The numerical analysis to understand the characteristics of flow around a honeycomb topography in relation to biofouling control
Cite as: AIP Conference Proceedings 2137, 040002 (2019); https://doi.org/10.1063/1.5121000 Published Online: 07 August 2019
Leow Zyncoln, and Felicia Wong Yen Myan
ARTICLES YOU MAY BE INTERESTED IN
A review of transformative learning theory with regards to its potential application in engineering education
AIP Conference Proceedings 2137, 050001 (2019); https://doi.org/10.1063/1.5121003 A study of vibration transmission on seated person in passenger vehicle
AIP Conference Proceedings 2137, 040001 (2019); https://doi.org/10.1063/1.5120999
Effect of graphene doping on the charge carrier and thermoelectric properties of RCF-Bi2S3 composites
AIP Conference Proceedings 2137, 020004 (2019); https://doi.org/10.1063/1.5120980
The Numerical Analysis to Understand the Characteristics of Flow Around a Honeycomb Topography in Relation to
Biofouling Control
Leow Zyncoln
1, b)and Felicia Wong Yen Myan
1, a)1School of Engineering, Faculty of Innovation and Technology. Taylor’s University, 1, Jalan Taylors, 47500 Subang Jaya, Selangor, Malaysia.
a) Corresponding author: [email protected]
Abstract. Biofouling, the undesirable accumulation of microorganisms on exposed or submerged surfaces, has been causing problems for humanity especially the marine industry. Regarding this, several antifouling solutions have been developed to eliminate or minimize the unnecessary maintenance cost associated with the damage done by biofouling.
Although the attempts were a success, they were either impractical or unsustainable. This leads researchers to develop effective yet ecological-friendly alternative. Surfaces with biomimetic micro-structures have been a hot topic of the current antifouling technology for decades as they have shown promising results in control biofouling. One of the micro-structures, honeycomb has also shown certain influences on fouling organisms, but lack of researches has left the efficacy of the distinct geometry rather mysterious. Past studies have considered static flow setting for the surface with honeycomb micro- structures. However, interaction between fouling organisms and the associated micro-textured surface in dynamic flow conditions still remains unanswered. Therefore, this research is crucial to study the hydrodynamic environment around a surface with honeycomb micro-structures via computational fluid dynamics (CFD) analysis. Three-dimensional (3D) models of honeycomb micro-structured surface alongside a non-patterned surface in distinct flow channels will be generated. Mesh generations will be carried out as a pre-processing stage onto the 3D models to perform CFD simulations.
This is an iterative process necessary to determine the optimum mesh element size using mesh independence analysis for more accurate results. The hydrodynamic variables, particularly flow velocity, wall shear stress, and shear strain rate within the honeycomb-structured topography as well as the non-patterned surface will be identified as the outcomes of the research. The data obtained from simulations will be carried on for comparisons, and the analysed results will be reported in the forms of contour and plot. Based on the hypotheses, the expected outcomes should prove that the edges of the honeycomb micro-structures induce high wall shears and hence stimulate strong antifouling behaviour, while the spacings between the structures promote biofouling. It is also expected that the honeycomb micro-patterned surface will yield greater hydrodynamic variations across the structure when compared to the non-patterned counterpart.
INTRODUCTION
Biofouling refers to the gradual process of any macro- or microorganisms to settle and grow on the surfaces of objects submerged in water. The consequences of having biofouling on engineering structures are that it would contribute to corrosion of metal structures and a decrease in the efficiency of moving parts. These have been issues especially to the marine industry. A common phenomenon of such issues is the increased drag due to biofouling on ship hulls that leads to a reduction in their speed and manoeuvrability and hence a huge increase in their fuel consumption. Likewise, biofouling has also impacted the water treatment industry where fouling of membrane systems such as reverse osmosis (RO) and nanofiltration (NF) increases their transmembrane pressure, which subsequently increases the operational cost of the plants.
While the consequences of biofouling have raised significant impact in economic aspect, it is crucial to manage the phenomena since the early stage of biofouling. Consequently, three well-known approaches were developed to
restrain the development of biofilms and settlement of microorganisms. The most effective one − the use of antifoulants or any other chemical agents such as biocidal antifouling coatings has by far been the best option among the antifouling technology to date [1]. These antifoulants deterred biofouling by effectively killing any fouling organisms that come in contact with the surface. However, it brings lethal side effects on innocent aquatic lives [2–
4].
