\left|k^{B}\Bigg{)}\right.\,, (71)

whereas the solution of tmBt_{m}^{B} remains the same form as in (67) in the corresponding limits. In the Schwarschild black hole where a,Q0a,Q\rightarrow 0, two event horizons, r+=2Mr_{+}=2M, r=0r_{-}=0 giving 𝒯m0\mathcal{T}_{m-}\rightarrow 0, together with rm1=0r_{m1}=0, lead to the further simplification from (71) and (67).

One usefulness of the analytical formulas is to explore the changes of azimuthal angle ϕ\phi as the motion crosses the horizon. In particular for the equatorial motion, as rr+r\rightarrow r_{+}, the term of 𝒥m+\mathcal{J}_{m+} in (66) due to the function of R1(α;ϕ|k)R_{1}(\alpha;\phi|k) in (63) gives the logarithmic divergence, namely ϕmBln(rr+)\phi_{m}^{B}\propto\ln(r-r_{+}). However, in the extremal case when r+=r=Mr_{+}=r_{-}=M, the extra divergence occurs in the coefficient of 𝒥m+𝒥m+\mathcal{J}_{m+}-\mathcal{J}_{m+} so that the leading order divergence becomes ϕmBln(rM)\phi_{m}^{B}\propto\ln(r-M). In the case of a=0a=0 for non spinning black hole, the coefficient of 𝒥m+𝒥m+\mathcal{J}_{m+}-\mathcal{J}_{m+} vanishes and the angle ϕ\phi smoothly changes across the horizon. The corresponding plot is shown in Fig. 8.

Refer to caption
Figure 8: The plots show the ΔϕmB\Delta\phi_{m}^{B} as a function of rr for bound motion with the various black holes including Kerr-Newman (Q=0.7M,a=0.7M,λm=M,ηm=0Q=0.7M,a=0.7M,\lambda_{m}=M,\eta_{m}=0), Kerr-extremal (Q=0M,a=M,λm=M,ηm=0Q=0M,a=M,\lambda_{m}=-M,\eta_{m}=0) and RN (Q=0.7M,a=0M,λm=M,ηm=0Q=0.7M,a=0M,\lambda_{m}=M,\eta_{m}=0). In this plot, the red and blue curves illustrate the examples of direct and retrograde orbits, respectively.

V Plunging ORBITS IN UNBOUND MOTION

For unbound motion (γm>1\gamma_{m}>1), the particle may start from the spatial infinity characterized by the constants of motion with the azimuthal angular momentum λm\lambda_{m}, the energy γm\gamma_{m}, and the Carter constant ηm\eta_{m}. In this section we consider the parameters mainly in the E regime shown in Fig. 9, in which the roots of the radial potential have the properties, rm3=rm4r_{m3}^{*}={r}_{m4} and ri>r+>r>rm2>rm1r_{i}>r_{+}>r_{-}>r_{m2}>r_{m1}. This means that there is no turning point in the black hole exterior and the particle starting from the spatial infinity will plunge directly into the black hole.

Refer to caption
Figure 9: The graphics shows the portion of parameter space limited by the double root solution, rm3=rm4r_{m3}=r_{m4} and rm1<rm2<r<r+r_{m1}<r_{m2}<r_{-}<r_{+}. For the region of the parameter space for E and F, the roots rm3r_{m3} and rm4r_{m4} are complex, rm3=rm4r_{m3}=r_{m4}^{*} and rm1<rm2<rr_{m1}<r_{m2}<r_{-}. The inset shows the details of the roots of illustrative cases E and F in the main figure . See the text for more discussion.

The main propose here is also to derive the exact solutions for the coordinates rmU(τm)r_{m}^{U}(\tau_{m}), θmU(τm)\theta_{m}^{U}(\tau_{m}), ϕmU(τm)\phi_{m}^{U}(\tau_{m}), and tmU(τm)t^{U}_{m}(\tau_{m}) (We have added the upper index UU for the unbound case). While the procedure is identical to that of the previous two sections, special care is required due to differences in the properties of the roots. The counterpart of Eq. (52) becomes

τmU=1(γm21)AmUBmU[F(ψ(r)|kU)F(ψ(ri)|kU)],\displaystyle\tau^{U}_{m}=-\frac{1}{\sqrt{(\gamma_{m}^{2}-1)A_{m}^{U}B_{m}^{U}}}\left[F\left(\psi(r)|k^{U}\right)-F\left(\psi(r_{i})|k^{U}\right)\right]\,, (72)

where

ψ(r)=cos1(AmU(rrm1)BmU(rrm2)AmU(rrm1)+BmU(rrm2)),\displaystyle\psi(r)=\cos^{-1}\left(\frac{A_{m}^{U}(r-r_{m1})-B_{m}^{U}(r-r_{m2})}{A_{m}^{U}(r-r_{m1})+B_{m}^{U}(r-r_{m2})}\right)\,, (73)
kU=(AmU+BmU)2(rm2rm1)24AmUBmU,\displaystyle k^{U}=\frac{(A_{m}^{U}+B_{m}^{U})^{2}-(r_{m2}-r_{m1})^{2}}{4A_{m}^{U}B_{m}^{U}}\,, (74)

and

AmU=(rm3rm2)(rm4rm2),BmU=(rm3rm1)(rm4rm1).\displaystyle A_{m}^{U}=\sqrt{(r_{m3}-r_{m2})(r_{m4}-r_{m2})}\;\;\;,\;\;\;B_{m}^{U}=\sqrt{(r_{m3}-r_{m1})(r_{m4}-r_{m1})}\,. (75)

