betaFeedback

WorksPhysics and Astronomy

Boiling and cavitation caused by transient heat transfer in superfluid helium-4

Abstract

Superfluid helium-4 (He II) has been widely utilized as a coolant in various scientific and engineering applications due to its superior heat transfer capability. An important parameter required in the design of many He II based cooling systems is the peak heat flux q∗q^*, which refers to the threshold heat flux above which boiling spontaneously occurs in He II. Past experimental and numerical studies showed that q∗q^* increases when the heating time tht_h is reduced, which leads to an intuitive expectation that very high q∗q^* may be achievable at sufficiently small tht_h. Knowledge on how q∗q^* actually behaves at small tht_h is important for applications such as laser ablation in He II. Here we present a numerical study on the evolution of the thermodynamic state of the He II in front of a planar heater by solving the He II two-fluid equations of motion. For an applied heat flux, we determine the heating time beyond which the He II near the heater transits to the vapor phase. As such, a curve correlating q∗q^* and tht_h can be obtained, which nicely reproduces some relevant experimental data. Surprisingly, we find that there exists a critical peak heat flux qc∗q^*_c, above which boiling occurs nearly instantaneously regardless of tht_h. We reveal that the boiling in this regime is essentially cavitation caused by the combined effects of the first-sound and the second-sound waves in He II. Based on this physical picture, an analytical model for qc∗q^*_c is developed, which reproduces the simulated qc∗q^*_c values at various He II bath temperatures and hydrostatic head pressures. This work represents a major progress in our understanding of transient heat transfer in He II.

Introduction

When saturated liquid 4He{}^{4}\mathrm{He} is cooled to below about 2.17 K, it undergoes a phase transition to the superfluid phase (known as He II) [1]. Phenomenologically, He II can be regarded as a mixture of two miscible fluid components: an inviscid and zero-entropy superfluid and a viscous normal fluid that consists of thermal quasiparticles (i.e., phonons and rotons) [2]. He II has many unique thermal and mechanical properties. For instance, it supports two sound-wave modes: an ordinary pressure-density wave (i.e., the first sound) where the two fluids oscillate in phase, and a temperature-entropy wave (i.e., the second sound) where the two fluids oscillate oppositely [2]. Furthermore, heat transfer in He II is via an extremely effective counterflow mode [2]: the normal fluid carries the heat and moves away from the heat source at a mean velocity vn=q/ρsTv_{n}=q/\rho sT, where qq is the heat flux, TT is the He II temperature, ρ\rho and ss are respectively the total density and the specific entropy of the He II; while the superfluid moves in the opposite direction at a mean velocity vs=−vnρn/ρsv_{s}=-v_{n}\rho_{n}/\rho_{s} to ensure zero net mass flow (here ρs\rho_{s} and ρn\rho_{n} denote the densities of the superfluid and the normal fluid, respectively). When the relative velocity of the two fluids exceeds a small critical value [3], a chaotic tangle of quantized vortex lines are created spontaneously in the superfluid, each carrying a quantized circulation κ≃10−3 cm2/s\kappa\simeq10^{-3}\ \mathrm{cm}^{2}/\mathrm{s} around its angstrom-sized core [4]. A mutual friction force between the two fluids then emerges due to the scattering of the thermal quasiparticles off the quantized vortices [5], which can lead to novel flow characteristics in both fluids [6, 7, 8, 9, 10].

Due to the superior heat transfer capability, He II has been widely utilized in various scientific and engineering applications for cooling devices such as superconducting magnets, superconducting accelerator cavities, and satellites [11]. Many of these applications involve high-flux transient heat transfer from a solid surface to He II, which is a complex process that involves the interplay of counterflow, second-sound emission, and vortex generation [12]. When the heat flux is higher than a threshold value denoted as the peak heat flux q∗q^{*}, boiling of He II can spontaneously occur on the heating surface. This peak heat flux q∗q^{*} is an important parameter needed in the design of many He II based cooling systems, which has been the subject of extensive experimental and numerical studies [13, 14, 15, 16, 17, 18, 19, 20]. A power-law dependence q∗∝th−nq^{*}\propto t_{h}^{-n} of q∗q^{*} on the heating time tht_{h} was reported in literature, where the power index nn varies in the range of 0.25 to 0.5 depending on the magnitude of the applied heat flux and other experimental conditions such as the He II bath temperature TbT_{b} and the hydrostatic head pressure Ph≃ρgHP_{h} \simeq\rho gH (where HH is the He II depth and gg is the gravitational acceleration) [15, 16, 17, 18, 19]. These results may give an intuitive expectation that very high q∗q^{*} may be achievable at sufficiently small tht_{h}. However, knowledge on how q∗q^{*} actually behaves in the high heat flux and short heating time regime is limited. Such knowledge could benefit various research fields such as nanomaterial fabrication via laser ablation in He II [21, 22].

In this paper, we present a numerical study of the q∗q^{*}-tht_{h} correlation with an emphasis on the high heat-flux regime. By solving the He II two-fluid equations of motion coupled with the Vinen’s equation for the vortex-line density [5], we examine the evolution of the thermodynamic state of the He II in front of a planar heater. For an applied heat flux qq, we determine the heating time tht_h beyond which the He II near the heater surface transits to the vapor phase. The obtained q∗−thq^*-t_h curves show good agreement with relevant experimental data. Surprisingly, we find that there exists a critical peak heat flux qc∗q_c^*, above which boiling occurs nearly instantaneously regardless of tht_h. Our analysis shows that the boiling in this regime is indeed heat-induced cavitation on the heater surface caused by combined effects of the first-sound and the second-sound waves in He II accompanying the heat transfer. We discuss an analytical model for evaluating qc∗q_c^* and show that this model can nicely reproduce the simulated qc∗q_c^* values at various He II bath temperatures and hydrostatic head pressures.