Due to rising ecological concern across the globe, the application of toxic antifouling coatings was restricted by several organizations and countries [5]. Consequently, fellow scientists and researchers have been prompted to research and develop inventive harmless alternatives to control biofouling [1, 6]. This led to another approach which involves the use of chemical agents, specifically foul-release coatings (FRCs) [7]. The coatings worked by reflecting any fouling microorganisms that approached the surface. However, they were not as efficient as the use of biocidal antifoulants, as the mechanism worked only beyond a specific threshold of hydrodynamic shear and the biofilm formation still persisted below the threshold [8].
The unsustainability of biocidal antifoulants and inefficiency of FRCs have rendered the first two approaches impractical, hence diverting the focus of antifouling technology to the third approach – surface modification by developing foul-resistant topographies. This approach is getting widespread in recent years due to its ecologically friendly characteristics. Previous studies including both experimental and numerical methods have shown that the antifouling properties of micro-topographies yield promising results, and this method was sought to eventually replace all the current chemical approaches [9].
Topography geometry is among the key features of a patterned surface in controlling biofouling. It is also one of the elements that affects the settlement trends of fouling organisms on a surface. Numerous micropatterned surfaces adapting naturally inspired topographies (biomimetics) such as micro-riblets and honeycomb gradient have been developed by researchers [10, 11]. Surface patterns based on non-biomimetic geometries such as meshes, microwells, prisms, and pyramids have also shown antifouling efficacy to certian levels in both laboratories and field-level experiments [9, 12–15]. That said, a hypothesis that was drawn by Halder et al. [9] stated that an enclosed pattern such as circular and rectangular wells developing great microfluidic conditions could be considered more effective for antifouling in a dynamic fluid environment, and hexagonal wells (depressions of honeycomb structures) happened to be one of them. This gave rise to the subject of the research which was applied as the geometry of the topography.
Geometry aside, Topography size is another key feature that could influence the settlement trends of biofoulers on a patterned surface. When it comes to performance, it is of primary importance as compared to the geometrical feature of a topography. Generally, fouling microorganisms favour topographies in the scale of macro (1 mm − 100 mm) rather than micro (1 µm − 100 µm) as the former are sufficiently large to fit the microorganisms [16]. A study by Aldred et al. [17] has shown that settlement of cyprids, which are the final larval forms of barnacles, was reduced on micro-topographies with widths ranging from 1 µm – 5 µm and 64 µm – 256 µm, regardless of geometry. Hence, this research was conducted based on the fouling resistance against cyprids settlement.
The antifouling efficacy of micro-patterned surfaces based on the surface modification can be quantified through both experimental and numerical approaches. The former which involves laboratory as well as fields experiments have manifested the efficacy in biofouling control with various topographies targeting certain species each time. On the other hand, the counterpart, numerical approach is also gaining higher interest among researchers as well as various engineering industries. Researches involving the theoretical approach are time consuming and costly, and these can be mitigated with the application of numerical approach, which led to the approach used in this research. Moreover, the numerical approach can have more control over the features and configurations of patterned surface, therefore rendering itself highly versatile.
In this regard, computational fluid dynamics (CFD) simulations were involved to evaluate patterns of flow and compare its performance between two models in relation to biofouling control. Considering the novelty of the subject, validation cannot be performed as there was no related experimental analysis existing as of the time the research was carried out. Hence, further investigations are required to validate the accuracy of the research outcome.
RESEARCH METHODOLOGY
Of all the existing approaches, foul inhibiting surface modification was applied in this project. A locally available SolidWorks 2019 (Student Edition) and ANSYS software package, version 2019 R1 (Academic Version) were used for modelling of geometry and simulation of fluid flow respectively. Some functions and settings applied in the process might not be available in older or later versions of both software. The following subsections describes the sequential
methods to yield the data of the project prior to their numerical analysis including geometry modelling as well as pre- processing and post processing.
Geometry Modelling
Two models of three-dimensional (3D) fluid domains can be portrayed by their distinct bottom surface – a honeycomb-structured surface and a plain, non-patterned surface. Both domains were modelled as 4 mm x 2 mm x 12 mm rectangular microchannels, with the honeycomb-structured domain having its topography located at 8 mm away from the inlet of the domains as shown in Fig. 1. The clearance must be allocated as such so that flows could be fully developed in terms of velocity prior to reaching the topographies. This was done to avoid velocity interference around the inlets of the domains. Formation of the aforementioned parabolic-curved flows can be found under the “Fully Developed Flow” sub-section. The honeycomb topography was formed by a 17-by-17 two-dimensional (2D) array of equilateral hexagons, with incircle diameter of 100 µm and spacing of 110 µm between structures.