Notice that AmUA^{U}_{m} and BmUB^{U}_{m} have different combinations of roots compared to the bound case (55). The evolution of the coordinate rU(τm)r^{U}(\tau_{m}) is then

rU(τm)=(BmUrm2AmUrm1)+(BmUrm2+AmUrm1)cn(XU(τm)|kU)(BmUAmU)+(BmU+AmU)cn(XU(τm)|kU),\displaystyle r^{U}(\tau_{m})=\frac{(B_{m}^{U}r_{m2}-A_{m}^{U}r_{m1})+(B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}){\rm{cn}}\left(X^{U}(\tau_{m})\left|k^{U}\right)\right.}{(B_{m}^{U}-A_{m}^{U}){+}(B_{m}^{U}+A_{m}^{U}){\rm{cn}}\left(X^{U}(\tau_{m})\left|k^{U}\right)\right.}\,, (76)

where

XU(τm)=(γm21)AmUBmUτmF(cos1(AmU(rirm1)BmU(rirm2)AmU(rirm1)+BmU(rirm2))|kU).\displaystyle X^{U}(\tau_{m})=\sqrt{\left(\gamma_{m}^{2}-1\right)A_{m}^{U}B_{m}^{U}}\tau_{m}-F\Bigg{(}\cos^{-1}\left(\frac{A_{m}^{U}(r_{i}-r_{m1})-B_{m}^{U}(r_{i}-r_{m2})}{A_{m}^{U}(r_{i}-r_{m1})+B_{m}^{U}(r_{i}-r_{m2})}\right)\left|k^{U}\Bigg{)}\right.\,. (77)

Once again, the specified conditions—BmU>AmU>0B_{m}^{U}>A_{m}^{U}>0, 0<kU<10<k^{U}<1, and rm1<rm2<rr_{m1}<r_{m2}<r, together with the inequality 1<AmU(rrm1)BmU(rrm2)AmU(rrm1)+BmU(rrm2)<1-1<\frac{A_{m}^{U}(r-r_{m1})-B_{m}^{U}(r-r_{m2})}{A_{m}^{U}(r-r_{m1})+B_{m}^{U}(r-r_{m2})}<1 ensure that the Jacobian elliptic cosine function in Eq. (76) remains a real-valued function.

The missing pieces for a complete description of the motions are the unbound version of equations (25) and (26), in which the integral (23) and (24) have been solved in Sec. IV. The results can be written as follows

I±U(τm)=1BmU(r±rm2)+AmU(r±rm1)[BmU+AmUAmUBmUXU(τm)\displaystyle I_{\pm}^{U}(\tau_{m})=-\frac{1}{B_{m}^{U}\left(r_{\pm}-r_{m2}\right)+A_{m}^{U}\left(r_{\pm}-r_{m1}\right)}\left[\frac{B_{m}^{U}+A_{m}^{U}}{\sqrt{A^{U}_{m}B^{U}_{m}}}X^{U}(\tau_{m})\right.
+2(rm2rm1)AmUBmUBmU(r±rm2)AmU(r±rm1)R1(β±U;ΥτmU|kU)]±iU,\displaystyle\quad\quad\quad\quad\quad\quad\quad\left.+\frac{2(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{B_{m}^{U}\left(r_{\pm}-r_{m2}\right)-A_{m}^{U}\left(r_{\pm}-r_{m1}\right)}R_{1}(\beta_{\pm}^{U};\Upsilon_{\tau_{m}}^{U}|k^{U})\right]-\mathcal{I}_{\pm_{i}}^{U}\;, (78)
I1U(τm)=(BmUrm2+AmUrm1BmU+AmU)XU(τm)AmUBmU+2(rm2rm1)AmUBmU(BmU)2(AmU)2R1(βU;ΥτmU|kU)1iU,\displaystyle I_{1}^{U}(\tau_{m})=\left(\frac{B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}}{B_{m}^{U}+A_{m}^{U}}\right)\frac{X^{U}(\tau_{m})}{\sqrt{A_{m}^{U}B_{m}^{U}}}+\frac{2(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{(B_{m}^{U})^{2}-(A_{m}^{U})^{2}}R_{1}(\beta^{U};\Upsilon_{\tau_{m}}^{U}|k^{U})-\mathcal{I}_{1_{i}}^{U}\;, (79)
I2U(τm)=(BmUrm2+AmUrm1BmU+AmU)2XU(τm)AmUBmU\displaystyle I_{2}^{U}(\tau_{m})=\left(\frac{B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}}{B_{m}^{U}+A_{m}^{U}}\right)^{2}\frac{X^{U}(\tau_{m})}{\sqrt{A^{U}_{m}B^{U}_{m}}}
+4(BmUrm2+AmUrm1BmU+AmU)(rm2rm1)AmUBmU(BmU)2(AmU)2R1(βU;ΥτmU|kU)\displaystyle\quad\quad\quad\quad\quad\quad\left.+4\left(\frac{B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}}{B_{m}^{U}+A_{m}^{U}}\right)\frac{(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{(B_{m}^{U})^{2}-(A_{m}^{U})^{2}}R_{1}(\beta^{U};\Upsilon_{\tau_{m}}^{U}|k^{U})\right.
+AmUBmU(2(rm2rm1)AmUBmU(BmU)2(AmU)2)2R2(βU;ΥτmU|kU)2iU,\displaystyle\quad\quad\quad\quad\quad\quad\quad\quad+\sqrt{A_{m}^{U}B_{m}^{U}}\left(\frac{2(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{(B_{m}^{U})^{2}-(A_{m}^{U})^{2}}\right)^{2}R_{2}(\beta^{U};\Upsilon_{\tau_{m}}^{U}|k^{U})-\mathcal{I}_{2_{i}}^{U}\;, (80)

where the functions R1R_{1} and R2R_{2} have been defined in (63) and (64) and the unbound version of the parameters now read as