The paper is organized as follows. In § 1.1, we outline the numerical model adopted in our study and briefly discuss some characteristic features of transient heat transfer in He II. In § 2, we explain how the q∗−thq^*-t_h correlation is determined at a given TbT_b and the He II depth HH. A comparison of the obtained q∗−thq^*-t_h curves with available experimental data is made, and the appearance of the unexpected critical peak heat flux qc∗q_c^* is discussed. In § 3, we present a systematic study on how qc∗q_c^* depends on TbT_b and HH. The underlying physical mechanism of qc∗q_c^* is explained in § 4. Finally, a brief summary is given in Sec. VI.

Numerical model

There have been a variety of methods developed for studying heat transfer and flow dynamics in He II under the influence of quantized vortices. For a comprehensive review of these methods, one may refer to Refs. [23, 24]. In this study, we adopt the two-fluid hydrodynamic model presented in our previous work [12] to examine 1D heat transfer in He II from a planar heater. This model consists of the following conservation equations for the He II mass, momentum, and energy:

∂ρ∂t+∇⋅(ρv)=0(1)\frac{\partial\rho}{\partial t}+\nabla\cdot(\rho\mathbf{v})=0 \tag*{(1)}
∂(ρv)∂t+∇(ρsvs2+ρnvn2)+∇P=0(2)\frac{\partial(\rho\mathbf{v})}{\partial t}+\nabla(\rho_s\mathbf{v}_s^2+\rho_n\mathbf{v}_n^2)+\nabla P=0 \tag*{(2)}
∂vs∂t+vs⋅∇vs+∇μ=Fnsρs(3)\frac{\partial\mathbf{v}_s}{\partial t}+\mathbf{v}_s\cdot\nabla\mathbf{v}_s+\nabla\mu=\frac{\mathbf{F}_{ns}}{\rho_s} \tag*{(3)}
∂(ρs)∂t+∇⋅(ρsvn)=Fns⋅vnsT(4)\frac{\partial(\rho_s)}{\partial t}+\nabla\cdot(\rho_s\mathbf{v}_n)=\frac{\mathbf{F}_{ns}\cdot\mathbf{v}_{ns}}{T} \tag*{(4)}

where ρv=ρsvs+ρnvn\rho\mathbf{v}=\rho_s\mathbf{v}_s+\rho_n\mathbf{v}_n denotes the total momentum density, PP is the He II pressure, and μ\mu is the chemical potential of He II. Considering the large relative velocity vns=∣vns∣=∣vn−vs∣v_{ns}=|\mathbf{v}_{ns}|=|\mathbf{v}_n-\mathbf{v}_s| of the two fluids at high heat fluxes, we also include the corrections to the thermodynamic properties of He II as proposed by Landau [2, 25]:

μ(P,T,vns)=μ(s)(P,T)−12ρnρvns2(5)\mu(P,T,v_{ns})=\mu^{(s)}(P,T)-\frac{1}{2}\frac{\rho_n}{\rho}v_{ns}^2 \tag*{(5)}
s(P,T,vns)=s(s)(P,T)+12vns2∂(ρn/ρ)∂T(6)s(P,T,v_{ns})=s^{(s)}(P,T)+\frac{1}{2}v_{ns}^2\frac{\partial(\rho_n/\rho)}{\partial T} \tag*{(6)}
ρ(P,T,vns)=ρ(s)(P,T)+12ρ2vns2∂(ρn/ρ)∂P(7)\rho(P,T,v_{ns})=\rho^{(s)}(P,T)+\frac{1}{2}\rho^2v_{ns}^2\frac{\partial(\rho_n/\rho)}{\partial P} \tag*{(7)}

where all the static properties (i.e., with superscript (s)(s)) are extracted from the Hepak dynamic library [26]. Typically, the corrections amount to no more than a few percent of the property values in static He II.

The Gorter-Mellink mutual friction Fns\mathbf{F}_{ns} per unit fluid volume depends on the vortex-line density LL (i.e., length of vortices per unit volume) and vns\mathbf{v}_{ns} as [27]:

Fns=κ3ρsρnρBLLvns(8)\mathbf{F}_{ns}=\frac{\kappa}{3}\frac{\rho_s\rho_n}{\rho}B_L L\mathbf{v}_{ns} \tag*{(8)}

where BLB_L is a temperature-dependent mutual friction coefficient [28]. To determine Fns\mathbf{F}_{ns}, we need to know the spatial and temporal evolution of L(r,t)L(r,t). For this purpose, we adopt the Vinen’s equation [5]:

∂L∂t+∇⋅(vLL)=αV∣vns∣L3/2−βVL2+γV∣vns∣5/2(9)\frac{\partial L}{\partial t}+\nabla\cdot(\mathbf{v}_L L)=\alpha_V|\mathbf{v}_{ns}|L^{3/2}-\beta_VL^2+\gamma_V|\mathbf{v}_{ns}|^{5/2} \tag*{(9)}

where αV\alpha_V, βV\beta_V, and γV\gamma_V are temperature-dependent phenomenological coefficients [5]. We use the values recommended by Kondaurova et al. for these coefficients since they give simulation results that agree well with experimental observations [29]. The term ∇⋅(vLL)\nabla\cdot(\mathbf{v}_L L) accounts for the drifting of the vortices [30, 31], where we take the drift velocity vL\mathbf{v}_L to be the local superfluid velocity vs\mathbf{v}_s [5]. Nonetheless, we find that this drifting effect is negligible in highly transient heat transfer processes [12]. The above model represents a coarse-grained description of the two-fluid hydrodynamics, since the action of individual vortices on the normal fluid is not resolved [32, 33]. Nonetheless, when the vortex-line density is relatively high, this model was shown to describe non-isothermal flows in He II very well [12, 34].

