Barneset al.Reply:

Abstract

Barnes et al. Reply: Greaves et al. [1] raise issues regarding our Letter [2] which contested the validity of a reported [3] first-order liquid-liquid transition (LLT) in supercooled ðY 2 O 3 Þ x ðAl 2 O 3 Þ 100Àx (or AYx) with x ¼ 20. The principal concerns are (i) our use of uncorrected pyrometric temperature data, (ii) the accuracy of the composition of our samples, and (iii) surface scattering dominating our SANS signal.(i) An emissivity correction of ¼ 0:92 [4] implies an underestimate in temperature of 30 K at 2273 K which is small compared to the temperature gradient across a levitated sample (see the pyrometry traces in Fig. 1 of [1]). An emissivity correction does not materially affect the conclusions presented in [2]. The calculated cooling curves in [2] are correct with respect to the explicitly stated molar normalizations.(ii) Gravimetric and electron probe methods show that our sample compositions are reliably reproduced to AE1%[5], consistent with their visual appearance [5]. Also, our AY20 samples supercooled and crystallized at ' 1500 K, well below the AY15 crystallization temperature of ' 1925 K [3]. The assertion in [1] that our AY20 sample corresponds to AY15 is mainly from the peak positions and heights in the measured x-ray structure factor SðQÞ. Figure 1 shows these parameters (determined from the numerical data sets of [1,3,6] and from additional experiments) and also shows that our measured SðQÞ for AY20 compares favorably with the revised Advanced Photon Source (APS) data [1]. Differences in peak positions will arise from systematic errors in different diffractometer calibrations.(iii) A component of the SANS signal will come from the surface of our levitated sample and, for a spherical sample, will have a cutoff at Q ¼ 0:015 A À1 followed by a Q À4 falloff. In the absence of sample density fluctuations this would constitute the only SANS signal. However, calculations based on simple models show that this would not mask a change taking place in the SANS signal from the reported LLT [3]. In contrast, SAXS experiments of the type used in [3] need to stably maintain the sample and incident beam positions to better than 10 m.Greaves et al.[1] interpret the pyrometry data from our SANS experiment (trace B in their Fig. 1) as arising from a polyamorphic rotor, without reference to any other experimental observations. The temperature variations were, however, clearly observed to result from a small gas bubble in the sample which led to rotational instability. Given the evidence in (ii) that our sample is AY20, the assumption of a polyamorphic rotor would imply a second LLT in AY20 at 1927 K, an unlikely scenario.

Dependence of intrinsic rotation reversals on collisionality in MAST

“…1 for simplicity (it was a small effect in other Ohmic intrinsic rotation studies [3] and cannot generate a reversal but only amplify flow generated by other mechanisms). The turbulent momentum and energy diffusivities, χ φ and χ i , are related by the turbulent Prandtl number P r = χ φ /χ i , with typical value P r ≈ 0.7 [26,27] used in the model. We want to identify contributions to Π int that reverse sign when the density changes.…”

Glass formation of rare earth aluminates by containerless processing

“…Many studies have focused on the liquid-liquid phase transition proposed in a Y 2 O 3 -Al 2 O 3 undercooled melt, which is always a controversial issue as to its existence [21][22][23][24][25]. On the other hand, in most studies on R 2 O 3 -Al 2 O 3 binary glasses, R was La or Y, and there are few reports of other R systems.…”

Intrinsic rotation with gyrokinetic models

“…only if the rotation is ordered to be sonic, RΩ ζ ∼ v ti . This limit is the high flow ordering [55][56][57][58], and it has been adopted by many of the numerical studies of turbulent momentum transport performed so far [3,5,6,[59][60][61][62]. If the rotation is sonic, most of the terms in (1) and (2) can be neglected to lowest order in ρ * , leaving only V i = Ω ζ Rζ, with Ω ζ = −c(∂φ 0 /∂ψ).…”

“…The presence of toroidal rotation is beneficial: it suppresses large-scale MHD instabilities [1,2] and reduces the turbulent radial flux of particles, energy and momentum [3][4][5][6][7][8]. In a number of experiments (e.g.…”