β±U=BmU(r±rm2)+AmU(r±rm1)BmU(r±rm2)AmU(r±rm1),βU=BmU+AmUBmUAmU\displaystyle\hskip 22.76219pt\beta_{\pm}^{U}=\frac{B_{m}^{U}(r_{\pm}-r_{m2})+A_{m}^{U}(r_{\pm}-r_{m1})}{B_{m}^{U}(r_{\pm}-r_{m2})-A_{m}^{U}(r_{\pm}-r_{m1})},\hskip 11.38109pt\beta^{U}=\frac{B_{m}^{U}+A_{m}^{U}}{B_{m}^{U}-A_{m}^{U}}\, (81)
ΥrU=cos1(AmU(rrm1)BmU(rrm2)AmU(rrm1)+BmU(rrm2)),ΥτmU=am(XU(τm)|kU).\displaystyle\Upsilon_{r}^{U}=\cos^{-1}\left(\frac{A_{m}^{U}(r-r_{m1})-B_{m}^{U}(r-r_{m2})}{A_{m}^{U}(r-r_{m1})+B_{m}^{U}(r-r_{m2})}\right),\hskip 11.38109pt\Upsilon_{\tau_{m}}^{U}={\rm am}\left(X^{U}(\tau_{m})\left|k^{U}\right)\right.\,. (82)

As before, the initial conditions ±iU\mathcal{I}_{\pm_{i}}^{U}, 1iU\mathcal{I}_{1_{i}}^{U}, 2iU\mathcal{I}_{2_{i}}^{U} are obtained by evaluating I±U{I}_{\pm}^{U}, I1U{I}_{1}^{U}, I2U{I}_{2}^{U} at r=rir=r_{i} of the initial condition. Also, for α\alpha in the functions (63) R1R_{1} and R2R_{2}, we have now α=βU,β±U\alpha=\beta^{U},\beta^{U}_{\pm}, where 0<α<10<\alpha<1, which ensure that the solutions are real-valued functions. Fig. 10 illustrates the orbits with the parameters of E in Fig. 9, where λm=0\lambda_{m}=0, ηm=10\eta_{m}=10, and γm=1.25\gamma_{m}=1.25.

Refer to caption
Figure 10: Illustration of a general nonequatorial plunging orbit exemplifies with the parameters E in Fig. 9. In this case ηm0\eta_{m}\neq 0 and the particle starts from spatial infinity and infalls directly into the black horizon.
Refer to caption
Figure 11: Illustration of an equatorial inspiral orbit with the parameters in the F region in Fig. 9 for ηm=0\eta_{m}=0 where the particle starts from spatial infinity and plunge directly into the horizon.

Following the same steps of the previous sections, one can straightforwardly calculate the trajectories on the equatorial plane. The solutions of ϕmU\phi_{m}^{U} and tmUt_{m}^{U} as the function of rr can be written as follows

ϕmU(r)=ImϕU(τm(r))+λmτm(r)+ϕmiU\displaystyle\phi_{m}^{U}\left(r\right)=I^{U}_{m\phi}\left(\tau_{m}\left(r\right)\right)+\lambda_{m}\tau_{m}\left(r\right)+\phi^{U}_{mi}
=γmγm21[2Mar+r(𝒦m𝒦m+)λmγmg(r)],\displaystyle=\frac{\gamma_{m}}{\sqrt{\gamma_{m}^{2}-1}}\left[\frac{2Ma}{r_{+}-r_{-}}\left(\mathcal{K}_{m-}-\mathcal{K}_{m+}\right)-\frac{\lambda_{m}}{\gamma_{m}}g\left(r\right)\right]\,, (83)
tmU(r)=ImtU(τm)+tmiU\displaystyle t_{m}^{U}\left(r\right)=I^{U}_{mt}\left(\tau_{m}\right)+t^{U}_{mi}
=γm1γm2{4M2r+r(𝒱m𝒱m+)+BmUrm2+AmUrm1BmU+AmU(BmUrm2+AmUrm1BmU+AmU+M)g(r)\displaystyle=\frac{\gamma_{m}}{\sqrt{1-\gamma_{m}^{2}}}\left\{\frac{4M^{2}}{r_{+}-r_{-}}\left(\mathcal{V}_{m-}-\mathcal{V}_{m+}\right)+\frac{B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}}{B_{m}^{U}+A_{m}^{U}}\left(\frac{B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}}{B_{m}^{U}+A_{m}^{U}}+M\right)g\left(r\right)\right.
+2(rm2rm1)AmUBmU(BmU)2(AmU)2[2(BmUrm2+AmUrm1BmU+AmU)+M]R1(βU;ψ(r)|kU)\displaystyle+\frac{2(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{(B_{m}^{U})^{2}-(A_{m}^{U})^{2}}\left[2\left(\frac{B_{m}^{U}r_{m2}+A_{m}^{U}r_{m1}}{B^{U}_{m}+A^{U}_{m}}\right)+M\right]R_{1}\left(\beta^{U};\psi(r)|k^{U}\right)
+4AmUBmU[(rm2rm1)AmUBmU(BmU)2(AmU)2]2R2(βU;ψ(r)|kU)+(4M2Q2)g(r)},\displaystyle\left.+4\sqrt{A_{m}^{U}B_{m}^{U}}\left[\frac{(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{(B_{m}^{U})^{2}-(A_{m}^{U})^{2}}\right]^{2}R_{2}\left(\beta^{U};\psi(r)|k^{U}\right)+\left(4M^{2}-Q^{2}\right)g\left(r\right)\right\}\,\,, (84)

