The Earth's core formation due to the Rayleigh-Taylor instability

The recent theories of planetary formation lead to a gravitationally unstable structure of the proto-Earth in the accretion stage, which is composed of three layers: an innermost undifferentiated solid core, an intermediate metal-melt layer, and an outermost silicate-melt layer. Taking this configuration as an initial state, we investigate the Earth's core formation due to the Rayleigh-Taylor instability by using the quantitative results on the instability in a self-gravitating fluid sphere obtained from another paper (S. Ida, Y. Nakagawa, and K. Nakagawa, submitted). We find that the instability occurs through the translational mode on a time scale of about 10 hr if the thickness of the metal-melt layer ⪆1 km. This leads to the conclusion that the Earth's core began to form through the translation of the innermost undifferentiated solid core as soon as the outer layer was melted and differentiated in the late accretion stage. In addition, we examine the rotational effects of the instability; the translation occurs most often along the rotational axis. But this preference is weak, since the rotational energy is small compared to the gravitational one.     

Dynamical behavior of nuclear reaction systems disturbed by fluid motion in the solar core

Using Euler's equation of motion, the equation for disturbed fluid motion against a hydrostatic equilibrium has been derived, and the nonequilibrium dynamical equation of a P-PI nuclear reaction system driven by He³ has been analysed using developed nonequilibrium theory. We find that the system in the solar core is unstable in the layer extending from about 0.2R to 0.4R if the core is disturbed by fluid motion; this instability may be related to thermal diffusion.     

Chemical composition of the Earth after the giant impact

The giant impact hypothesis for the origin of the Moon has been widely accepted. One of the most important features of this hypothesis is that the impactor's metallic core was incorporated in the Earth after impact. If the mass of the impactor is 0.82 × 10²⁷ g, the mass of the impactor core was estimated to be 0.19 × 10²⁷ g, which is about 1/10 of present Earth's core. Liu (1982) derived the bulk composition of the Earth from CI chondrites, and concluded that the Fe content of his model appears to be low in comparison with the present Earth, which, however, can be rationalized by the addition of impactor core into the proto-Earth developed by Liu (1982). If the impactor's mantle contains 14 wt% FeO as suggested, the mass ratio of impactor/proto-Earth should not exceed 0.22. The same ratio is not likely to exceed 0.30, if a giant blowoff did not occur during impact.     

Two-dimensional stochastic motions and the problem of differential rotation

This paper deals with the spatial dependence of the angular velocity in a rotating turbulent fluid sphere. The original turbulence unaffected by the global rotation is assumed to be two-dimensional where the stochastic force field producing the turbulence does not possess a radial component. By using results of earlier papers we proceed to the treatment of a rotational rate, , no longer small compared to _c (frequency of turbulent mode). It is shown that for _c the angular velocity increases with increasing radius but no latitudinal dependence exists. Contrary to this, for 2 _c an equatorial acceleration is possible and related to negativity of the two-dimensional eddy viscosity. Furthermore, the outer layers rotate faster than the inner ones. These findings coincide with Gilman's numerical results. Ward's observations, as well as the characteristic scales of supergranulation and giant cells, suggest the presence of negative two-dimensional eddy viscosity on the Sun.     

Evidence for a small,high-Z,iron-like solar core

Radial and nonradial oscillation equations without and with the gravitation perturbation (with and without the Cowling approximation, CA) are solved numerically using the profile from a more accurate high-Z core (HZC) solar model. This more accurate HZC model was generated with the CRAY X-MP/48 supercomputer at the San Diego Supercomputer Center. Frequencies of oscillation in the five-min band (5MB) and frequencies with period near 160 min are presented in tables and plotted in echelle diagrams. The model was generated by integrating the stellar structure equations from the center to he surface, as done in Rouse (1964), using a maximum space step, ;x _m = 5 × 10^–4, decreasing to 10^–6 in the hydrogenionization zone just below the photosphere. Two subsets of space mesh points are used to calculate the oscillation frequencies, viz., one with a maximum space step of 5 × 10^–3, decreasing to 10^–6 with a total of 621 points (mesh 5I) and the other with a maximum space step of 2 × 10^–3, with a total of 867 points (mesh 5J).With the surface boundary condition applied at x = 1.0, the l – 1 degree nonradial frequencies with CA and the l-degree frequencies without CA are in very good agreement with the frequency spacings for observed frequencies of oscillation labeled l = 1 to 5, but with the l – 1 frequencies with CA about 10 Hz or so less than the observations and the l frequencies without CA about 10 Hz or so greater than the observations. And for the Duvall and Harvey (1983) observations labeled l = 10 and l = 20, the l = 9 and l = 19 nonradial solutions with CA agree to about 5 Hz or less with the observations. Considering from the two preceeding papers in this series that increasing the density in the outer envelope and photosphere will increase the 5MB frequencies and applying the outer boundary condition at x > 1.0 will decrease the 5MB frequencies, the net affects of such changes could move one or the other set of frequencies closer to the observations — or require a slightly different model structure to obtain accurate agreements with the values of the observed frequencies throughout the 5MB.In either case, it is concluded that the first-order, radially-symmetric structure of the model outside the HZC is close to the structure of the real Sun. This is of fundamental importance because a real gas adiabatic temperature gradient (Rouse, 1964, 1971) is used in the outer convective region without free parameters.Other aspects of agreements and differences between radial and nonradial solutions, with CA and without CA are discussed. In particular, the l = 4, 6, 8, and 9 g-mode solutions with CA indicate that the observed 160.01 min period may be a common l-mode period of oscillation. More research is proposed.     

