Circular pipe flow and particle image velocimetry (PIV).

Pressure drop reduction and heat transfer deterioration in circular pipe flow, and the PIV method.

技術情報1


 In the transportation of slush hydrogen or slush nitrogen via pipeline for use as the refrigerant for high-temperature superconducting (HTS) power transmission and equipment, it is important to consider pressure drop due to straight pipes, as well as flow restrictions such as valves, converging-diverging pipes, and orifices. Such pipelines also often make use of corrugated pipes to absorb the cryogenic contraction (stainless-steel pipe contracts by approximately 3 mm per 1 m). Increased pressure drop during slush fluid transport requires greater pump power, thus reducing the overall performance of the transport system.

 When slush nitrogen is to be used as the refrigerant for HTS power transmission, heat generation can be anticipated due to superconductor quench, making it important to also consider pressure drop of slush nitrogen during heat transfer, together with forced convection heat transfer to slush nitrogen.

 Flow and heat transfer experiments were performed with respect to slush nitrogen flow in a horizontal heat transfer pipe. From the experimental results, we firstly reported that the pressure drop reduction and heat transfer deterioration (i.e., pressure drop and heat transfer coefficient for liquid nitrogen are greater than those of slush nitrogen) emerged simultaneously during pipe flow, and the mechanisms involved in pressure drop reduction and heat transfer deterioration have been elucidated [20, 21].
 In addition to the quantitative determination of pressure drop reduction and heat transfer deterioration, We have also reported on the visualization of solid particle behavior by means of a high-speed video camera and flow velocity distribution of solid particles measured using the PIV method* as presented in the figure below, as well as the elucidation of the reduction of the reduction and deterioration mechanisms through flow analysis, employing a numerical analysis code (SLUSH-3D) that we developed.

 From our previous research results, it was found that in the pseudo-homogeneous flow of slush nitrogen (63 K) in horizontal circular pipes having diameters of 10 mm and 15 mm, at flow velocities of over 2.0 m/s and over 3.6 m/s, the maximum reductions in pressure drop were 25% and 23%, respectively, in comparison with liquid nitrogen (63 K). Also, we have reported that at heat flux of 10 kW/m2, the maximum pressure drop reductions were 23% and 17%, while at the same time the maximum deteriorations in heat transfer coefficient were 27% and 20% as shown in the above figure [20, 21].

 The mechanism involved in pressure drop reduction and heat transfer deterioration is summarized as follows:
 When pipe flow velocity increases and heterogeneous flow changes to pseudo-homogeneous flow in the case of heated or unheated flow, such that the solid particles move towards the center region of the pipe (refer to the figure in "Numerical analysis of slush flow by SLUSH-3D"), interference decreases between the solid particles and the pipe wall. Here, the pipe center region refers to the region inside the pipe excluding the thin liquid layer near the pipe wall in which there are few solid particles.
 This movement of solid particles to the center results in a grouping of particles that suppresses the development of turbulence at the pipe wall, thus resulting in pressure drop reduction. Although heat transfer takes place near the pipe wall where the liquid layer contains few solid particles, the comparative lack of turbulence in this region means that the heated liquid near the wall is prevented from dispersing towards the pipe center, causing heat transfer deterioration.

 When slush fluid is subjected to pumped transport over a long distance in combination with superconducting electrical power transmission, pressure drop reduction and heat transfer deterioration are extremely important from the standpoint of design data required to reduce the required power for pumps and to prevent quench propagation.

 In other research on pressure drop reduction in slush fluid, only one experimental result obtained by Sindt et al. has been reported indicating the occurrence of this phenomenon in slush hydrogen flowing in a circular pipe having an inner diameter of 16.6 mm [32]. Only the pressure drop of slush hydrogen was measured, and there was no attempt to explain the reduction or to consider the heat transfer properties.
 The relationship between the Reynolds number and the pipe friction factor for slush nitrogen flowing in a circular pipe with an inner diameter of 15 mm and slush hydrogen is shown in the figure below. The solid line represents the Prandtl-Karman eq. for a single-phase liquid.

 Pressure drop reductions for slush hydrogen and slush nitrogen appear for the Reynolds (Re) numbers of 1.0×105 and above, and as the solid fraction increases, the Re number at the onset of pressure drop reduction becomes greater.
 However, the velocity at which pressure drop reduction initiated with respect to slush hydrogen for an inner diameter of 16.6 mm was lower than that in the case of slush nitrogen for a 15 mm diameter pipe (nearly the same diameter) at identical solid fractions. That is, as indicated in the figure, for a solid fraction of 20 wt.%, the Re number at which the reduction initiates is 1.0×105 for slush hydrogen and 1.5×105 for slush nitrogen. Since the kinetic viscosities are about the same for both, the reduction initiates at lower flow velocity in the case of slush hydrogen. The solid-liquid density ratio is lower for slush hydrogen (1.12 vs. 1.18 for slush nitrogen), and, as the viscosity of liquid hydrogen is 1/11 that of liquid nitrogen, pseudo-homogeneous flow occurs at lower flow velocity. These results suggest that the pressure drop reduction effect is manifested in pseudo-homogeneous flow [20] (refer to the figure in "Numerical analysis of slush flow by SLUSH-3D").

 Taking unheated pressure drop test results for circular pipe, we applied the least squares approximation method to obtain a correlation equation between the slush Reynolds number ReSL using the apparent viscosity μSL = μL [1- (x/0.6)]-1.8 [33], and the pipe friction factor λ. Here, μL is liquid nitrogen viscosity. The correlation equation, as shown in the figure below, can accurately estimate the pipe friction factor from the flow velocity and volumetric solid fraction x (not mass solid fraction) [20].
 The correlation equation takes pressure drop reduction into account, and will be effective for design work when estimating required pumping power for the transport of slush nitrogen.

 * PIV method: The velocity of solid particles is directly measured using the Particle Image Velocimetry method.