FIGURE 1. Isometric view of the geometry of a honeycomb-structured model
Mesh Generation
For the honeycomb-structured model, a tetrahedron mesh method with patch conforming algorithm was applied.
Since the subject of the research revolves around the topography, denser elements were required in that particular region. In this regard, a body of influence (BOI) was applied. The size of an independent body of 2 mm x 1.8 mm x 0.2 mm which was pre-sketched in ANSYS DesignModeler was sufficient to cover the entire topography as well as the flow field around it. Beyond the BOI, the elements were set to grow at the rate of 1.2. With the mesh settings anchored, mesh was then refined subsequently to obtain mesh convergence. The mesh refinement of the honeycomb- structured model was strategized as shown in Table 1.
TABLE 1. Variation of mesh refinements for honeycomb-structured model No. of Variation Mesh Method Universal Element
Size (µm)
In-Body Element Size (µm)
1 Tetrahedron 80 40
2 Tetrahedron 80 35
3 Tetrahedron 80 30
4 Tetrahedron 80 25
5 Tetrahedron 80 20
2 mm 4 mm
12 mm 8 mm
As for the non-patterned model, a sweep method was chosen as it yielded meshes with the best mesh metrics.
Since there were no special features on the surface of the model, no additional mesh settings were needed. The refinement of the non-patterned model was tabulated in Table 2.
TABLE 2. Variation of mesh refinements for non-patterned model No. of Variation Mesh Method Universal Element
Size (µm)
1 Sweep 400
2 Sweep 300
3 Sweep 200
4 Sweep 100
5 Sweep 50
To minimize the variables, the distinct variations of mesh for the honeycomb-structured model were set to be only manipulated by element size inside the BOI. Subsequent in-body element size was reduced by a fixed interval of 5 µm. As for the non-patterned model, the mesh variations were created by manipulating the universal element size by a fixed interval of 100 µm. Both models were meshed until a data extracted from its simulation, particularly the average wall shear successfully converged.
Mesh Independence Analysis
Prior to performing convergence study, mesh metrics were evaluated to ensure that all meshes were acceptable. In this regard, the orthogonal quality and skewness of the meshes were taken into account. Table 3 and Table 4 show the mesh metrics of each mesh variation and the relative error between the average wall shear yielded from simulation for both models.
TABLE 3. Mesh refinements with relative errors for honeycomb-structured model No. of
Varia tion
Mesh Metric Statistics Averag
e Wall Shear (Pa)
Relative Error
Orthogonal Quality Skewness Nodes Elemen
Min Max Avg Min Max Avg t
1 0.20524 0.99185 0.76998 6.8810 e-4
0.79476 0.22921 139477 735016 3.3326 e-4
0.000 2 0.22471 0.99494 0.76972 1.7784
e-4 0.77529 0.22945 156997 823948 5.2817
e-4 58.486 3 0.29251 0.99342 0.77500 2.4482
e-5 0.70749 0.22415 179306 946615 5.4157
e-4 2.537 4 0.29578 0.99394 0.77480 5.2892
e-4 0.70422 0.22442 218873 1171378 5.2633
e-4 -2.814 5 0.24128 0.99257 0.77735 2.3299
e-4
0.75872 0.22190 319438 1720573 5.4797 e-4
4.111
TABLE 4. Mesh refinements with relative errors for non-patterned model No. of
Varia tion
Mesh Metric Statistics
Average Wall Shear
(Pa)
Relative Error Orthogonal
Quality Skewness
Nodes Element Min Max Avg Min Max Avg
1 1 1 1 1.3057
e-10 1.3059
e-10 1.3058
e-10 2046 1500 3.9815e-4 0.000
2 1 1 1 1.3057
e-10 1.3062
e-10 1.3058
e-10 4592 3640 4.4196e-4 11.944
3 1 1 1 1.3057
e-10
1.3068 e-10
1.3058
e-10 14091 12000 4.7486e-4 11.003
4 1 1 1 1.3057
e-10
1.3110 e-10
1.3061
e-10 104181 96000 4.948e-4 5.185
5 1 1 1 1.3057
e-10 1.3395
e-10 1.3071
e-10 800361 768000 5.0612e-4 1.329
For the honeycomb-structured model, it was analysed that in all cases the mesh orthogonal quality was in the range of 0.20524−0.99494 (acceptable value >= 0.1) and the skewness was in the range of 2.4482e-5−0.79476 (acceptable value <=0.8). As for the non-patterned model, the mesh orthogonal quality was at the maximum value of 1 in all cases and the skewness ranged between 1.3057e-10−1.3071e-10. Thus, all the meshes have successfully passed the assessment of mesh metrics.