Dynamical instability of the envelope of red supergiants and the lower mass limit for carbon detonation supernovae

We investigated the lower mass limitM _l for the carbon detonation supernovae by testing the dynamical instability of the envelopes of red supergiants. It was found that the dependence ofM _l on the mixing lengthl of convection is appreciable. As a smaller value ofl is assumed,M _l becomes larger. It may be as large as 8M ifl is a third of the pressure scale-height. This is one of the ways to remove the difficulty of overproduction of iron-peak elements involved in the model of the carbon detonation supernovae.     

A comparison between two internal structure models of the Earth

In view of the elastic deformation of the Earth we performed the comparative study for the Earth's models 1066 A and PREM, calculated the static Love numbers from degree 2 to 30, and discussed the relative variations of the second degree Love numbers and their combinations due to the variation of the position of the core-mantle boundary, due to the redistribution of V _p, V _s, and in the lower mantle, and due to the possible rigidity in the outer core. From the above-mentioned discussions we recommended that the Standard Earth Model (SEM) should include two kinds of models—one is oceanless, and another has an oceanic surface. Finally, we calculated the astronomicgeodetic parameters, which are consistent with the primary constants in the IERS Standards, of the SEM.     

Permissible modes of non-thermal convection in planetary mantles

Boundary conditions are imposed upon the solutions of the conservation equations for non-thermal convective motion in a self-gravitating, homogeneous, non-rotating sphere of radiusR, consisting of a core extending to a fraction of the radius of the sphere and with a viscous mantle overlying the core. It is shown that convective modes are permissible in the mantle only for certain values ofn and .     

Rayleigh-Taylor instability of rotating fluid in the presence of a vertical magnetic field

The Rayleigh-Taylor instability of two rotating superposed fluids in the presence of a vertical magnetic field has been investigated. It is shown thatn ² is purely real, wheren is the growth rate of a perturbation. In the basis of this fact it is shown that a unique dispersion relation exists if the lighter fluid lies beneath the heavier one. However, if the heavier fluid lies beneath the lighter fluid, then no unique dispersion relation exists. The effect of rotation is to slow down the rate at which potentially unstable stratification departs from the equilibrium position.     

On properties of the fourier harmonics in the limit-cycle models of classical cepheids

The Fourier coefficients of the hydrodynamic variables are calculated for the limit-cycle models of classical Cepheids having periods from 7.2 to 10.9 days. In adiabatically pulsating layers of the stellar envelope, each Fourier harmonic of orderk 8 is shown to be identified with a corresponding standing wave, so that the pulsation motions of the adiabatic layers may be represented as a superposition of standing waves. Each Fourier harmonic of orderk may also be identified with the eigenfunction of orderl of the linear adiabatic wave equation when the resonance condition _l /₀ =k is fulfilled. The spectra of the oscillatory moment of inertia and the spectra of kinetic energy obey the power law for the Fourier harmonics of orderk 15, the spectrum slope being steeper for shorter pulsation periods. In the helium and hydrogen ionizing regions all of the Fourier harmonics drive the pulsation instability, whereas in the radiative damping region the mechanical work done by each Fourier harmonic is negative. In classical Cepheids having periods shorter than 10 days the period dependence of the secondary bump is due to phase changes of the second order Fourier harmonic in the outer nonadiabatic layers of the stellar envelope. At a pulsation period of II 9.7 days the second order Fourier harmonic is identified with the second overtone. At periods II > 10 days the second order Fourier harmonic tends to be attracted by the fundamental mode in such a way that their phases coincide in the outer layers of the stellar envelope.     