In our study of transient heat transfer from a planar heater in He II, we consider a 1D computational domain with the heater located at x=0x=0 as shown in Figure 1 (a). To generate a rectangular heat pulse of duration tht_h with a surface heat flux qhq_h, we adopt a time-dependent boundary condition on the heater surface: vn=qh/ρsTv_n=q_h/\rho sT for the normal fluid and vs=−vnρn/ρsv_s=-v_n\rho_n/\rho_s for the superfluid during 0<t≤th0<t\leq t_h; and vn=vs=0v_n=v_s=0 at t>tht>t_h. An initial vortex-line density of L0=102 cm−2L_0=10^2\ \mathrm{cm}^{-2} is also assumed, which is comparable to typical densities of remnant vortices in He II containers [35]. Indeed, the exact value of L0L_0 does not affect the simulation results for the range of qhq_h considered in our work [12]. We evolve Eqs. (1)–(9) using the MacCormack’s predictor-corrector scheme with a spatial step Δx=10−5 m\Delta x=10^{-5}\ \mathrm{m} and a time step Δt=10−8 s\Delta t=10^{-8}\ \mathrm{s}, and we also adopt a flux-corrected transport method to suppress numerical instabilities [36].

Image not extracted

Figure 1. (a) A schematic diagram showing the transient heat transfer process from a planar heater located at x=0x = 0. (b) Simulated curves showing the temperature increment ΔT=T−Tb\Delta T = T - T_b in He II at various times when a heat pulse of qh=30q_h = 30 W/cm2^2 with a duration th=0.295t_h = 0.295 ms is applied. The bath temperature is Tb=1.78T_b = 1.78 K, and a He II depth H=1H = 1 m is assumed. (c) The spatial profiles of ΔT\Delta T near the heater at various tt.

To illustrate the key features of transient heat transfer in He II, we show in Figure 1 (b) the simulated spatial profiles of the temperature increment ΔT=T−Tb\Delta T = T - T_b at various tt for a representative case with qh=30q_h = 30 W/cm2^2 and th=0.295t_h = 0.295 ms. This simulation was performed at a bath temperature Tb=1.78T_b = 1.78 K with a He II depth H=1H = 1 m. One can see that a second-sound pulse featured by a positive ΔT\Delta T emerges in He II, which carries the heat energy and propagates away from the heater surface at the known second-sound speed (i.e., c2=19.6c_2 = 19.6 m/s at 1.78 K [28]). Near the heater surface where the vortex-line density LL grows rapidly, the interaction between the vortices and the second-sound pulse can convert the thermal energy carried by the pulse to locally deposited heat. This process results in a heated layer of He II (i.e., the thermal layer) adjacent to the heater surface [12]. To make this thermal layer more visible, we plot the ΔT\Delta T profile near the heater in Figure 1 (c). As the heat pulse ends, ΔT\Delta T on the heater surface reaches the highest value. After that, the thermal layer spreads out diffusively [11] so that ΔT\Delta T on the heater surface decreases. This example clearly shows that the largest temperature increment in He II occurs near the heater surface by the end of the heat pulse. Therefore, in our subsequent study on the onset of boiling in He II, we shall focus on the state variation of the He II adjacent to the heater.

Peak heat flux calculation

For a given helium bath condition (i.e., TbT_b and HH), the peak heat flux q∗q^* depends on the heat-pulse duration tht_h. To determine the correlation between q∗q^* and tht_h, we adopt a method by scanning tht_h as illustrated in Figure 2 (a). This figure shows the evolution of the thermodynamic state (P,T)(P,T) of the He II at the first grid point x=Δxx = \Delta x for heat pulses with the same flux qh=30q_h = 30 W/cm2^2 but with different duration tht_h. All the curves start from the same initial state marked by the black open circle, i.e., T∣t=0=Tb=1.78T|_{t=0} = T_b = 1.78 K and P∣t=0=Pb=Ps(Tb)+ρgHP|_{t=0} = P_b = P_s(T_b) + \rho gH where Ps(Tb)P_s(T_b) is the saturation pressure at TbT_b and H=1H = 1 m. The He II states at the end of the heat pulses are marked by the filled circles of the respective colors. It is clear that when the heater turns on, there is a pressure drop followed by an increase of the He II temperature. As the heat pulse ends, the pressure spikes up in all the cases and the state curves evolve back to the starting point. Obviously, the He II state at the end of the heat pulse gets closer to the saturation line at larger tht_h. We consider that boiling occurs when the state curve reaches the saturation line. For the example shown in Figure 2 (a), boiling occurs at th=0.297t_h = 0.297 ms.