where

𝒱m±=(r±Q22M)𝒦m±,\displaystyle\mathcal{V}_{m\pm}=\left(r_{\pm}-\frac{Q^{2}}{2M}\right)\mathcal{K}_{m\pm}\,, (85)
𝒦m±=(r±a(λmγm)+Q22M){(BmU+AmU)g(r)BmU(r±rm2)+AmU(r±rm1)\displaystyle\mathcal{K}_{m\pm}=\left(r_{\pm}-\frac{a\left(\frac{\lambda_{m}}{\gamma_{m}}\right)+Q^{2}}{2M}\right)\left\{\frac{(B_{m}^{U}+A_{m}^{U})g\left(r\right)}{B_{m}^{U}(r_{\pm}-r_{m2})+A_{m}^{U}(r_{\pm}-r_{m1})}\right.
+2(rm2rm1)AmUBmU[BmU(r±rm2)]2[AmU(r±rm1)]2R1(β±U;ψ(r)|kU)},\displaystyle\quad\quad\left.+\frac{2(r_{m2}-r_{m1})\sqrt{A_{m}^{U}B_{m}^{U}}}{\left[B_{m}^{U}(r_{\pm}-r_{m2})\right]^{2}-\left[A_{m}^{U}(r_{\pm}-r_{m1})\right]^{2}}R_{1}\left(\beta_{\pm}^{U};\psi(r)|k^{U}\right)\right\}\,, (86)
g(r)=1AmUBmUF(ψ(r)|kU).\displaystyle g\left(r\right)=\frac{1}{\sqrt{A_{m}^{U}B_{m}^{U}}}F\left(\psi(r)|k^{U}\right)\,. (87)

Fig. 11 shows the orbit of the particle with the parameters in F in Fig. 9.

Refer to caption
Figure 12: The plots show the ΔϕmU\Delta\phi_{m}^{U} as a function of rr for unbond infalling motion, with the various black hole including Kerr-Newman (Q=0.7M,a=0.7M,λm=M,ηm=0Q=0.7M,a=0.7M,\lambda_{m}=M,\eta_{m}=0), Kerr extremal (Q=0M,a=M,λm=M,ηm=0Q=0M,a=M,\lambda_{m}=-M,\eta_{m}=0) and RN (Q=0.7M,a=0M,λm=M,ηm=0Q=0.7M,a=0M,\lambda_{m}=M,\eta_{m}=0). The blue and red curves show the examples of direct and retrograde orbits, respectively.

In the Kerr black hole, for Q0Q\rightarrow 0, the solutions are given by taking rm1=0r_{m1}=0 to (83) and (84). In the RN black hole for a0a\rightarrow 0 but rm10r_{m1}\neq 0, Eq. (83) can be significantly simplified as

ϕmU(r)=λm(γm21)AmUBmUF(cos1(AmU(rrm1)BmU(rrm2)AmU(rrm1)+BmU(rrm2))|kU).\displaystyle\phi_{m}^{U}\left(r\right)=-\frac{\lambda_{m}}{\sqrt{(\gamma_{m}^{2}-1)A_{m}^{U}B_{m}^{U}}}F\Bigg{(}\cos^{-1}\left(\frac{A_{m}^{U}(r-r_{m1})-B_{m}^{U}(r-r_{m2})}{A_{m}^{U}(r-r_{m1})+B_{m}^{U}(r-r_{m2})}\right)\left|k^{U}\Bigg{)}\right.\,. (88)

In the Schwarschild black hole, a,Q0a,Q\rightarrow 0 and the root rm1=0r_{m1}=0 further simplify the RN solution above. Nevertheless, the solution of tmUt_{m}^{U} in various black holes remains the same form as (84) after taking the proper limits.

The behavior of the azimuthal angle ϕ\phi as the motion crosses the horizon shares the same features as in bound motion. In particular for the equatorial motion, as crossing the horizon, the term of 𝒦m±\mathcal{K}_{m_{\pm}} in (83) due to the function of R1(α;ϕ|k)R_{1}(\alpha;\phi|k) in (63) gives the logarithmic divergence in general Kerr-Newman black holes. However, in the extremal case, the leading order divergence becomes ϕmBln(rM)\phi_{m}^{B}\propto\ln(r-M). In the case of a=0a=0 for non spinning black hole, again the coefficient of 𝒦m+𝒦m+\mathcal{K}_{m+}-\mathcal{K}_{m+} vanishes and the angle ϕ\phi smoothly changes across the horizon. The corresponding plot is shown in Fig. 12.