Representation of the Earth's gravitational potential

The paper represents the Earth's gravitational potentialV, outside a sphere bounding the Earth, by means of its difference V from the author's spheroidal potential. The difference V is in turn represented as arising from a surface density on the sphere bounding the Earth. Because of the slow decrease with ordern of the normalized coefficients in the spherical harmonic expansion ofV, the density anomalies from which the higher coefficients arise must occur in regions close to the Earth's surface. The surface density is thus an idealization of the product of the density anomaly and the crustal thicknessb. Values of are computed from potential coefficients obtained from two sources, Rapp and the Smithsonian Astrophysical Observatory. The two sources give qualitative agreement for the values of and for its contour map. The numerical values obtained for are compatible with the idea that the responsible density anomalies are reasonably small, i.e., less than 0.05 g/cm³, and occur in the crust alone.This paper was prepared under the sponsorship of the Electronics Research Center of the National Aeronautics and Space Administration through NASA Grant NGR 22-009-311.     

Spectrum of the galactic magnetic fields

The magnetic fields observed in the galactic disc are generated by the differential rotation and the helical turbulent motions of interstellar gas. On the scalesl=2k ^–1 which lie in the intervall ₀<l<l _e (l ₀100 pc is the energy-range scale of the galactic turbulence), the spectral density of the kinetic energy of turbulence (k ^–5/3) exceeds the magnetic energy spectral density (k ^–1). The equipartition between magnetic and kinetic energies is reached atl=l _e 6 pc in the intercloud medium and is maintained down to the scalel=l _d 0.03 pc. In dense interstellar cloudsl _e is determined by the individual cloud size andl _d 0.1 pc.The internal turbulent velocities in Hi clouds (cloud size is assumed to be 10 pc) lie in the range from 1.8 to 5.6km s^–1, fitting well within the observed range of internal rms velocities. Dissipation of the interstellar MHD turbulence leads to creation of temperature fluctuations with amplitudes of 150 K and 65 K in dense clouds and intercloud medium, respectively. The small-scale fluctuations observed in the interstellar medium may arise from such perturbations due to the thermal instability, for instance. Dissipation of the MHD turbulence energy provides nearly half of the energy supply needed to maintain the thermal balance of the interstellar medium.     

Sensitivity of solar eigenfrequencies to the age of the sun

The increasing central concentration of the Sun with age modifies the acoustic eigenfrequencies. In particular, the frequency separation d _l =3(2l+3)^–1v _n,l –v _n–1,l+2 for modes with l + 1/2 n decreases as nuclear reactions augment the molecular-weight gradient in the energy-generating core. If, for example, the Sun were older than is generally believed, one might therefore expect d _l to be smaller than current theoretical predictions. On the other hand, to ensure that the luminosity is consistent with observations, the presumed initial hydrogen abundance would need to be enhanced, thereby reducing the resultant molecular-weight gradient. Thus there is some degree of cancellation of the two major factors that determine d _l .Various authors have either reported directly on the sensitivity of d _l , or have provided the information from which it can be calculated. We have added our own computations. There is broad agreement amongst the results: d _l diminishes with the presumed age of the Sun at the rate of about 1 Hz per Gy for l = 0; the magnitude of the rate appears to decline with increasing l.     

Stability of the stellar structure and nonequilibrium dynamical characteristics of (g — T) effects in nuclear reactions in stellar cores

The nonequilibrium dynamical theory of thermonuclear reaction in the stars is used to analyse the stabilities of P-PI reactions in the solar core and 3 reactions of helium burning in the core of red giant stars. The constant stability of stars on the main sequence and the helium flash instability have been explained from a new point of view. Calculations show that the (g - T) term in the dynamical equation has a damaging effect on the stability of the thermonuclear reaction, but its intensity in the core of red giant stars is ~ 10⁷ times that in the solar core. It may be intimately related to some instability of the stellar structure. The effect on the flash instability is especially analysed.     

Shear flow instability near the cusp resonance in magnetopause boundary layer

Shear flow instability is studied in the planar magnetopause boundary layer region by treating the plasma as compressible. A necessary criterion for instability near the cusp resonance is obtained analytically. The criterion depends on plasma, Alfvén Mach numberM _A and the ratio of the scale lengths of the gradients in the flow and Alfvén velocities. The instability at the cusp resonance layer can be excited rather easily for the low plasma and for shear flow scale length smaller than the typical scale length over which Alfvén velocity varies. The growth rate for instability is obtained for any from a cubic equation. The unstable modes may contribute to the ULF wave activity at the magnetopause.     