Based on Table 3 and Table 4 again, it can be seen that the average wall shear increased as the number of nodes and elements increased. The pro-processing data showed that the average wall shear converged at Variation Number 2 and Number 4 for the honeycomb-structured and non-patterned models respectively, where both average wall shears have started to increase/decrease at a much slower pace. Since the subsequent refinement yielded a relative error of less than 5%, mesh convergence was assumed satisfied. Hence, relative data produced by meshes of Variation Number 2 and Number 4 for the respective honeycomb-structured and non-patterned models were selected for future assessment in this research.
Numerical Setup (Pre-Processing)
Governing Equations
A native fluid flow software, ANSYS CFX was used to perform the necessary CFD simulations in this project by solving both continuity and Navier-Stokes equations computationally for pressure and velocity fields. Since all the flows were assumed steady-state and incompressible, both governing equations can be simplified as shown in (1) and (2).
∙ 0 (1)
P (2)
The velocity vectors, can be described as shown in (3:
(3) where , , and represent the velocity components in the -, -, and -directions respectively.
Initial and Boundary Conditions
Steady-state laminar flow was applied as the viscous model of the simulations. Water was set to flow inside the fluid domains from the inlets at 1.0 10-6 kg s-1 normal to the boundary. With the ambient temperature set at 20 °C, the density and dynamic viscosity of water were 1.0 103 kg m-3 and 1.0 10-3 N s m-2 respectively. The domains were enclosed by a four-side no-slip stationary wall. The fluid flowing in the -direction would eventually exit at the outlet with relative pressure of 0 Pa.
Data Acquisition (Post-Processing)
Three hydrodynamic variables that were deemed vital in these simulations of fluid flow were flow velocity, wall shear, and shear strain rate. A vertical, longitudinal midplane (which intersected the centre of topographies for the case of honeycomb-structured model) was created at the centre of each fluid domain along the whole span. This was to image both velocity and shear strain rate contours for non-patterned and honeycomb-structured models. Bottom solid boundaries that had been explicitly defined as walls including those of honeycomb structures were treated as channel beds to generate wall shear contours in both models. For the honeycomb-structured model, three horizontal lines with different depths parallel to -axis were created at the centre of topography to measure all three hydrodynamic variables along the lines. The first line (green) was located on the top face of topography, whereas the second and third lines (red and blue, respectively) were positioned at 40 µm and 70 µm below the first line respectively, as illustrated in Fig. 2A. These positions were selected because contrasting wall shears would possibly be found
between these lines of different depths. As for the non-patterned model, the hydrodynamic variables were measured at only one line (green), which position corresponded to that of the first line of honeycomb-structured model. Figure 2B shows the side-by-side comparison of the positions of measurement lines for both models.
FIGURE 2. Measurement lines for (A) honeycomb and (B) non-patterned models
Analyse of data obtained from the CFD simulations can be found under the “Results” sub-section. Discussion of results and subsequent comparisons of performance between both honeycomb and non-patterned models can be traced under the “Discussions” sub-section.
RESULTS AND DISCUSSIONS Results
Fully Developed Flow
Prior to performing simulations for data acquisition, the position of topography was assessed to ensure that fully developed flows had been formed before reaching the topography. Fully developed flows are important in this project as they could ensure one-direction flow with constant velocity and mitigate disturbance of flows from other directions.
In this regard, a finely meshed model with non-patterned surface was employed to undergo a simulation in order to investigate the point where the flows would be fully developed. Following the simulation, a velocity contour was generated at the middle of the fluid domain to illustrate the formation of its fully developed flow from the inlet at the mass flow rate of 1.0 10-6 kg s-1 as shown in Fig. 3. The gradual increment of the flow velocities from both top and bottom solid boundaries to the centre of the domain was observed as the colour changes from blue to red.
