Bubble layer

The same conclusion is evident from results obtained by Hino and Ueda (1975) and presented above in Fig. 6.4. The conclusion that A7s is almost unaffected by inlet flow velocity as at 7) -C 1 as at Z) < 1 was established from experiments carried out in the channels of diameters about d = 1—10 mm. What has been commonly observed at incipient boiling for subcooled flow in channels of this size is that small bubbles nucleate, grow and collapse while still attached to the wall, as a thin bubble layer formed along the channel wall. 

Figure 3.19 Configuration of bubble layer as affected by flow rate at high subcooling (Freon-118) (a) low-velocity boiling flow (b) high-velocity boiling flow. (From Tong et al., 1966b. Copyright 5 1966 by American Society of Mechanical Engineers, New York. Reprinted with permission.)...

In saturated annular flow boiling, on the other hand, a liquid film annulus normally covers the heating surface and acts as a cooling medium. Thinning of the liquid film annulus, therefore, indicates approaching dryout. The behavior of bubble layers and liquid annuli are of interest to visual observers. 

Two visual studies of Freon boiling crisis were conducted at the University of Pittsburgh (Lippert, 1971) and at Michigan University (Mattson et al., 1973). Both programs succeeded in identifying the DNB under the saw-shaped bubble layer of subcooled Freon flows as shown in Figures 5.8 and 5.9, respectively... 

The bubble behavior near the boiling crisis is three-dimensional. It is hard to show a three-dimensional view in side-view photography, because the camera is focused only on a lamination of the bubbly flow. Any bubbles behind this lamination will be fussy or even invisible on the photograph, but they can be seen by the naked eye and recorded in sketches as shown in Section 5.2.3. For further visual studies, the details inside bubble layers (such as the bubble layer in the vicinity of the CHF) would be required. Therefore, close-up photography normal and parallel to the heated surf ace is highly recommended. 

Figure 5.8 Saw-shaped bubble layer at DNB in a high-pressure, subcooled, vertical Freon flow p = 190 psig (1.4 MPa) G = 3.2 X 106 lbm/hr ft2 (4,320 kg/m2 s) A7 ub = 29°F (16°C) subcooled (qlA)t = 1.2 x 10s Btu/hr ft2 (3.8 x 105 W/m2). (From Dougall and Lippert, 1973. Reprinted with permission of NASA Scientific Technical Information, Linthicum Heights, MD.)...

The subcooled core bubble-layer liquid exchange and bubble condensation analysis was suggested by Tong (1975). 

Bankoff calculated Tx by using Gunter s experimental data and obtained the interesting result that, in each series of runs, Tx rises steeply toward the saturation temperature as burnout is approached. This gives a fairly thick bubble layer, which increases the degree of superheat near the wall. Bankoff concluded that burnout occurs when the core is unable to remove the heat as fast as it can be transmitted by the wall layer. ... 

We further consider that the boiling crisis occurs when the bubble-layer shielding effect reaches a maximum at bubble stagnation. A criterion for bubble-layer stagnation is suggested by Tong (1968b) to be... 

Several approaches to analyzing the thermal shielding effect of a bubble layer on CHF are presented here chronologically. 

The analysis of critical enthalpy in a bubble layer was suggested by Tong et al. (1966a). 

The analysis of turbulent mixing at the core-bubble layer interface was suggested by Weisman and Pei (1983). 

The analysis of mass and energy balance on the bubble layer was suggested by Chang and Lee (1989). 

These different approaches are complementary to each other in basic concept. However, these analyses have not provided clear insight information of the bubble layer at the CHF about the bubble shape (spherical or flat elliptical), bubble population and its effect on turbulent mixing, and bubble behavior. The bubble behavior in a bubble layer could involve bubble rotation caused by flow shear, normal bubble velocity fluctuation, and bubble condensation in the bubble layer caused by the subcooled water coming from the core. Further visual study and measurements in this area may be desired. 

From the simplified model of Figure 5.17, the energy equation for the bubble layer is written in terms of average conditions across the liquid layer ... 

T = average temperature of bubble layer Tb = temperature of bulk flow q" = surf ace heat flux... 

The bubble layer is assumed to have constant void fraction along the length before DNB, with a balanced rate of bubble detachment and bubble condensation in the layer. Hence, the average properties p, p, and c of the bubble layer are assumed to be independent of position. 

The thickness, s, and average velocity, V, of the bubble layer are approximately constant along the flow direction before DNB. The layer thickness includes the thin layer of superheated liquid in contact with the wall and is considered a homogenized inside layer. 

The average temperature of liquid in the bubble layer is approximately constant before DNB. 

The heat transfer coefficient h from the bubble layer to the bulk flow is constant before DNB. 

Postulating that the inception of DNB is determined by a limiting value of the enthalpy of the bubble layer, we can write... 

Figure 5.20 Radial flow interchange in bubble layer.

Weisman and Pei Model Weisman and Pei (1983) assumed that CHF occurs when the steam void fraction in the bubble layer just exceeds the critical void fraction (they estimated it to be 0.82) at which an array of ellipsodal bubbles can be maintained without significant contact between the bubbles. The steam void fraction in the bubble layer is determined by a balance between the outward flow of vapor and the inward flow of liquid at the bubble layer-core interface. 

By assuming no bubble layer mass gradient along the flow direction, z, a simplified radial mass balance equation for the bubble layer can be established according to Figure 5.20 as... 

The quantity G of the effective mixing mass flux is determined by the turbulent velocity fluctuations at the bubble-layer edge. The distance of the edge of the bubble layer from the wall is taken as the distance at which the size of the turbulent eddies is k times the average bubble diameter. Weisman and Pei have determined empirically that k equals 2.28. Only a fraction of the turbulent velocity fluctuations produced are assumed to be effective in reaching the wall. The effective velocity fluctuations are those in which the velocity exceeds the average velocity away from the wall produced by evaporation heat flux q"b. At the bubble layer-core interface, the effective mass flux to the wall is computed as... 

---