6.2. Bubble velocity correlations¶
See also
The discussion here is based on order-of-magnitudes of forces relevant to bubble motion [TH18]. This approach is simple and easy to follow. However, for readers those who want to see derivation in which the relation with the jump condition is stressed, see also
In the terminal state of a bubble in still liquid, the force balance, as first order expression, may be written as
The \(c\)s are constants. Let us first consider a bubble of very small size. Considering the limiting case of \(d \rightarrow 0\), we may also have \(V_{T} \rightarrow 0\). Therefore, \(F_{\sigma} \gg F_{i}, F_{\mu}, F_{b}\). Because of the strong action of surface tension, the bubble maintain spherical shape. The distribution of surface tension force acting on the spherical surface is symmetric, so that the surface tension force cancels out when integrating the distribution for the entire surface. Thus, for tiny bubbles, the surface tension force keeps their shapes spherical, but plays no role in the rise velocity. Omitting \(F_{\sigma}\) from the force balance, we have
The remaining forces show the dependence: \(F_{b} \propto d^{3}\), \(F_{\mu} \propto V_{T} d\), and \(F_{i} \propto V_{T}^{2} d^{2}\). By a detailed analysis we will discuss later, \(V_{T} \propto d^{2}\) for small bubble sizes, so that \(F_{\mu} \propto d^{3}\) and \(F_{i} \propto d^{4}\). Thus, \(F_{b}\) and \(F_{\mu}\) are comparable, but \(F_{i}\) is smaller than the first two, leading to
from which we obtain
The rise velocity is therefore
where \(c_{1} = c_{b} / c_{\mu}\). This equation shows that the increase in buoyancy makes a bubble faster, but the viscosity retards the bubble. Dividing this equation with the velocity scale of momentum diffusion, \(\mu_{L} / \rho_{L} d\) [m/s], yields the following non-dimensional expression of bubble velocity:
Substituting this result into Eq. (6.12) yields
where \(c_{St} = 4 c_{1} / 3\). The drag coefficient is therefore inversely proportional to \(Re\). The constant, \(c_{St}\), can be analytically obtained for \(Re \ll 1\) (the Stokes regime) as \(c_{St} = 16\). See Drag acting on a solid sphere in Stokes flow (LL) and Drag acting on a fluid sphere in Stokes flow for rigorous derivation. On the other hand, in the limiting case of \(Re \rightarrow \infty\) for spherical bubble, the liquid flow can be considered as potential flow, but the work done by drag must have a finite value and balances with the viscous dissipation (Viscous dissipation). That is,
where \(V\) is the volume of the liquid phase, and \(e_{ij}\) is the rate of strain tensor defined by
and \(v_{i}\) is the \(i\)th component of the liquid velocity. The order of magnitude of the velocity gradient may be estimated as \(V_{T} / d\). Therefore, the order of \(e_{ij}\) is \(V_{T}/d\). The viscous dissipation may take place in the vicinity of the bubble due to the high \(Re\), so that \(V\) can be replaced with \(d^{3}\). Thus,
Substituting \(F_{D} = \frac{C_{D}}{2} \rho_{L} V_{T}^{2} \frac{\pi d^{2}}{4}\) into the L.H.S. yields
By putting constants into a single value \(c_{L}\), we obtain
where \(c_{L} = 48\). The derivation of this case is given in Drag acting on a spherical bubble at high Re. In summary, for bubbles in the viscous force dominant regime,
Bubbles cannot maintain their spherical shape as \(d\) increases. The shape of a large bubble may be no longer neither spherical nor ellipsoidal, but a slice of a sphere called a spherical cap. Because of its large size, the inertial and buoyancy forces are dominant rather than the viscous and surface tension forces. Therefore,
Hence,
This functional form is similar to that of the phase velocity of gravitational water wave. In dimensionless form, the bubble velocity can be written as
where \(c_{T} = \sqrt{c_{b}/c_{i}}\). Hence, the drag coefficient is given by
The Bernoulli theorem (see Large bubbles in liquid) gives \(c_{T} = 1/\sqrt{2}\), so that \(C_{D} = 8/3\). Using the relationship \(Fr^{2} = We/ Eo\) gives an alternative dimensionless form:
Consider bubbles of intermediate sizes. They cannot maintain spherical shape but may be ellipsoidal or distorted ellipsoidal. Assuming that the viscous contribution to the drag is negligible compared to the other forces, we have
First, we assume that the bubble rise motion is governed by the inertial and surface tension forces. In this case,
Hence,
The phase velocity of capillary water wave has a similar functional form. Interestingly, in this limiting case, the bubble rise velocity decreases with increasing bubble size. Nondimensionalizing this equation yields
where \(c_{M} = c_{\sigma} / c_{i}\) and \(c_{M}\) will be found to be \(2\) in Wave analogy for bubble rise velocity. If the inertial and buoyancy forces are competitive, we combine Eqs. (6.28) and (6.5) to obtain
With help of Eq. (6.12) (the most right equation),
This is valid for bubbles in the surface-tension and inertial force dominant regime.