Evolution of the He II thermodynamics state and simulated peak heat flux

Figure 2. (a) Evolution of the He II thermodynamics state at x=Δxx = \Delta x when heat pulses of qh=30q_h = 30 W/cm2^2 with different duration tht_h are applied. The filled circles indicate the He II states at the end of the heat pulses. (b) The simulated curve showing the dependance of the peak heat flux q∗q^* and the pulse duration tht_h. The dashed line denotes that above a critical peak heat flux qc∗q_c^*, the onset time of boiling suddenly drops to the order of Δx/c2\Delta x/c_2.

By repeating the above analysis at various applied heat fluxes, we can determine the corresponding pulse durations beyond which boiling occurs. The results for Tb=1.78T_b = 1.78 K and H=1H = 1 m are collected in Figure 2 (b). We see that when tht_h is greater than about 10−410^{-4} s, q∗q^* increases with decreasing tht_h, which agrees with the trend reported in literature [15, 16, 17, 18, 19]. However, when the applied heat flux reaches a critical value qc∗≃55q_c^* \simeq55 W/cm2^2, we find surprisingly that the onset time of boiling suddenly jumps from about 10−410^{-4} s to an extremely small value. This value is found to be of the order Δx/c2\Delta x/c_2, i.e., the traveling time of the second-sound pulse to reach the first grid point. At heat fluxes higher than qc∗q_c^*, boiling always occurs on a similar time scale. This time scale Δx/c2\Delta x/c_2 suggests that the onset time for boiling would become arbitrarily small as one approaches the heater surface. However, in practice, the onset time will be limited by various factors such as the time it takes for vapor bubbles to grow on the heater surface, which is about a few microseconds for the bubbles to reach a radius of about 1010 μ\mum [37, 38]. The appearance of the critical peak heat flux qc∗q_c^* and the associated sudden drop of the onset time of boiling to the order of Δx/c2\Delta x/c_2 are previously unreported phenomena, which indicates the existence of an unusual mechanism of boiling in He II.

Before moving to the next section to present our systematic study of qc∗q_c^*, we would like to compare our model simulations with some available experimental data. In Figure 3, we show the experimental data of q∗q^* versus tht_h obtained by Tsoi and Lutset [15] and by Shimazaki et al [18]. The work of Tsoi and Lutset adopted a thin-film nichrome heater (surface area: 3×33 \times3 cm2^2) immersed in He II at 1.7941.794 K, and the boiling was detected by monitoring the pressure change in He II using a piezosensor. The experiment of Shimazaki et al. was conducted at Tb=2T_b = 2 K and utilized a slightly smaller heater (area: 2.7×2.72.7 \times2.7 cm2^2), where the boiling was detected by measuring thermal shockwaves using a superconducting temperature sensor. The exact hydrostatic head pressures in these experiments were not reported. Nonetheless, we can perform simulations at the corresponding TbT_b with a range of HH compatible with the expected He II depths estimated based on their setup schematics. The simu-

Comparison of simulated peak heat flux and experimental data

Figure 3. Comparison of the simulated peak heat flux q∗q^* and some relevant experimental data. The blue circles and squares are data obtained by Tsoi and Lutset at 1.794 K [15]. The open and filled circles are data obtained by Shimazaki et al. at 2 K [18], where the open circles indicate possible onset of the boiling while the filled circles denote firm observation of the boiling. The purple and the green curves are our simulation at Tb=1.794T_b = 1.794 K with H=1H = 1 m and Tb=2T_b = 2 K with H=0.3H = 0.3 m, respectively. The purple band indicates the span of the curve when HH is varied in the range of 0.8−1.50.8 - 1.5 m, while the green band is for HH in the range of 0.2−0.40.2 - 0.4 m.

lated q∗−thq^*-t_h curves are shown in Figure 3, which agree well with these experimental data. This agreement validates our model. Note that due to the limited sensor response times, the sudden drop of the onset time of boiling at heat fluxes above qc∗q_c^* was not resolvable in these experiments.

Critical peak heat flux variation

In order to understand the physical mechanism underlying the critical peak heat flux qc∗q_c^*, we need to first conduct a systematic study on its dependence on the helium bath condition. For this purpose, we have repeated the afore-mentioned analysis at various TbT_b and HH. In Figure 4 (a), we show the calculated q∗−thq^*-t_h curves for TbT_b in the range of 1.3−2.11.3 - 2.1 K with a fixed He II depth of H=0.5H = 0.5 m. The critical peak heat flux qc∗q_c^* at each TbT_b is identified and marked by the filled circle. From this study, the dependence of qc∗q_c^* on TbT_b is obtained, which is shown in Figure 4 (b). qc∗q_c^* first increases with increasing TbT_b before reaching a maximum at Tb≃1.95T_b \simeq1.95 K. Then, qc∗q_c^* decreases as TbT_b further increases. The maximum value of qc∗q_c^* appears to be achieved at the bath temperature where the two fluids have about the same densities.

Image not extracted

(a)

Image not extracted

(b)

Figure 4. (a) Calculated q∗q^*-tht_h curves at various TbT_b with a fixed He II depth H=0.5 mH=0.5\ \mathrm{m}. The critical peak heat flux qc∗q_c^* for each curve is marked by the filled circle of the respective color. (b) The obtained critical peak heat flux qc∗q_c^* as a function of TbT_b. The black dots are simulation data. The solid and the dashed curves are calculated using Eq. (13) with and without the ΔP1\Delta P_1 term, respectively.