The physical properties of an analytic model for a relativistic star

An exact solution of Einstein's equations for a static isentropic perfect fluid sphere is examined in detail. The analysis yields a strong indication that the model isstable with respect to infinitesimal radial pulsations. This means that the temperature is decreasing outwards. We prove that the adiabatic speed of sound is everywhere less than the speed of light if and only if the radius of the sphere is larger than 1.61 times its Schwarzschild radius. We further show that the strong energy condition is fulfilled everywhere if and only if the radius is larger than 1.76 times the Schwarzschild radius. The necessary and sufficient condition for the speed of sound to be decreasing outwards is given, and we find that this criterion is fulfilled if the fluid is causal. Taking the values of the pressure and the density to be somewhere given by the maximum values from Baymet al.'s equation of state, i.e., ₀=5.1×10¹⁴ g cm^–3 andp ₀=7.4×10³³ dyne cm^–2, we find the maximum mass of the fluid sphere to be 2.5 solar masses.Dedicated to the memory of the late George Cunliffe McVittie (1904–1988).     

Thermal convection instability of a two-component fluid layer in hydromagnetics

The thermal convection instability of a two component fluid layer subjected to a temperature gradient has been studied in the presence of an applied magnetic field. The associated thermal diffusion separation has a predominant effect even when the separations are small. Solutions for the non-oscillatory marginal states have been obtained. It is shown that the concentration gradient has a stabiliting or destabilizing effect according as T<or>0. Approximate solutions for the oscillatory solutions have been obtained by the method of variational principle and the dispersion relation has been solved numerically.     

On magnetoconvection in a rotating fluid layer of high thermal conductivity

This paper explores the thermal instability of a plane fluid layer rotating rapidly about a vertical axis in the presence of a uniform vertical magnetic field. The thermal diffusivity is taken to be large compared with the magnetic diffusivity . For a range of parameters it is shown that an increase of the magnetic field leads to adecrease in the critical Rayleigh number, and two quite distinct physical mechanisms appear to be involved.     

Rayleigh-Taylor instability as a mechanism for corona formation on Venus

In this study we explore the idea that coronae have formed on Venus as a result of gravitational (Rayleigh-Taylor) instability of the lithosphere. The lithosphere is represented by a system of stratified homogeneous viscous layers (low-density crust over high density mantle, over lower density layer beneath the lithosphere). A small harmonic perturbation imposed on the base of the lithosphere is observed to result in gravitational instability under the constraint of assumed axisymmetry. Topography develops with time under the influence of dynamic stress associated with downwelling or upwelling, and spatially variable crustal thickening or thinning. Topography may therefore be elevated or depressed above a mantle downwelling, but the computed gravity anomaly is always negative above a mantle downwelling in a homogeneous asthenosphere. The ratio of peak gravity to topography anomaly depends primarily on the ratio of crust to lithospheric viscosity. Average observed ratios are well resolved for two groups of coronae (∼40 mgal km⁻¹), consistent with models in which the crust is perhaps 5 times stronger than the lithosphere. Group 3a (rim surrounding elevated central region) coronae are inferred to arise from a central upwelling model, whereas Group 8 (depression) coronae are inferred to arise from central downwelling. Observed average coronae radii are consistent with a lithospheric thickness of only 50 km. An upper low-density crustal layer is 10-20 km thick, as inferred from the amplitude of gravity and topography anomalies.     

The flow of an infinite compressible fluid past a rigid,gravitating sphere and interacting degenerate star-red giant binaries

We compute two examples of the flow structure of an infinite medium flowing hypersonically past a non-accreting, gravitating, rigid sphere. The resulting flow depends strongly on the ratioA of kinetic energy at infinity to potential energy on the sphere's surface per unit mass.A=0.25 yields a flow rather like that past a hard, gravitationless sphere upstream, but with a closed shock downstream.A=0.028 yields a circulating eddy flow downstream of the sphere which causes the isodensity contours to be extended upstream. Application to a compact object immersed in a binary companion is discussed. We pictorially illustrate the fluid flow past a degenerate star starting to spiral into its giant companion. The accretion rates onto hard gravitating objects can be many orders of magnitude less than the classical Hoyle-Lyttleton-Bondi rates unless cooling dominates the flow.