Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
E VOLUTION OF ECCENTRICITY AND ORBITAL INCLINATION OF MIGRATING PLANETS IN 2:1 MEAN MOTION RESONANCE Jean Teyssandier & Caroline Terquem Institut d’Astrophysique de Paris & University of Oxford A BSTRACT We determine, analytically and numerically, the conditions needed for a system of two migrating planets trapped in a 2:1 mean motion resonance to enter an inclination-type resonance. We provide an expression for the asymptotic equilibrium value that the eccentricity of the inner planet reaches under the combined effects of migration and eccentricity damping. We also show that, for a ratio of inner to outer masses below unity, the inner eccentricity has to pass through a value of order 0.2-0.3 for the system to enter an inclination-type resonance. Numerically, we confirm that an inclination-type resonance may also be excited at another, larger, eccentricity of 0.6, as found by previous authors. We find that, for a mass ratio below unity, the system cannot enter an inclination-type resonance if the ratio of eccentricity damping timescale to migration timescale is smaller than 0.2. As the disc/planet interaction is characterized by such ratios to be of the order of 10−2, we conclude that excitation of inclination through the type of resonance described here is very unlikely to happen in a system of two planets migrating in a disc. I NTRODUCTION I LLUSTRATIVE CASES As of today, 98 extrasolar multiple planet systems have been detected by radial velocity surveys and 420 by the Kepler mission. A significant fraction of these systems contain planet pairs in or near a 2:1 mean motion resonance. Capture in mean motion resonance is thought to be the result of convergent migration of planets. The evolution of planetary systems in mean motion resonance has been studied with three– dimensional simulations [e.g., Thommes and Lissauer, 2003]. It was found that a system of planets in an eccentricity–type resonance may enter an inclination–type resonance if the eccentricity of the inner planet becomes large enough. In this context, high orbital inclinations can be reached starting from nearly coplanar configurations. Studies of resonant inclination excitation are important as this mechanism has been proposed to explain the fact that some extrasolar planets have an orbit which is inclined with respect to the stellar equatorial plane. In Teyssandier and Terquem [2014], we derive analytically a necessary condition for the onset of inclination–type resonance. The analysis has to be done to second order in eccentricities and inclinations. We also perform numerical simulations to investigate the regime of high eccentricities. Left column, from top to bottom: Semi–major axes (in AU), eccentricities and inclinations of the two planets, ∆ϖ = ϖi − ϖo and ∆Ω = Ωi − Ωo versus time (in years). Right column, from top to bottom: resonant angles ϕ1, ϕ2, ϕ6, ϕ7 and ϕ8 †† versus time (in years). All the angles are given in degrees. In the plots displaying the eccentricities and inclinations, the upper curves represent e i and I i, respectively, whereas the lower curves represent e o and I o, respectively. R ESONANT M IGRATION We are interested in the case where convergent migration of two planets has led to capture into a 2:1 mean motion resonance. The inner planet has a mass m i , and the outer one a mass m o. We define q = m i /m o and α = a i /a o, the constant semi-major axis ratio. The orbital evolution of each planet is a combination of disc-planet interactions (in the form of orbital migration and eccentricity damping) and planet–planet interactions (in the form of resonant and secular interactions). The evolution of semi-major axis and eccentricity is : 0 time (years) 0 0 time (years) time (years) Evolution of a system in 2:1 mean motion resonance with q = 0.7 and t e /t a = 0.25. Shortly after the beginning of the simulation, the planets are captured into an eccentricity–type resonance: e i grows until it reaches e i,eq = 0.2 and ∆ϖ, ϕ1 and ϕ2 librate about 0◦. After t ≃ 8 × 105 years, e i starts decreasing while e o gets larger, and the value about which ϕ2 librates switches rather abruptly to 180◦. Throughout the evolution, e i stays too small to allow for inclination–type resonance. Accordingly, the inclinations remain small, the resonant angles ϕ6, ϕ7 and ϕ8 behave chaotically and ∆Ω = 0 throughout the evolution. Here t a and t e are the migration and eccentricity damping timescales. Migration is only applied to the outer planet. The resonant part is modelled by a disturbing function which is expanded to second-order in eccentricity and inclination [Murray and Dermott, 2001]. From these equations we find that the eccentricities are excited by the resonant migration, but are damped by the planetdisc interactions. The inner planet’s eccentricity reaches the equilibrium value ( 1 t e /t a e i,eq = p 2 1+ q α time (years) Evolution of a system in 2:1 mean motion resonance with q = 0.7 and t e /t a = 0.8. Here again, shortly after the beginning of the simulation, the planets are captured into an eccentricity– type resonance: e i grows until it reaches e i,eq = 0.35 and ∆ϖ, ϕ1 and ϕ2 librate about 0◦. After t ≃ 3.2 × 106 years, e i starts decreasing while e o gets larger, and the value about which ϕ2 librates switches to 180◦. When e i ≃ 0.3, an inclination–type resonance starts to develop. Accordingly, the resonant angles ϕ6, ϕ7 and ϕ8 start librating about 180◦, 0◦ and 180◦, respectively, while ∆Ω librates about 180◦. The inclinations grow quickly. 10 ) 2a i e2i da i da i = − , dt dt res te ( ) da o a o 2a o e2o da o = − − , dt ( dt ) res t a te ei de i de i = − , dt ( dt )res t e eo de o de o = − . dt dt res t e 0 0 time (years) time (years) Evolution of a system in 2:1 mean motion resonance with q = 0.7 and t e /t a = 4. Here again, shortly after the beginning of the simulation, the planets are captured into an eccentricity–type resonance and e i and e o grow. In the present case, ϕ1 and ϕ2 librate about 0◦ throughout the simulation. When e i ∼ 0.6, the system enters an inclination–type resonance and ∆Ω, ϕ6, ϕ7 and ϕ8 start librating about 180◦, 180◦, 0◦ and 180◦, respectively, while the inclinations start growing. †† ϕ1 = 2λo − λi − ϖi, ϕ2 = 2λo − λi − ϖo, ϕ6 = 4λo − 2λi − 2Ωi, ϕ7 = 4λo − 2λi − Ωi − Ωo, ϕ8 = 4λo − 2λi − 2Ωo. PARAMETER S PACE τe / τa ( 0 1 0.1 0.1 1 q = mi / mo no I-res I-res, φ2 = 180 I-res, φ2 = 0 I-res, φ2 = 180, tres > 3 Myr I-res, φ2 = 0, tres > 3 Myr This plot shows the occurrence of inclination–type resonance as a function of q = m i /m o and t e /t a in systems evolved between 0 and 6 Myr. Crosses represent systems that did not enter an inclination–type resonance, open and filled symbols represent systems that entered an inclination–type resonance with (ϕ1, ϕ2) = (0◦, 180◦) or (ϕ1, ϕ2) = (0◦, 0◦), respectively. Circles and triangles represent systems that entered an inclination–type resonance before or after t = 3 Myr, respectively. The dashed ( ) p lines are the curves t e /t a = 4e2i,res 1 + q α with e i,res = 0.3 (lower line) and e i,res = 0.57 (upper line). In agreement with theoretical expectations, systems in between those two lines entered an inclination–type resonance with (ϕ1, ϕ2) = (0◦, 180◦) if evolved long enough, whereas systems above the upper line entered an inclination– type resonance with (ϕ1, ϕ2) = (0◦, 0◦). Hence, it clearly appears that if t e /t a ≲ 0.2, an inclination–type resonance cannot be achieved. For mass ratios of unity and larger, t e /t a has to be of order of unity, at least, for an inclination–type resonance to occur. )1/2 . An inclination–type resonance can be triggered when the nodes precess at the same rate, and match the periaspis precession. It imposes a condition on the inner eccentricity. We find that for most mass ratios, e i,res ≃ 0.3. An inclination–type resonance will be triggered if e i,res < e i,eq, i.e., t e /t a p > 4e2i,res. 1+ q α R EFERENCES C ONCLUSIONS C.D. Murray and S.F Dermott. Solar System Dynamics. Cambridge University Press, 2001. • We have shown that, for a system of two resonant migrating planets to enter an inclination–type resonance, the eccentricity of the inner planet has to reach a value e i,res ≃ 0.2 − 0.3. J. Teyssandier and C. Terquem. . Submitted to MNRAS, 2014. E. W. Thommes and J. J. Lissauer. . ApJ, 597: 566–580, 2003. • A necessary condition for the system to enter an inclination–type resonance is that e i,res < e i,eq. This leads to a condition on the ratio of the eccentricity damping timescale to the migration timescale, t e /t a, as a function of the mass ratio. For q ≤ 1, we find that the system cannot enter an inclination–type resonance if t e /t a < 0.2. • For a wide range of planet masses, the eccentricity damping timescale due to planet/disc interaction is of a few 103 years. As type I and type II migration timescales are on the order of 105 years, it implies that t e /t a ∼ 10−2, too low for the value needed for e i,res to be reached. We conclude that it is unlikely that misaligned systems formed via this mechanism.