It has been observed that the catch probability strongly
depends on initial particle depth, which can be investigated
in the future work.
In addition, it is recommended to introduce realistic
waves and buoyant motion of the boom, and to perform
transient simulations.
3.4.3 Boom angle modeling conclusions
The conclusions regarding catch probability obtained
during the 3D CFD modeling in both cases support the
findings of the 2D CFD models. The results do not match
perfectly, but within the range of the highest significance
(macroplastics), and partially for mesoplastics the results
of the 2D and 3D CFD studies are in agreement.
Since the 2D CFD study covered a wider range of simulation
setups, the results of the 3D CFD catch efficiency
were used for reference, benchmarking and comparison,
but not in the catch efficiency calculations.
The results related to the interaction of the flow with The
Array and the movement of plastic particles along The
Array are in good agreement between the two 3D CFD
studies. Even though the way of modeling this particular
part of the problem was different in the case of the
two studies, the results are similar. In the case of Comsol
modeling, the computational domain was increased significantly
in order to be able to capture particle velocities
in the far field away from the sink.
In the case of ANSYS CFX periodic boundary conditions
were implemented in order to measure the particle transportation
rate towards the sink.
The particle velocities in the direction parallel to The Array
for Array angles ranging from 5 to 35 degrees show
similar behavior and reach average values of 0.55 cm/s
– 4.4 cm/s (Comsol) and 1.0 cm/s – 4.7 cm/s (ANSYS CFX).
Both particle velocity ranges correspond to a current
velocity of 10 cm/s. The deviation in the results can be
explained by the different geometric setups used in the
simulations. The Comsol simulation uses a 100 m barrier
while the ANSYS simulation uses a quasi-infinite barrier
modeled by periodic boundary conditions. The length of
the barrier affects the current velocity that builds up in
front of the barrier.
The obtained values serve as a solid basis for defining the
Array angle and length. They also define the achievable
plastic collection rate as a function of the Array angle.
The results of the 3D CFD study strengthen the findings
of the 2D CFD investigation and confirm the weak points
about the uncertainties currently associated with setting
up CFD simulations of the particle collection efficiency
problems.