When the He II depth HH is changed, the dependence of qc∗q_c^* on TbT_b remains similar to that presented in Figure 4 (b) but its exact value changes. To illustrate how qc∗q_c^* varies with HH quantitatively, we fix the bath temperature at Tb=1.78T_b = 1.78 K and calculate the q∗−thq^*-t_h curves at various He II depth HH. Representative results for HH in the range of 0.3−2 m0.3-2\ \mathrm{m} are shown in Figure 5a, where qc∗q_c^* can be determined (marked by the filled circles). The obtained qc∗q_c^* is then plotted as a function of HH in Figure 5b. It is clear that qc∗q_c^* increases monotonically with increasing HH.

Explanation of Critical Peak Heat Flux

The studies presented in the previous sections show that the critical peak heat flux qc∗q_c^* depends on both TbT_b and HH and is likely associated with the propagation of the second-sound pulse since the corresponding boiling time is on the order of Δx/c2\Delta x/c_2. To better understand the physical processes that controls qc∗q_c^*, we show in Figure 6(a) the evolution of the He II state adjacent to the heater (i.e., x=Δxx=\Delta x) when the applied heat flux gradually increases.

Image not extracted

Figure 6. (a) Evolution of the He II state at x=Δxx=\Delta x when heat pulses of different qhq_h are applied. The asterisk in each curve denotes the end state of the fast process, and the filled circle marks where the state curve reaches the saturation line. The inset shows the obtained correlation between qc∗q_c^* and tht_h. (b) Time evolution of the He II pressure PP, temperature TT, and the vortex-line density LL at x=Δxx=\Delta x, 1 mm1\ \mathrm{mm}, and 2 mm2\ \mathrm{mm} for the case with qh=50 W/cm2q_h=50\ \mathrm{W/cm^2}. All the simulations are conducted at Tb=1.78 KT_b=1.78\ \mathrm{K} and H=1 mH=1\ \mathrm{m}.

All the cases start from the same initial state as in Figure 2(a), i.e., T=Tb=1.78 KT=T_b=1.78\ \mathrm{K} and P=Ps(Tb)+ρgHP=P_s(T_b)+\rho gH with H=1 mH=1\ \mathrm{m}. In what follows, we present a few important features observed in this study.

First, when the heater turns on, there is a fast process during which the pressure drops by ΔP\Delta P and the temperature increases by ΔT\Delta T. This process occurs on a time scale of Δx/c2\Delta x/c_2. The end state of this fast process is marked by the asterisk for each curve in Figure 6(a). It is clear that the magnitudes of both ΔP\Delta P and ΔT\Delta T increase with increasing the heat flux qq. To provide a more direct view of the fast process, we show the time evolution of the He II pressure, temperature, and the vortex-line density at x=Δxx=\Delta x, 1 mm1\ \mathrm{mm}, and 2 mm2\ \mathrm{mm} for a representative case with qh=50 W/cm2q_h=50\ \mathrm{W/cm^2} in Figure 6(b). For the x=Δxx=\Delta x curve, following the fast process the He II temperature and pressure remain nearly constant for over 0.1 s0.1\ \mathrm{s}. During this period, the vortex-line density LL gradually grows following an initial rapid increase of LL is controlled by the generation term (i.e., the last term) in Eq. (9), which is also the reason why the exact value of the initial line density L0L_0 does not affect the simulation result. When LL builds up to a sufficient level (i.e., of the order 108 cm−210^8\ \mathrm{cm^{-2}}), local heating due to the mutual friction makes the He II temperature rise again. This heating process drives the He II state at x=Δxx=\Delta x further towards the saturation line. Interestingly, at the critical peak heat flux qc∗≃55 W/cm2q_c^*\simeq55\ \mathrm{W/cm^2}, the state curve of the He II at x=Δxx=\Delta x reaches the saturation line during the fast process without involving any subsequent slower heating process. Therefore, the onset time for boiling suddenly drops to the order of Δx/c2\Delta x/c_2 (see the inset in Figure 6(a)). At heat fluxes higher than qc∗q_c^*, the boiling is largely controlled by the sudden drop in pressure across the saturation line, a phenomenon that is known as cavitation [39].

Based on the physical picture presented above, we can indeed develop an analytical model to evaluate qc∗q_c^*. When the heater turns on, a second-sound pulse emerges from the heater surface. At short times when the vortex-line density near the heater is relatively low, the temperature increment ΔT\Delta T within the second-sound pulse zone is related to the applied heat flux qhq_h as [2]:

qh=c2T(ρs∣2nd−ρs∣bath)≃c2ρCpΔT+12vns2c2T(∂ρn∂T)Tb+ΔT,Pb(10)q_h=c_2T(\rho s|_{2nd}-\rho s|_{bath}) \simeq c_2\rho C_p\Delta T+\frac{1}{2}v_{ns}^2c_2T\left(\frac{\partial\rho_n}{\partial T}\right)_{T_b+\Delta T,P_b} \tag*{(10)}

where the subscripts “2nd” and “bath” denote the parameters evaluated in the second-sound pulse zone and in the He II bath, respectively. Cp=T(∂s/∂T)PC_p=T(\partial s/\partial T)_P is the specific heat of He II, and the second term in the above equation comes from the correction term in Eq. (6). Within the second-sound pulse zone, a counterflow of the two fluids establishes as shown schematically in Figure 7. The velocities of the two fluids are given by vn=qh/(ρsT)∣Tb+ΔTv_n=q_h/(\rho s T)|_{T_b+\Delta T} and vs=−vnρn/ρsv_s=-v_n\rho_n/\rho_s. Accordingly, due to Eq. (2), the finite vnv_n and vsv_s lead to a pressure change ΔP2\Delta P_2 in the second-sound pulse zone as given by:

Image not extracted

Figure 7. A schematic diagram showing the temperature and pressure changes as well as the motion of the two fluids in the first-sound and the second-sound pulse zones.