VI Conclusions

In this paper, we analytically derive the solutions of infalling orbits in the context of general nonequatorial motion in the Kerr-Newman black holes, considering both bound and unbound motion. These solutions can be written in terms of the elliptical integrals and the Jacobian elliptic functions of manifestly real functions of the Mino time. Various limits have been taken to show the respective solutions in Kerr, Reissner-Nordstro¨\ddot{o}m, and Schwarzschild black holes. In the case of the bound motion, we extend the study of [13, 14] on equatorial motion to consider that the particle starts from rrISSOr\leq r_{\rm ISSO} and then inspirals into the black hole on general nonequatorial motion. In the limits of Q0Q\rightarrow 0 and restricting on the equatorial plane, the obtained solutions reduces to the ones in [13, 14] . We also consider the other types of the plunge motion, with the values of γm\gamma_{m}, λm\lambda_{m}, and ηm\eta_{m} shown in Fig.5. In these cases, one has two real-valued roots of the radial potential, one inside the horizon, rm1r_{m1}, and the other outside the horizon, rm4r_{m4}. Thus, the particle starts from rrm4r\leq r_{m4} and can either travel toward the rm4r_{m4}, come back and then plunge directly into the black hole or travel directly into the black hole. As for the unbound state, we showed the parameters γm\gamma_{m}, λm\lambda_{m}, and ηm\eta_{m} in Fig.9. Interestingly, while the two real-valued roots rm2r_{m2} and rm1r_{m1} are inside the event horizon, the other two roots are complex conjugate pair, rm3=rm4r_{m3}=r_{m4}^{*}. The particle starts from the spatial infinity and will plunge directly into the black holes. The analytical solutions allow us, in particular, to explore the behavior of the variation of azimuthal angle ϕ\phi as the equatorial motion crosses the horizon. In general Kerr-Newman black holes, the angle ϕ\phi diverges as ϕmln(rr+)\phi_{m}\propto\ln(r-r_{+}). However, in the extremal case, the leading order divergence becomes ϕm1/(rM)\phi_{m}\propto 1/(r-M). In the case of non spinning black hole, the angle smoothly changes across the horizon. The dramatic difference in the behavior of the azimuthal angle ϕ\phi across the horizon, which has been found in equatorial infalling orbits in Kerr black holes [13, 14], is also seen in Kerr-Newman black holes in general nonequatorial orbits. This may have some implications for the associated gravitational wave emission observed far away from the black holes.

These exact solutions of the spiral and plunge motions into the black hole are also of astrophysical interest due to the fact that they have direct relevance to black hole accretion phenomena. These explicit solutions may have applications to the numerical accretion as well as extending current theories of black hole accretion [15, 16, 18, 17].

Appendix A The angular potential Θ(θ)\Theta(\theta) and the integrals GmθG_{m\theta}, GmϕG_{m\phi}, and GmtG_{mt}

The detailed studies related to the Θm\Theta_{m} potential in the θ\theta direction can be found in the papers [28, 26, 36]. Here we summarize some of the relevant parts for the completeness of presentation. The angular potential (12) for the particle can be rewritten in terms of u=cos2θu=\cos^{2}\theta and the equation of motion requires Θm0\Theta_{m}\geq 0, which restricts the parameter space of λm\lambda_{m}, ηm\eta_{m}, and γm\gamma_{m} (see Fig. 9 in [26]). The roots of Θm(θ)=0\Theta_{m}(\theta)=0 can be written as [28],

um,±=Δm,θ±νmΔm,θ2+4a2ηmγm212a2Δmθ=a2ηm+λm2γm21\displaystyle{u_{m,\pm}=\frac{\Delta_{m,\theta}\pm\nu_{m}\sqrt{\Delta_{m,\theta}^{2}+\frac{4\,{a}^{2}\,\eta_{m}}{\gamma_{m}^{2}-1}}}{2{a}^{2}}\,\;\;\Delta_{m\theta}={a}^{2}-\frac{\eta_{m}+\lambda_{m}^{2}}{\gamma_{m}^{2}-1}} (89)

with νm=sign(γm21)\nu_{m}={\rm sign}(\gamma_{m}^{2}-1), which give the boundaries of the parameter space. For positive ηm\eta_{m} and nonzero λm\lambda_{m} the particle starts off from the black hole exterior, 1>u+>0{1>u_{+}}>0 is the only positive root, which in turn gives two roots at θm+=cos1(u+),θm=cos1(u+)\theta_{m+}=\cos^{-1}\left(-\sqrt{u_{+}}\right),\theta_{m-}=\cos^{-1}\left(\sqrt{u_{+}}\right). The particle travels between the southern and northern hemispheres crossing the equator at θ=π2\theta=\frac{\pi}{2}.

The solution of the coordinate θm(τm)\theta_{m}(\tau_{m}) can be obtained by an inversion of (14) [28, 26]

θ(τm)=cos1(νθium+sn(uma2(γm21)(τm+νθi𝒢mθi)|um+um)),\theta(\tau_{m})=\cos^{-1}\left(-\nu_{\theta_{i}}\sqrt{u_{m+}}{\rm sn}\left(\sqrt{-u_{m-}{a}^{2}\left(\gamma_{m}^{2}-1\right)}\left(\tau_{m}+\nu_{\theta_{i}}\mathcal{G}_{m\theta_{i}}\right)\left|\frac{u_{m+}}{u_{m-}}\right)\right.\right)\;, (90)

where Mino time

τm=Gmθ=p(𝒢mθ+𝒢mθ)+νθi[(1)p𝒢mθ𝒢mθi]\displaystyle\tau_{m}=G_{m\theta}=p(\mathcal{G}_{m\theta_{+}}-\mathcal{G}_{m\theta_{-}})+\nu_{\theta_{i}}\left[(-1)^{p}\mathcal{G}_{m\theta}-\mathcal{G}_{m\theta_{i}}\right] (91)

and sn denotes the Jacobi Elliptical sine function. In (91) pp counts the times the trajectory passes through the turning points and νθi=sign(dθidτ)\nu_{\theta_{i}}={\rm sign}\left(\frac{d\theta_{i}}{d\tau^{\prime}}\right). The function 𝒢mθ\mathcal{G}_{m\theta} is

