Fig. 4
From: Iron diapirs entrain silicates to the core and initiate thermochemical plumes
Diapir descent and conduit formation. a The measured onset time (two standard deviations) versus the viscosity of fluid S2 (black diamonds). The onset time is defined by the time when the horizontal metal pond layer begins to go unstable and sinks crossing the S1−S2 interface (yellow line, inset photograph). Theoretical prediction (black line) for onset of a Rayleigh−Taylor instability, Eq. (1) . Shaded region is revised theory using Eq. (5) for the range h1 from 2.5 to 4.2 cm. Experimental Data 8, 10, 11, 18, 21, and 24. b Descent distance versus time for a diapir during terminal velocity for three experiments which vary gallium layer thickness, hm = 0.25 cm (black squares), hm = 0.5 cm (gray circles), and hm = 1.0 cm (open triangles). Classical Stokes theory for an inviscid sphere (solid line) considers a spherical radius using the actual volume of metal in the diapir. Revised Stokes theory considering reduced drag with a conduit including conduit mass from Eq. (2) is shown by dotted lines for each experiment (Experiments 5, 6, and 8). Average rdiapir is 0.017 m, 0.020 m, 0.025 m and θ is 20°, 30°, 55° for experiments 8, 5, and 6, respectively. The average conduit height, hc, is approximately half the descent distance. Errors in distance and time are smaller than the symbol. c Conduit radius versus time for a smooth metal pond with μ1 = 0.01 Pa ⋅ s and μ2 = 10,987 Pa ⋅ s (Experiment 34). Open square indicates the time when the metal diapir reaches the base of the tank. Three regimes for conduit behavior are identified (see text). d Conduit radius versus time for an emulsified metal pond (this study) with μ1 = 0.1 Pa ⋅ s and μ2 = 11,175 Pa ⋅ s (Experiment 14). Only regimes 1 and 3 are observed in all emulsified experiments