ΔP2=−(ρsvs2+ρnvn2)≃−qh2ρns2T2ρs∣Tb+ΔT,Pb.(11)\Delta P_2=-(\rho_s v_s^2+\rho_n v_n^2)\simeq-\left.\frac{q_h^2\rho_n}{s^2T^2\rho_s}\right|_{T_b+\Delta T,P_b}. \tag*{(11)}

This pressure drop is essentially a manifestation of the Bernoulli effect due to the motion of the two fluids in the second-sound zone.

Besides the effects due to the second-sound pulse, there is another subtle effect. Note that He II has a negative thermal expansion coefficient at temperatures above about 1.1 K1.1\ \mathrm{K} [28]. Therefore, the He II density ρ\rho in the second-sound zone must increase due to the temperature rise ΔT\Delta T, which requires a mass flow towards this region. To supply the mass, a first-sound pulse is generated where the two fluids move in phase at a velocity v\mathbf{v} towards the second-sound wavefront, as shown in Figure 7. The mass flux ρv\rho\mathbf{v} should balance the needed mass associated with the expansion of the second-sound pulse zone, i.e., ρ∣v∣=c2(ρ∣2nd−ρ∣1st)\rho|\mathbf{v}|=c_2(\rho|_{2nd}-\rho|_{1st}). This finite ρv\rho\mathbf{v} leads to a pressure drop ΔP1\Delta P_1 in the first-sound pulse zone. To the lowest order in v\mathbf{v}, one can derive ΔP1\Delta P_1 from Eq. (2) as:

ΔP1≃−ρc1∣v∣=−c1c2(ρ∣2nd−ρ∣1st)≃−c1c2[(ρ2nd(s)−ρ1st(s))+ρ2vn22(∂ρn/ρ∂P)∣2nd](12)\begin{aligned} \Delta P_1\simeq-\rho c_1|\mathbf{v}|=-c_1c_2(\rho|_{2nd}-\rho|_{1st}) \\ &\simeq-c_1c_2\left[\left(\rho_{2nd}^{(s)}-\rho_{1st}^{(s)}\right)+\left.\frac{\rho^2v_n^2}{2}\left(\frac{\partial\rho_n/\rho}{\partial P}\right)\right|_{2nd}\right] \tag*{(12)} \end{aligned}

where c1c_1 is the speed of the first sound in He II (i.e., about 231 m/s231\ \mathrm{m/s} at 1.78 K1.78\ \mathrm{K}). The pressure drop in the first-sound zone generates a pulling force exerting on the wavefront of the second-sound zone, which results in a total pressure drop in the second-sound zone as ΔP=ΔP1+ΔP2\Delta P = \Delta P_1 + \Delta P_2. This effect can be clearly observed in Figure 6 (b) at x=1x = 1 mm (and x=2x = 2 mm), since the first-sound pulse and the second-sound pulse arrive at different times.

At the critical peak heat flux qc∗q_c^*, the temperature increment ΔT\Delta T and the total pressure drop ΔP\Delta P associated with the second-sound pulse would drive the He II at x=Δxx = \Delta x from the initial state (Tb,Pb)(T_b,P_b) to the saturation line upon its arrival. Therefore, the following equation must hold:

Ps(Tb)+ρgH+ΔP1+ΔP2=Ps(Tb+ΔT),(13)P_s(T_b) + \rho g H + \Delta P_1 + \Delta P_2 = P_s(T_b + \Delta T), \tag*{(13)}

where qc∗q_c^* enters the equation through ΔT\Delta T, ΔP1\Delta P_1, and ΔP2\Delta P_2 via Eqs. (10)–(12). Using this model, we have calculated qc∗q_c^* as a function TbT_b at H=0.5H = 0.5 m and as a function of HH at Tb=1.78T_b = 1.78 K. The results are included in Figure 4 (b) and Figure 5 (b) as the solid curves. Excellent agreement between the model’s calculation and the simulated qc∗q_c^* values is observed, which proves that our understanding of the mechanism underlying the critical peak heat flux is correct. To see how large the first-sound effect is, we have also repeated the calculation using Eq. (13) but without the ΔP1\Delta P_1 term. The results are shown as the dashed

Image not extracted

(a)

Image not extracted

(b)

Figure 5. (a) Calculated q∗q^*-tht_h curves at various HH with a fixed bath temperature Tb=1.78 KT_b=1.78\ \mathrm{K}. The critical peak heat flux qc∗q_c^* for each curve is marked by the filled circle of the respective color. (b) The obtained critical peak heat flux qc∗q_c^* as a function of HH. The black dots are simulation data. The solid and the dashed curves are calculated using Eq. (13) with and without the ΔP1\Delta P_1 term, respectively.

curves in Figure 4 (b) and Figure 5 (b). Obviously, the first-sound effect is non-negligible at these large heat fluxes.