FIGURE 3. Formation of a fully developed flow from inlet of the domain (flow direction from left to right)
On the other hand, a plot was extracted along a longitudinal line created at the exact centre of the fluid domain along the whole span to examine the magnitude of developing flow velocities as shown in Fig. 4.
Based on the curve, it can be observed that the velocity magnitude increased proportionally along the flow direction when the flow first entered the domain. The increment started to pace down and converge at approximately 1.5 mm from the inlet until it went completely stagnant at ~4 mm from the inlet. This indicates that the flow had been fully
A B
converged and developed. Figure 5 shows a parabolic curve of the flow velocities along a vertical line intersecting the longitudinal line at x-coordinate of -0.002.
FIGURE 4. Flow development from inlet to outlet of the domain (flow direction from left to right)
FIGURE 5. Full form of parabolic-curved flow at 4 mm away from the inlet
Velocity
Based on (3), the velocity of a fluid flow can be categorized into four types, namely velocity magnitude, velocity , velocity , and velocity which vectors of velocity are based on the , , and -direction respectively. In this regard, velocity magnitude was applied on the midplanes for generation of contours, whereas velocity which has higher significance as compared to the others was applied along the measurement lines for generation of plots.
Figure 6 and Figure 7 present parts of the contours of velocity magnitude at the midplane on the non-patterned and honeycomb-structured surfaces respectively from the centre of height of the domain to the bottom solid boundary.
The position of the contour on the non-patterned surface was taken to surround the area of topography which corresponded to that on a honeycomb-structured surface. Both non-patterned and honeycomb-structured domains show little to no signs of fluid flow at the bottom walls. This manifested the no-slip condition of the solid boundaries applied to the domains. Based on Fig. 6, the flow above the non-patterned surface shows a gradual increment of velocity magnitude as it approaches the centre of the height. On the other hand, the contour of flow in Fig. 7 shows similar increment of velocity magnitude to that on the non-patterned surface, as the velocity magnitude increases uniformly from the lowest to the highest values. Inside the honeycomb structures, it was found that the velocity magnitude was of minimum value.
FIGURE 6. Velocity contour at the mid plane on the non-patterned surface (flow direction from left ot right)
FIGURE 7. Velocity contour at the mid plane on the honeycomb-structured surface (flow direction from left ot right)
FIGURE 8. Velocity at various measurement lines for both models (flow direction from left to right)
Figure 8 shows the integrated plot of velocity at all the measurement lines for both models. Based on the data obtained from all the measurement lines, the average velocity was -3.3 10-14 m s-1 for the non-patterned surface.
Likewise, the averages of the velocity obtained at depths of 0 µm, 40 µm, and 70 µm were 7.164 10-9 m s-1, -1.222 10-10 m s-1, and 3.411 10-9 m s-1 respectively for the honeycomb-structured surface.
Wall Shear
Figure 9 and Figure 10 shows plan views of wall shear contours at the channel bed on the non-patterned as well as honeycomb-structured surfaces respectively. The position of the contour on the non-patterned surface was taken to surround the area of topography which corresponded to that on a honeycomb-structured surface. The contour in Figure 9 shows that the wall shear on the channel bed of the non-patterned surface was constant along the flow direction with slightly lower wall shear near the solid boundaries. On the other hand, it was observed from the wall shear contour in Fig. 10 that higher wall shears were formed between the gaps of honeycomb structures especially at the centre of the flow. As the position moves from the centre of the gaps to the upper periphery of the honeycomb structures, the wall shear decreases gradually to a moderate value. Likewise, lower wall shears were found at the bottom of the depressions of honeycomb structure.
Figure 11 shows the integrated plot of wall shear at all the measurement lines for both models. The average wall shear recorded along the measurement line in non-patterned surface was 4.995 10-4 Pa. Correspondingly, the average wall shears measured at the depths of 0 µm, 40 µm and 70 µm in honeycomb-structured surface were 5.282 10-4 Pa, 1.942 10-5 Pa, and 2.211 10-5 Pa respectively.
Shear Strain Rate
Figure 12 and Figure 13 illustrate parts of the shear strain rate contours at the midplane on the non-patterned and honeycomb-structured surfaces respectively from the bottom solid boundary to the centre of height of the domain.