𝒢mθ=1uma^2(γm21)F(sin1(cosθum+)|um+um).\mathcal{G}_{m\theta}=-\frac{1}{\sqrt{-u_{m-}\hat{a}^{2}\left(\gamma_{m}^{2}-1\right)}}F\left(\sin^{-1}\left(\frac{\cos\theta}{\sqrt{u_{m+}}}\right)\left|\frac{u_{m+}}{u_{m-}}\right)\right.\,. (92)

The evolution of coordinates ϕm(τm)\phi_{m}(\tau_{m}) and tm(τm)t_{m}(\tau_{m}) in (15) and (16) involves the integrals (21) and (22), which can expressed as follows [26]

Gmϕ(τm)=1uma2(γm21)Π(um+;am(uma2(γm21)(τm+νθi𝒢θi)|um+um)|um+um)νθi𝒢mϕi,G_{m\phi}(\tau_{m})=\frac{1}{\sqrt{-u_{m-}{a}^{2}\left(\gamma_{m}^{2}-1\right)}}\Pi\left(u_{m+};{\rm am}\left(\sqrt{-u_{m-}{a}^{2}\left(\gamma_{m}^{2}-1\right)}\left(\tau_{m}+\nu_{\theta_{i}}\mathcal{G}_{\theta_{i}}\right)\left|\frac{u_{m+}}{u_{m-}}\right)\right.\left|\frac{u_{m+}}{u_{m-}}\right)\right.-\nu_{\theta_{i}}\mathcal{G}_{m\phi_{i}}\,, (93)
𝒢ϕi=1uma2(γm21)Π(um+;sin1(cosθium+)|um+um),\mathcal{G}_{\phi_{i}}=-\frac{1}{\sqrt{-u_{m-}{a}^{2}\left(\gamma_{m}^{2}-1\right)}}\Pi\left(u_{m+};\sin^{-1}\left(\frac{\cos\theta_{i}}{\sqrt{u_{m+}}}\right)\left|\frac{u_{m+}}{u_{m-}}\right)\right.\;, (94)
Gmt(τm)=2um+uma2(γm21)E(am(uma2(γm21)(τm+νθi𝒢mθi)|um+um)|um+um)νθi𝒢mti,G_{mt}(\tau_{m})=-\frac{2u_{m+}}{\sqrt{-u_{m-}{a}^{2}\left(\gamma_{m}^{2}-1\right)}}E^{\prime}\left({\rm am}\left(\sqrt{-u_{m-}{a}^{2}\left(\gamma_{m}^{2}-1\right)}\left(\tau_{m}+\nu_{\theta_{i}}\mathcal{G}_{m\theta_{i}}\right)\left|\frac{u_{m+}}{u_{m-}}\right)\right.\left|\frac{u_{m+}}{u_{m-}}\right)\right.-\nu_{\theta_{i}}\mathcal{G}_{mt_{i}}\,, (95)
𝒢mti=2u+ua^2(γm21)E(sin1(cosθiu+)|u+u).\mathcal{G}_{mt_{i}}=\frac{2u_{+}}{\sqrt{-u_{-}\hat{a}^{2}\left(\gamma_{m}^{2}-1\right)}}E^{\prime}\left(\sin^{-1}\left(\frac{\cos\theta_{i}}{\sqrt{u_{+}}}\right)\left|\frac{u_{+}}{u_{-}}\right)\right.\;. (96)

In the formulas above EE and Π\Pi are the incomplete elliptic integral of the second and third kinds, respectively. Also the prime denotes the derivative with respect to the second argument,

E(φ|k)=kE(φ|k)=E(φ|k)F(φ|k)2k.\displaystyle E^{\prime}\left(\varphi\left|k\right)\right.=\partial_{k}E\left(\varphi\left|k\right)\right.=\frac{E\left(\varphi\left|k\right)\right.-F\left(\varphi\left|k\right)\right.}{2k}\,. (97)

Appendix B The radial potential Rm(r)R_{m}(r) and its roots

As for the radial potential (11), it is a quartic polynomial. We then rewrite Rm(r)R_{m}({r}) as follows

Rm(r)=Smr4+Tmr3+Umr2+Vmr+Wm,\displaystyle R_{m}({r})=S_{m}{r}^{4}+T_{m}{r}^{3}+U_{m}{r}^{2}+V_{m}{r}+W_{m}\,, (98)

where the coefficients functions are given in terms constants of motion and parameters of the black hole as

Sm=γm21,S_{m}=\gamma_{m}^{2}-1, (99)
Tm=2M,T_{m}=2M, (100)
Um=a2(γm21)Q2ηmλm2,U_{m}={a}^{2}\left(\gamma_{m}^{2}-1\right)-{Q}^{2}-\eta_{m}-\lambda_{m}^{2}, (101)
Vm=2M[(aγmλm)2+ηm],V_{m}=2M\Bigl{[}\left({a}\gamma_{m}-\lambda_{m}\right)^{2}+\eta_{m}\Bigr{]}, (102)
Wm=a2ηmQ2[(aγmλm)2+ηm].W_{m}=-{a}^{2}\eta_{m}-{Q}^{2}\Bigl{[}\left({a}\gamma_{m}-\lambda_{m}\right)^{2}+\eta_{m}\Bigr{]}\,. (103)

Furthermore, it is useful to represent the radial potential using its roots, namely

Rm(r)=(γm21)(rrm1)(rrm2)(rrm3)(rrm4).R_{m}({r})=\left(\gamma_{m}^{2}-1\right)({r}-r_{m1})({r}-r_{m2})({r}-r_{m3})({r}-r_{m4})\,. (104)