Summary

We have conducted a numerical study on 1D transient heat transfer in He II from a planar heater. The peak heat flux q∗q^* for the onset of boiling in He II is determined as a function of the heat-pulse duration tht_h. A major finding in our study is the observation of a critical peak heat flux qc∗q_c^* above which boiling occurs almost instantaneously. Our analysis shows that the boiling at heat fluxes lower than qc∗q_c^* is caused by a heating process, which is associated with the relatively slow buildup of the quantized vortices and the thermal layer in front of the heater. When the applied heat flux is higher than qc∗q_c^*, the boiling is essentially a cavitation on the heater surface due to the combined effects of the first-sound and the second-sound waves in He II. A theoretical model for evaluating qc∗q_c^* is developed, which nicely reproduces the simulated qc∗q_c^* values at various He II bath temperatures and hydrostatic head pressures. Inspired by this work, a question we plan to address next is how the boiling physics may change in non-homogeneous heat transfer in He II. As shown in our early work [12], heat transfer of He II in non-homogeneous geometries (such as from cylindrical or spherical heaters) can exhibit new features. Understanding the behavior of the peak heat flux in these geometries could benefit research work such as quench-spot detection on He II cooled superconducting accelerator cavities [40, 41] and the heat and mass transfer processes due to a vacuum failure in He II cooled accelerator beamline tubes [42, 43, 44, 45].

Acknowledgments

The authors would like to acknowledge the support from the U.S. Department of Energy under Grant No. DE-SC0020113. The work was conducted at the National High Magnetic Field Laboratory at Florida State University, which is supported through the National Science Foundation Cooperative Agreement No. DMR-1644779 and the state of Florida.