The position of the contour on the non-patterned surface was taken to surround the area of topography which corresponded to that on a honeycomb-structured surface. The shear strain contour in Fig. 12 shows that the shear strain rate was at its highest value near the bottom wall of the non-patterned surface. As the position approaches the centre of the flow, the shear strain rate was found decreasing gradually to the lowest value. Similarly, the contour in Fig.13 shows that the shear strain rate decreases uniformly as it approaches the centre of the flow. However, fluctuations of
-1.00E-06 -8.00E-07 -6.00E-07 -4.00E-07 -2.00E-07 0.00E+00 2.00E-07 4.00E-07 6.00E-07 8.00E-07 1.00E-06
0.002 0.0022 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 0.0038 0.004 Velocity v(m s-1)
Position X (m)
Non-Patterned Surface Honeycomb-Structured Surface at 0 um Honeycomb-Structured Surface at 40 um Honeycomb-Structured Surface at 70 um
shear strain rate were found along the flow direction above the honeycomb-structured surface. The shear strain rate was also found decreasing gradually to the minimum value as it approaches the bottom of the structures.
FIGURE 9. Wall shear contour at the channel bed on the non-patterned surface (flow direction from left ot right)
FIGURE 10. Wall shear contour at the channel bed on the honeycomb-structured surface (flow direction from left ot right)
FIGURE 11. Wall shear at various measurement lines for both models (flow direction from left to right)
FIGURE 12. Shear strain rate at the mid plane on a non-patterned surface (flow direction from left ot right) 0.00E+00
1.00E-04 2.00E-04 3.00E-04 4.00E-04 5.00E-04 6.00E-04 7.00E-04 8.00E-04
0.002 0.0022 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 0.0038 0.004
Wall Shear (Pa)
Position X (m)
Non-Patterned Surface Honeycomb-Structured Surface at 0 um Honeycomb-Structured Surface at 40 um Honeycomb-Structured Surface at 70 um
FIGURE 13. Shear strain rate contour at the channel bed on a honeycomb-structured surface (flow direction from left ot right)
FIGURE 14. Shear strain rate at various measurement lines for both models (flow direction from left to right) Figure 14 shows the integrated plot of shear strain rate at all the measurement lines for both models. For the non- patterned surface, the average shear strain rate obtained at the measurement line was 4.463 10-1 s-1. As for the honeycomb-structured surface, the average shear strain rates at the depths of 0 µm, 40 µm, and 70 µm were found 3.984 10-1 s-1, 1.022 10-1 s-1, 2.898 10-2 s-1 respectively.
0.00E+00 1.00E-01 2.00E-01 3.00E-01 4.00E-01 5.00E-01 6.00E-01
0.002 0.0022 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 0.0038 0.004 Shear Strain Rate (s-1)
Position X (m)
Non-Patterned Surface Honeycomb-Structured Surface at 0 um Honeycomb-Structured Surface at 40 um Honeycomb-Structured Surface at 70 um
Discussions
Based on the integrated plot of velocity in Fig. 8, minimal values of the velocity were found along the measurement line on the non-patterned surface. This is ideal for fouling microorganisms to settle and grow, since there were negligible flows and hence little flow disturbance to interrupt their settlement process. Conversely, alternately large and small fluctuations of flow velocity were found on the honeycomb-structured surface along the measurement line at the depth of 0 µm. The larger velocity fluctuations corresponded to flow conditions at the openings of the depressions of honeycomb structures, whereas the smaller velocity fluctuations appeared to be on the flat surfaces above the structures, as illustrated in Fig. 15. This indicates that any microorganisms approaching the opening of the depressions would have higher possibilities to be washed away by the flow as compared to those on the flat surfaces.
On the other hand, low fluctuations of velocity were observed in the same model along the measurement lines at the depths of 40 µm and 70 µm. These corresponded to the flow conditions inside the honeycomb structures where vortices could be found. The wavy characteristic of the associated velocity values resembled the direction of the flow, forming the shape of a vortex. Any potential fouling microorganisms getting into the structures would be washed out by the flow current and vortices, similarly to the conditions on the flat surfaces above. It is also possible that the microorganisms could be entrapped at the bottom edges or corners (as known as ‘kink’ sites) of the structures, as the flows were insignificant in the regions.
FIGURE 15. Indication of velocities along the measurement lines in the fluid domain
The distribution pattern of wall shear could also have an impact on settlement scenarios. Based on the integrated plot in Fig. 11, it was observed that the wall shear distribution was uniform throughout the measurement line on the non-patterned surface, as reflected in the contour in Fig. 9. This signifies that some high shear disturbance but no fluctuations of it were traced. On the other hand, multiple fluctuations of wall shear on the honeycomb-structured surface along the measurement line at the depth of 0 µm were noticed. The sharp increases and decreases of the fluctuations in wall shear corresponded to the upper peripheries of the honeycomb structures, while the regions at between resembled the flat surface above as visualized in Fig. 16.
FIGURE 16. Indication of wall shear values along the measurement lines in the fluid domain
On a side note, it appeared that there was a slight drop in wall shear between each fluctuation. This can be reflected by the physical characteristic of the topography where the drop was located at between two edges of distinct honeycomb structures as shown in Fig. 17. In contrast, peaks of wall shear were formed at the centre of the Y-junctions.
FIGURE 17. Varying shear stress pattern along the line measured at 0 µm depth on the honeycomb-structured surface The honeycomb structures developed high shear bounded zones along the upper peripheries of the structure.
However, the values were nowhere near those at the centre of the structures. Likewise, the shear values on the said zones were lower than the those in the region outside the topography, contrasting the hypothesis made by Halder et al. 9. However, more research with different topography size and configuration is needed to proof the hypothesis incorrect. Regardless of the values, both peaks of the wall shear and its large fluctuations could make the region on the upper surface unfavourable for settlement of microorganisms. Conversely, the wall shear distribution at the centre
Slight drop in wall shear
Peak of wall shear Peak of wall shear
and near bottom of the honeycomb structures was significantly lower than that at the top. Random wall shear values were observed at the centre as it was measured perpendicularly to the vertical walls of the structures, whereas consistent shear patterns were noticed near the bottom surface of the topography. It can be noted that wall shear decreased as it approached the bottom of the structures. Thus, these areas would be an ideal choice for microorganisms since they settle preferably in regions with low wall shear.
The last hydrodynamic variable of the research, differential strain rate could be one of the key factors to disrupt the process of settlement. From the simulations, it was evident that the shear strain rate was constant on the non- patterned surface as illustrated in Fig. 12 and Fig 14. However, there were no fluctuations of strain rate forming at the bottom wall of the domain, hence rendering it a hospitable area for microorganisms despite the high rates of shear strain. Similarly to wall shear, fluctuations of shear strain rate occurred at the top surface of the topography as seen in Fig. 13 and Fig. 18 where sharp increases and decreases in the rate indicate the positions of the upper peripheries of honeycomb structures. The regions at between fluctuations resembled the flat surface above the structures. Differential strain rates can be seen in multiple points across the top surface of the topography, where the values decreased as they approached the bottom of the structures. Majority of the sharp fluctuations occurred frequently at the top and centre of the topography, thus further increasing the possibility of the microorganisms to be washed away on top of the influence by wall shear. These were crucial in designing topographies as sharply fluctuating shear strain rates were among the microfluidic conditions necessary for inhibiting the settlement 9.
FIGURE 18. Indication of shear strain rates along the measurement lines in the fluid domain
Table 5 shows the comparison of averages of various flow variables between all the measurement lines for both non-patterned and honeycomb-structured models.
TABLE 5. Flow variables at various measurement lines for both models Flow Variable
Measurement Line Non-patterned
Surface
Honeycomb- Structured Surface at 0 µm
Honeycomb- Structured Surface at 40 µm
Honeycomb- Structured Surface at 70 µm Velocity (m s-1) -3.3 10-14 7.164 10-9 -1.222 10-10 3.411 10-9
Wall Shear (Pa) 4.995 10-4 5.282 10-4 1.942 10-5 2.211 10-5 Shear Strain Rate (s-1) 4.463 10-1 3.984 10-1 1.022 10-1 2.898 10-2
From the comparison of performance made between both models, it is evident that surface with the honeycomb- structured topography improves the overall antifouling capability in comparison to a surface without it. Furthermore, the averages of all flow variables on the top of the surface with the honeycomb structures were recorded at the highest values in all three measurement lines. This leads to a recommendation where height reduction can be done onto the honeycomb structures as one of the ways to optimize the topography. Other suggestions include manipulation of areas with low hydrodynamic variations by reproducing the areas with high variations. The size of honeycomb structures can be decreased so that the low shear region can be minimized to make space for larger high shear zones at the top region.