The different dynamical behaviors of the system are characterized by the positions of these roots. See figures (1), (5), (9), and also References [28, 26]. The representation of roots of a quartic equation are well known, but cumbersome. We will write them down for the sake of unifying notation and ensuring the completeness of the work

rm1=M2(γm21)zmXm2zm2+Ym4zm,r_{m1}=-\frac{M}{2\left(\gamma_{m}^{2}-1\right)}-z_{m}-\sqrt{-\hskip 2.84526pt\frac{{X}_{m}}{2}-z_{m}^{2}+\frac{{Y}_{m}}{4z_{m}}}\,, (105)
rm2=M2(γm21)zm+Xm2zm2+Ym4zm,r_{m2}=-\frac{M}{2\left(\gamma_{m}^{2}-1\right)}-z_{m}+\sqrt{-\hskip 2.84526pt\frac{{X}_{m}}{2}-z_{m}^{2}+\frac{{Y}_{m}}{4z_{m}}}\,, (106)
rm3=M2(γm21)+zmXm2zm2Ym4zm,r_{m3}=-\frac{M}{2\left(\gamma_{m}^{2}-1\right)}+z_{m}-\sqrt{-\hskip 2.84526pt\frac{{X}_{m}}{2}-z_{m}^{2}-\frac{{Y}_{m}}{4z_{m}}}\,, (107)
rm4=M2(γm21)+zm+Xm2zm2Ym4zm,r_{m4}=-\frac{M}{2\left(\gamma_{m}^{2}-1\right)}+z_{m}+\sqrt{-\hskip 2.84526pt\frac{{X}_{m}}{2}-z_{m}^{2}-\frac{{Y}_{m}}{4z_{m}}}\,, (108)

where

zm=Ωm++ΩmXm32,z_{m}=\sqrt{\frac{\Omega_{m+}+\Omega_{m-}-\frac{{X}_{m}}{3}}{2}}\;, (109)

and

Ωm±=ϰm2±(ϖm3)3+(ϰm2)23\Omega_{m\pm}=\sqrt[3]{-\hskip 2.84526pt\frac{{\varkappa}_{m}}{2}\pm\sqrt{\left(\frac{{\varpi}_{m}}{3}\right)^{3}+\left(\frac{{\varkappa}_{m}}{2}\right)^{2}}}\, (110)

with

ϖm=Xm212Zm,ϰm=Xm3[(Xm6)2Zm]Ym28.{\varpi}_{m}=-\hskip 2.84526pt\frac{{X}_{m}^{2}}{12}-{Z}_{m}\,,\quad\quad{\varkappa}_{m}=-\hskip 2.84526pt\frac{{X}_{m}}{3}\left[\left(\frac{{X}_{m}}{6}\right)^{2}-{Z}_{m}\right]-\hskip 2.84526pt\frac{{Y}_{m}^{2}}{8}\,. (111)

XmX_{m}, YmY_{m}, and ZmZ_{m} are the short notation for

Xm=8UmSm3Tm28Sm2,\displaystyle{X}_{m}=\frac{8U_{m}S_{m}-3T_{m}^{2}}{8S_{m}^{2}}\,, (112)
Ym=Tm34UmTmSm+8VmSm28Sm3,\displaystyle{Y}_{m}=\frac{T_{m}^{3}-4U_{m}T_{m}S_{m}+8V_{m}S_{m}^{2}}{8S_{m}^{3}}\,, (113)
Zm=3Tm4+256WmSm364VmTmSm2+16UmTm2Sm256Sm4.\displaystyle{Z}_{m}=\frac{-3T_{m}^{4}+256W_{m}S_{m}^{3}-64V_{m}T_{m}S_{m}^{2}+16U_{m}T_{m}^{2}S_{m}}{256S_{m}^{4}}\,. (114)
Acknowledgements.
This work was supported in part by the National Science and Technology council (NSTC) of Taiwan, Republic of China.

References

  • [1] B. P. Abbott et al. (LIGO Scientific and Virgo Collaborations), Observation of Gravitational Waves from a Binary Black Hole Merger, Phys. Rev. Lett. 116, 061102 (2016).
  • [2] B. P. Abbott et al. (LIGO Scientific and Virgo Collaborations), GWTC-1: A Gravitational-Wave Transient Catalog of Compact Binary Mergers Observed by LIGO and Virgo during the First and Second Observing Runs, Phys. Rev. X 9, 031040 (2019).
  • [3] R. Abbott et al. (LIGO Scientific and Virgo Collaborations), GWTC-2: Compact Binary Coalescences Observed by LIGO and Virgo During the First Half of the Third Observing Run, Phys. Rev. X 11, 021053 (2021).
  • [4] K. Akiyama et al. (Event Horizon Telescope), First M87 Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole, Astrophys. J. 875, L1 (2019).
  • [5] K. Akiyama et al. (Event Horizon Telescope), First Sagittarius A* Event Horizon Telescope Results. I. The Shadow of the Supermassive Black Hole in the Center of the Milky Way, Astrophys. J. Lett. 930, L12 (2022).
  • [6] C. W. Misner, K. S. Thorne, and J. A. Wheeler, Gravitation (W. H. Freeman and Company, San Francisco, 1973).
  • [7] S. Chandrasekhar, The Mathematical Theory of Black Holes, International Series of Monographs on Physics (Clarendon Press/Oxford University Press, 1992).
  • [8] eLISA Consortium et al., The gravitational universe, arXiv:1305.5720.
  • [9] Enrico Barausse et al., Prospects for fundamental physics with LISA, Gen. Relativ. Gravit. 52, 81 (2020).
  • [10] LISA Consortium Waveform Working Group, Waveform Modelling for the Laser Interferometer Space Antenna, arXiv:2311.01300.
  • [11] Stanislav Babak et al., Science with the space-based interferometer LISA. V. Extreme mass-ratio inspirals, Phys.Rev.D 95, 103012 (2017).
  • [12] Bence Kocsis, Nicolas Yunes,and Abraham Loeb, Observable signatures of extreme mass-ratio in spiral black hole binaries embedded in thin accretion disks, Phys. Rev.D 84, 024032 (2011).
  • [13] Andrew Mummery and Steven Balbus, Inspirals from the Innermost Stable Circular Orbit of Kerr Black Holes: Exact Solutions and Universal Radial Flow, Phys. Rev. Lett. 129, 161101 (2022).
  • [14] Andrew Mummery and Steven Balbus, A complete characterisation of the orbital shapes of the non-circular Kerr geodesic solutions with circular orbit constants of motion, Phys. Rev. D 107, 124058 (2023).
  • [15] A. C. Fabian et.al, The soft state of the black hole transient source MAXIJ1820+070: emission from the edge of the plunge region?, MNRAS 493, 5389 (2020)
  • [16] Don N. Page and Kip S. Thorne, Disk-Accretion onto a Black Hole. Time-Averaged Structure of Accretion Disk, The Astrophysical Journal 191, 499 (1974).
  • [17] Jeremy D. Schnittman, Julian H. Krolik, and Scott C. Noble, Disk Emission from Magneto-hydrodynamic Simulations of Spinning Black Holes, The Astrophysical Journal 819, 48 (2016).
  • [18] Christopher S. Reynolds and Mitchell C. Begelman,Intermittent radio galaxies and source statistics, The Astrophysical Journal 487, L135 (1997).
  • [19] Puttarak Jai-akson, Auttakit Chatrabhuti, Oleg Evnin, Luis Lehner, Black hole merger estimates in Einstein-Maxwell and Einstein-Maxwell-dilaton gravity, Phys. Rev. D 96, 044031 (2017).
  • [20] Conor Dyson, Maarten van de Meent, Kerr-fully Diving into the Abyss: Analytic Solutions to Plunging Geodesics in Kerr, Class.Quant.Grav. 40, 19 (2023).
  • [21] Conor Dyson, David Pereñiguez, Magnetic Black Holes: from Thomson Dipoles to the Penrose Process and Cosmic Censorship, arXiv:2306.15751 [gr-qc].
  • [22] B. Carter, Global Structure of the Kerr Family of Gravitational Fields, Phys. Rev. 174, 1559 (1968).
  • [23] Y. Mino, Perturbative approach to an orbital evolution around a supermassive black hole, Phys. Rev. D 67, 084027 (2003).
  • [24] W. Schmidt, Celestial mechanics in Kerr spacetime, Classical Quantum Gravity 19, 2743 (2002).
  • [25] Ryuichi Fujita and Wataru Hikida, Analytical solutions of bound timelike geodesic orbits in Kerr spacetime, Classical Quantum Gravity 26, 135002 (2009).
  • [26] Chen-Yu Wang, Da-Shin Lee, and Chi-Yong Lin, Null and time-like geodesics in Kerr-Newman black hole exterior, Phys. Rev. D 106, 084048 (2022).
  • [27] M. Abramowitz and I. A. Stegun, Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th ed. (Dover, New York, 1964).
  • [28] S. E. Gralla and A. Lupsasca, Null geodesics of the Kerr exterior, Phys. Rev. D 101, 044032 (2020).
  • [29] R. Abuter, et al. (GRAVITY Collaboration), Detection of orbital motions near the last stable circular orbit of the massive black hole SgrA* Astronom. Astrophys. 618, L10 (2018).
  • [30] M. Zajacek, A. Tursunov, A. Eckart, and S. Britzen, On the charge of the Galactic centre black hole, Mon. Not. R. Astron. Soc. 480, 4408 (2018).
  • [31] M. Zajacek, A. Tursunov, A. Eckart, S. Britzen, E. Hackmann, V. Karas, Z. Stuchlik, B. Czerny, J.A. Zensus, Constraining the charge of the Galactic Centre black hole, J. Phys. Conf. Ser. 1258, 012031 (2019).
  • [32] Eva Hackmann and Hongxiao Xu, Charged particle motion in Kerr-Newmann space-times, Phys. Rev. D 87, 124030 (2013).
  • [33] G. Compère, Y. Liu, and J. Long, Classification of radial Kerr geodesic motion, Phys. Rev. D 105, 024075 (2022).
  • [34] L. C. Stein and N. Warburton, Location of the last stable orbit in Kerr spacetime, Phys. Rev. D 101, 064007 (2020).
  • [35] Janna Levin, Gabe Perez-Giz, Homoclinic orbits around spinning black holes I: Exact solution for the Kerr separatrix, Phys. Rev. D 79, 124013 (2009).
  • [36] Yi-Ting Li, Chen-Yu Wang, Da-Shin Lee, and Chi-Yong Lin, Homoclinic orbits in Kerr-Newman black holes, Phys. Rev. D 108, 044010 (2023).
  • [37] Amos Ori and Kip S. Thorne, Transition from inspiral to plunge for a compact body in a circular equatorial orbit around a massive, spinning black hole, Phys. Rev. D 62, 124022 (2000).
  • [38] R. O’Shaughnessy, Transition from inspiral to plunge for eccentric equatorial Kerr orbits, Phys. Rev. D 67, 044004 (2003).