H. Furukawa et al.
10.4236/wjm.2019.97012 179 World Journal of Mechanics
The governing equations are the unsteady three-dimensional incompressible
Navier-Stokes equation with cylindrical coordinates (
r
,
θ
,
z
) and the continuity
equation.
We use both SOR and ILUCGS methods to solve Poisson’s equation for pres-
sure. The stress-free boundary condition was used for the upper-end wall and
the stationary (non-slip) condition is used for the lower end wall. We applied
Neumann conditions based on the momentum equation for pressure. As the ini-
tial condition, all velocity components are zero. Mixed solution of water and
glycerin is assumed to be the working fluid, and its dynamic viscosity is 6.0 ×
10
–6
m
2
/s. For the discretization method, we apply the QUICK method for con-
vection terms, the second-order central difference method for the other space
integration, and Euler’s method for time integration. Grids are staggered and
equidistant in each direction. The number of grid points is 41 in the radial direc-
tion, and the number of grid points in the axial direction is proportionally ad-
justed so that it becomes 41 for the aspect ratio of 1.0. The number of grid points
in the circumferential direction is 74. In order to examine the validity of the
number of grid points, we analyzed Taylor vortex flow using several types of gr-
ids under various numerical conditions, and concluded that there are no differ-
ences among the modes that are finally formed, the formation of modes up to
the final mode, and the manner of decay of the vortexes.
3. Mode Analysis Method
In the preceding studies on the numerical analysis, the changes in kinetic energy
of axial velocity in every cell have been used to identify whether the flow state is
the Taylor vortex flow or wavy Taylor vortex flow. This identification is possible
because the Taylor vortex flow is defined to be the wavy Taylor vortex flow when
the Taylor vortex flow becomes non-linear and unstable due to the cyclic axial
oscillation. The criterion for this identification is the amplitude of the kinetic
energy changes, and the flow state is identified to be the wavy Taylor vortex flow
when the amplitude in the neighborhood of the estimated critical Reynolds
number reaches or exceeds this criterion. When the amplitude is less than (does
not reach) this criterion, the flow state is identified to be the Taylor vortex flow.
The mode is decided based on the amplitude when TS is 2,000,000 (at TS =
2,000,000). The critical Reynolds number when the flow transits from the Taylor
vortex flow to wavy Taylor vortex flow are decided. The computation for a
relatively short period of time when TS is 2,000,000 (at TS = 2,000,000) reveals
that the flow between inner and outer cylinders at around the critical Reynolds
number is unsteady, and the computation is not extended to the point where the
kinetic energy changes converge. Therefore, the flow state may have been de-
cided to be the wavy Taylor vortex flow based on the kinetic energy changes
even when the Taylor vortex flow is unsteady. The criterion of the energy am-
plitude is decided by the trend of energy amplitude changes in the neighborhood
of the estimated critical Reynolds number, and thus the wavy Taylor vortex flow
评论0
最新资源