REFERENCES
1. K. A. Dafforn, J. A. Lewis, and E. L. Johnston, Antifouling strategies: History and regulation, ecological impacts and mitigation, Mar. Poll. Bull. 62, (2011) pp. 453–465.
2. P. E. Gibbs and G. W. Bryan, Reproductive failure in populations of the Dog-Whelk, Nucella Lapillus, caused by imposex induced by tributyltin from antifouling paints, J. Mar Biol. Assoc. United Kingdom 66, (1986) pp.
767–777.
3. C. Alzieu, J. Sanjuan, P. Michel, M. Borel, and J. P. Dreno, Monitoring and assessement of butyltins in Atlantic coastal waters, Mar. Poll. Bull. 20, (1989) pp. 22–26.
4. Z. Xie, N. C. Wong, P. Y. Qian, and J. W. Qiu, Responses of polychaete Hydroides elegans life stages to copper stress, Mar. Ecol. Prog. Ser. 285, (2005) pp. 89–96.
5. K. Schiff, D. Diehl, and A. Valkirs, Copper emissions from antifouling paint on recreational vessels, Mar. Poll.
Bull. 48, (2004) pp. 371–377.
6. J. A. Callow and M. E. Callow, Trends in the development of environmentally friendly fouling-resistant marine coatings, Nat. Commun. 2, (2011) p. 244.
7. T. Ekblad, G. Bergström, T. Ederth, Poly(ethylene glycol)-containing hydrogel surfaces for antifouling applications in marine and freshwater environments, Biomacromolecules 9, (2008) pp. 2775–2783.
8. M. Lejars, A. Margaillan, and C. Bressy, Fouling release coatings: A nontoxic alternative to biocidal antifouling coatings, Chem. Rev. 112, (2012) pp. 4347–4390.
9. P. Halder, M. Nasabi, F. Lopez, A novel approach to determine the efficacy of patterned surfaces for biofouling control in relation to its microfluidic environment, Biofouling 29, (2013) pp. 697–713.
10. J. F. Schumacher, M. L. Carman, T. G. Estes, A. W. Feinberg, L. H. Wilson, M. E. Callow, J. A. Callow, J. A.
Finlay, A. B. Brennan, Engineered antifouling microtopographies – effect of feature size, geometry, and roughness on settlement of zoospores of the green alga Ulva, Biofouling 23, (2007) pp. 55–62.
11. L. Xiao, S. E. M. Thompson, M. Röhrig, M. E. Callow, J. A. Callow, M. Grunze, A. Rosenhahn, Hot embossed microtopographic gradients reveal morphological cues that guide the settlement of zoospores, Langmuir 29, (2013) pp. 1093–1099.
12. Y. K. Lee, Y. J. Won, J. H. Yoo, K. H. Ahn, and C. H. Lee, Flow analysis and fouling on the patterned membrane surface, J. Memb. Sci. 427, (2013) pp. 320–325.
13. C. Picioreanu, J. S. Vrouwenvelder, and M. C. M. van Loosdrecht, Three-dimensional modeling of biofouling and fluid dynamics in feed spacer channels of membrane devices, J. Memb. Sci. 345, (2009) pp. 340–354.
14. P. Halder, M. Nasabi, N. Jayasuriya, J. Shimeta, M. Deighton, S. Bhattacharya, A. Mitchell, M. A. Bhuiyan, An assessment of the dynamic stability of microorganisms on patterned surfaces in relation to biofouling control, Biofouling 30, (2014) pp. 695–707.
15. Y. J. Won, S. Y. Jung, J. H. Jang, J. W. Lee, H. R. Chae, D. C. Choi, K. H. Ahn, C. H. Lee, P. K. Park, Correlation of membrane fouling with topography of patterned membranes for water treatment, Memb. Sci. 498, (2016) pp.
14–19.
16. F. W. Y. Myan, J. Walker, and O. Paramor, The interaction of marine fouling organisms with topography of varied scale and geometry: A review, Biointerphases 8, (2013) pp. 1–13.
17. N. Aldred, A. Scardino, A. Cavaco, R. de Nys, and A. S. Clare, Attachment strength is a key factor in the selection of surfaces by barnacle cyprids ( Balanus amphitrite ) during settlement, Biofouling 26, (2010) pp. 287–299.