References

  1. [1]D. R. Tilley and J. Tilley, Superfluidity and Superconductivity, 2nd ed., Graduate Student Series in Physics (Adam Hilger Ltd, Bristol, 1986).
  2. [2]L. D. Landau and E. M. Lifshitz, Fluid Mechanics, 2nd ed., Vol. 6 (Pergamon Press, Oxford, 1987).
  3. [3]W. F. Vinen, Mutual friction in a heat current in liquid helium II. I. experiments on steady heat currents, Proc. R. Soc. Lond. A 240, 114 (1957).
  4. [4]R. J. Donnelly, Quantized Vortices in Helium II, Vol. 2 (Cambridge University Press, Cambridge, 1991).
  5. [5]W. F. Vinen, Mutual friction in a heat current in liquid helium II. II. experiments on transient effects, Proc. R. Soc. A 240, 128 (1957).
  6. [6]A. Marakov, J. Gao, W. Guo, S. W. Van Sciver, G. G. Ihas, D. N. McKinsey, and W. F. Vinen, Visualization of the normal-fluid turbulence in counterflowing superfluid ⁴He, Phys. Rev. B 91, 094503 (2015).DOI
  7. [7]J. Gao, E. Varga, W. Guo, and W. F. Vinen, Energy spectrum of thermal counterflow turbulence in superfluid helium-4, Phys. Rev. B 96, 094511 (2017).
  8. [8]J. Gao, W. Guo, S. Yui, M. Tsubota, and W. F. Vinen, Dissipation in quantum turbulence in superfluid ⁴He above 1 K, Phys. Rev. B 97, 184518 (2018).
  9. [9]S. Bao, W. Guo, V. S. L’vov, and A. Pomyalov, Statistics of turbulence and intermittency enhancement in superfluid ⁴He counterflow, Phys. Rev. B 98, 174509 (2018).DOI
  10. [10]B. Mastracci and W. Guo, Exploration of thermal counterflow in He II using particle tracking velocimetry, Phys. Rev. Fluids 3, 063304 (2018).
  11. [11]S. W. Van Sciver, Helium Cryogenics, 2nd ed., International Cryogenics Monograph Series (Springer, New York, 2012).
  12. [12]S. Bao and W. Guo, Transient heat transfer of superfluid ⁴He in nonhomogeneous geometries: Second sound, rarefaction, and thermal layer, Phys. Rev. B 103, 134510 (2021).
  13. [13]H. Kobayashi and K. Yasukōchi, Maximum and minimum heat flux and temperature fluctuation in film-boiling states in superfluid helium, in Advances in Cryogenic Engineering (Springer, 1980) pp. 372–377.DOI
  14. [14]S. K. Nemirovskii and A. N. Tsoi, Transient thermal and hydrodynamic processes in superfluid helium, Cryogenics 29, 985 (1989).DOI
  15. [15]A. Tsoi, Boiling up of helium I, helium II, and nitrogen during transient heat release, Inzhenerno-Fizicheskii Zhurnal 51, 5 (1985).DOI
  16. [16]B. A. Danilchenko, M. O. Lutset, and V. N. Poroshin, Limit of transient heat absorption by superfluid helium for very large heat pulses, Cryogenics 29, 444 (1989).DOI
  17. [17]R. Wang, Criterion for quantum turbulence onset after rectangular heat pulse in superfluid helium, Cryogenics 35, 883 (1995).DOI
  18. [18]T. Shimazaki, M. Murakami, and T. Kanari, Measurement of characteristic time for quantized vortex tangle development in He II, Cryogenics 38, 601 (1998).DOI
  19. [19]L. Kondaurova, V. Efimov, and A. Tsoi, Influence of quantum turbulence on the processes of heat transfer and boiling in superfluid helium, J. Low Temp. Phys. 187, 80 (2017).DOI
  20. [20]L. P. Kondaurova, Dynamics of vortex line density and heat transfer processes in superfluid helium, Low Temp. Phys. 44, 29 (2018).DOI
  21. [21]E. Latimer, D. Spence, C. Feng, A. Boatwright, A. M. Ellis, and S. Yang, Preparation of ultrathin nanowires using superfluid helium droplets, Nano Lett. 14, 2902 (2014).DOI
  22. [22]E. B. Gordon, M. E. Stepanov, M. I. Kulish, A. V. Karabulin, V. I. Matyushenko, and I. I. Khodos, The nanowires growth by laser ablation of metals inside rotating superfluid helium, Laser Phys. Lett. 16, 026002 (2018).DOI
  23. [23]S. K. Nemirovskii and W. Fiszdzon, Chaotic quantized vortices and hydrodynamic processes in superfluid helium, Rev. Mod. Phys. 67, 37 (1995).DOI
  24. [24]S. K. Nemirovskii, On the closure problem of the coarse-grained hydrodynamics of turbulent superfluids, J. Low Temp. Phys. 201, 254 (2020).DOI
  25. [25]I. M. Khalatnikov, An Introduction to the Theory of Superfluidity, 1st ed. (CRC Press, 2000).
  26. [26]V. D. Arp, R. D. McCarty, and B. A. Hands, HEPAK - Thermophysical properties of helium from 0.8K or the melting line to 1500K, Cryodata Inc. (2005).
  27. [27]W. F. Vinen, Mutual friction in a heat current in liquid helium II. III. Theory of the mutual friction, Proc. R. Soc. Lond. A 242, 493 (1957).DOI
  28. [28]R. J. Donnelly and C. F. Barenghi, The observed properties of liquid helium at the saturated vapor pressure, J. Phys. Chem. Ref. Data 27, 1217 (1998).DOI
  29. [29]L. Kondaurova, V. Efimov, A. N. Tsoi, Influence of quantum turbulence on the processes of heat transfer and boiling in superfluid helium, J. Low Temp. Phys. 187, 80 (2017).DOI
  30. [30]K. W. Schwarz, Three-dimensional vortex dynamics in superfluid ⁴He: Homogeneous superfluid turbulence, Phys. Rev. B 38, 2398 (1988).pubmed.ncbi.nlm.nih.gov/9946544
  31. [31]S. K. Nemirovskii, Macroscopic dynamics of superfluid turbulence, Low Temp. Phys. 45, 841 (2019).DOI
  32. [32]S. Yui, H. Kobayashi, M. Tsubota, and W. Guo, Fully coupled two-fluid dynamics in superfluid ⁴He: Anomalous anisotropic velocity fluctuations in counterflow, Phys. Rev. Lett. 124, 155301 (2020).DOI
  33. [33]B. Mastracci, S. Bao, W. Guo, and W. F. Vinen, Particle tracking velocimetry applied to thermal counterflow in superfluid ⁴He: Motion of the normal fluid at small heat fluxes, Phys. Rev. Fluids 4, 083305 (2019).DOI
  34. [34]Y. A. Sergeev and C. F. Barenghi, Turbulent radial thermal counterflow in the framework of the HVBK model, EPL 128, 26001 (2019).
  35. [35]D. D. Awschalom and K. W. Schwarz, Observation of a remanent vortex-line density in superfluid helium, Phys. Rev. Lett. 52, 49 (1984).DOI
  36. [36]C. A. J. Fletcher, Specific Techniques for Different Flow Categories, 2nd ed., Computational Techniques for Fluid Dynamics, Vol. 2 (Springer, Berlin, 2003).
  37. [37]W. Guo and H. J. Maris, Observations of the motion of single electrons in liquid helium, J. Low Temp. Phys. 148, 199 (2007).DOI
  38. [38]W. Guo, D. Jin, G. M. Seidel, and H. J. Maris, Experiments with single electrons in liquid helium, Phys. Rev. B 79, 054515 (2009).
  39. [39]C. E. Brennen, Cavitation and Bubble Dynamics (Cambridge University Press, Cambridge, 2013).
  40. [40]S. Bao and W. Guo, Quench-spot detection for superconducting accelerator cavities via flow visualization in superfluid helium-4, Phys. Rev. Appl. 11, 044003 (2019).DOI
  41. [41]S. Bao, T. Kanai, Y. Zhang, L. N. Cattafesta, and W. Guo, Stereoscopic detection of hot spots in superfluid ⁴He (He II) for accelerator-cavity diagnosis, Int. J. Heat Mass Transf. 161, 120259 (2020).DOI
  42. [42]N. Garceau, S. Bao, W. Guo, and S. Van Sciver, The design and testing of a liquid helium cooled tube system for simulating sudden vacuum loss in particle accelerators, Cryogenics 100, 92 (2019).DOI
  43. [43]N. Garceau, S. Bao, and W. Guo, Heat and mass transfer during a sudden loss of vacuum in a liquid helium cooled tube - Part I: Interpretation of experimental observations, Int. J. Heat Mass Transf. 129, 1144 (2019).DOI
  44. [44]S. Bao, N. Garceau, and W. Guo, Heat and mass transfer during a sudden loss of vacuum in a liquid helium cooled tube - Part II: Theoretical modeling, Int. J. Heat Mass Transf. 146, 118883 (2020).DOI
  45. [45]N. Garceau, S. Bao, and W. Guo, Heat and mass transfer during a sudden loss of vacuum in a liquid helium cooled tube - Part III: Heat deposition in He II, Int. J. Heat Mass Transf. 181, 121885 (2021).DOI

Paper details

vlnc:W_8cyqe6r9
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents