Simulations of small solid accretion on to planetesimals in the presence of gas
MNRAS 472, 3543–3553 (2017)
doi:10.1093/mnras/stx1964
Advance Access publication 2017 August 2
Simulations of small solid accretion on to planetesimals in the presence
of gas
A. G. Hughes‹ and A. C. Boley‹
Department of Physics and Astronomy, University of British Columbia, 2329 West Mall, Vancouver, BC V6T 1Z4, Canada
Accepted 2017 July 28. Received 2017 July 28; in original form 2017 June 5
The growth and migration of planetesimals in a young protoplanetary disc are fundamental to planet formation. In all models of early growth, there are several processes that can
inhibit grains from reaching larger sizes. Nevertheless, observations suggest that growth of
planetesimals must be rapid. If a small number of 100 km sized planetesimals do manage to
form in the disc, then gas drag effects could enable them to efficiently accrete small solids
from beyond their gravitationally focused cross-section. This gas-drag-enhanced accretion can
allow planetesimals to grow at rapid rates, in principle. We present self-consistent hydrodynamics simulations with direct particle integration and gas-drag coupling to estimate the rate
of planetesimal growth due to pebble accretion. Wind tunnel simulations are used to explore a
range of particle sizes and disc conditions. We also explore analytic estimates of planetesimal
growth and numerically integrate planetesimal drift due to the accretion of small solids. Our
results show that, for almost every case that we consider, there is a clearly preferred particle
size for accretion that depends on the properties of the accreting planetesimal and the local
disc conditions. For solids much smaller than the preferred particle size, accretion rates are
significantly reduced as the particles are entrained in the gas and flow around the planetesimal. Solids much larger than the preferred size accrete at rates consistent with gravitational
focusing. Our analytic estimates for pebble accretion highlight the time-scales that are needed
for the growth of large objects under different disc conditions and initial planetesimal sizes.
Key words: hydrodynamics – planets and satellites: formation – planets and satellites: physical evolution.
1 I N T RO D U C T I O N
The planet formation process is critically dependent on the coagulation of dust grains into larger bodies. The time-scale for
significant solid growth is thought to be a few million years,
based on the chronologies of meteorites (Villenueve, Chaussidon &
Libourel 2009), the fraction of inferred protoplanetary discs relative
to cluster ages (Mamajek et al. 2009), and the incidence of gas giant planets (Cumming et al. 2008). The latter provides a constraint
on core-nucleated instability (Pollack et al. 1996) as a dominant
formation model for giant planets, which requires the formation of
a large solid core before the gaseous disc dissipates. Furthermore,
morphological features in discs such as HL Tau (ALMA Partnership 2015) and TW Hydrae (Andrews et al. 2016) may indicate the
presence of embedded planets, which if correct, would require the
rapid growth of large solids in some discs.
E-mail: (AGH); (ACB)
The early stages of planet formation rely on the growth of micron
to submillimetre grains into planetary sizes. As grains reach submillimetre and millimetre radii, further growth may be hindered due
to bouncing at low collisional speeds and fragmentation at higher
collisional speeds (Blum & Wurm 2008; Testi et al. 2014). Regions
of high solid density could potentially build planets or planetesimals
through direct gravitational collapse (Goldreich & Ward 1973) if
solids become concentrated due to some mechanism such as vertical
settling (D’Alessio, Calvet & Hartmann 2001). However, vertical
settling is opposed by turbulent mixing, which can lift grains away
from the mid-plane of the disc (Dubrulle 1995), slowing or preventing significant grain growth. Furthermore, gas is expected to orbit at
sub-Keplerian speeds due to the negative outward pressure gradient
in the disc. This causes objects that are on Keplerian orbits to face
a headwind, resulting in the loss of angular momentum for small
solids. Without any mitigating factor, centimetre to metre-sized objects would consequently spiral into their stars on short time-scales
(Adachi, Hayashi & Nakazawa 1976; Weidenschilling 1977). While
spiral arms or other pressure perturbations could concentrate solids
C 2017 The Authors
Published by Oxford University Press on behalf of the Royal Astronomical Society
ABSTRACT
3544
A. G. Hughes and A. C. Boley
2 T H E O RY A N D M E T H O D S
Gas in a protoplanetary disc is expected to orbit at sub-Keplerian
speeds due to support from a negative gas pressure gradient. In the
absence of such support, solids will orbit at Keplerian speeds. This
velocity difference causes the solids to encounter a headwind with
the gas, and alters the orbit of solids depending on size. Small solids
quickly become entrained with the gas flow, while larger solids carry
MNRAS 472, 3543–3553 (2017)
Table 1. Disc and object properties used in the hydrodynamics simulations, following the flared disc profile outlined in equation (1) along with
predicted best-accreted pebble size for each distance as found using equation (9). While most disc conditions change dramatically at different stellar
separations, the relative velocity between the gas and the planetesimal stays
roughly constant.
r (au)
ρ g g cm−3
T (K)
v rel m s−1
s (cm)
10
3
1
0.3
0.1
3.2 × 10−12
95
170
300
550
950
54.8
54.7
54.7
54.6
54.6
0.0005
0.01
0.2
3.0
50.0
6.4 × 10−11
1.0 × 10−9
2.0 × 10−8
3.1 × 10−7
enough momentum that gas drag only affects their motion over very
long time-scales.
Throughout this paper, we assume a disc model with the following
mid-plane profiles:
r −n
;
ρ = ρ0
au
r −m
;
(1)
T = T0
au
where r is the stellar separation measured in au,
ρ 0 = 1 × 10−9 g cm−3 and T0 = 300 K are gas density and
temperature values at 1 au. For our model, we use the exponents m
R
= 0.5 and n = 2.5. We take the gas to be ideal with P = μg T ρ and
μ = 2.3 g mol−1 .
At any location in the disc, the difference between the circular
Keplerian orbit and the circular gas motion is given by
r dP
GM
GM
−
+
(2)
vrel =
r
r
ρ dr
⎛
= vK ⎝1 −
(1 − (n + m)
c0
vK
2
⎞
r0 m ⎠
r
(3)
for circular Keplerian speed v K at r and temperature and density
power-law exponents m and n, respectively. The isothermal sound
speed at r0 = 1 au is given by c0 . The gas conditions at various stellar
separations are shown in Table 1, along with the corresponding
relative wind speeds. For our assumptions, the wind speed v rel stays
roughly constant with stellar separation.
As a solid experiences a relative wind, drag forces couple them
to the gas. The characteristics of this coupling depend on the drag
regime, which can be divided into two basic types: Epstein and
Stokes. A solid falls into the Epstein regime when its radius is
much smaller than the mean free path (...truncated)