THE TURBULENT BOUNDARY LAYER STRUCTURE OVER DIATOMACEOUS SLIME FOULING

Biofilm fouling has a significant effect on ship performance. Here, the impact of biofilm fouling on boundary layer structure is investigated. Turbulent boundary layer measurements were performed over diatomaceous-slime-fouled plates using high resolution PIV. The mean velocity profile over biofilm shows the expected downward shift (∆U), producing higher drag, and hence higher friction velocity. This increase in drag is seen in enhanced turbulent kinetic energy and Reynolds shear stress. Due to the complex nature of the biofilm’s topography, the flow is heterogeneous in the streamwise direction when compared with smooth-wall flows. INTRODUCTION Many biological surfaces are rough, and man-made surfaces, such as ship hulls, tidal turbine blades, and canals often become rough due to biological activity, such as the attachment and growth of organisms, also known as biofouling. This roughness impacts the performance of these engineered systems (Townsin 2003; Walker et al. 2013a; Walker et al. 2013b). Surface roughness due to biofouling on ship hulls has major economic consequences for shipping and Naval activities. For example, for mid-sized vessels alone, biofouling costs the U.S. Navy an estimated $56 million per year due to increased fuel consumption and the costs of cleaning and painting the hull (Schultz et al. 2011). The primary biofouling community seen on Navy vessels is a biofilm, which is composed of bacterial or algal cells embedded in a viscoelastic extracellular polymeric substance (EPS) (Stoodley et al. 1999). The hydrodynamic regime a biofilm grows in, as well as the organismal makeup of a biofilm determines its physical structure. Different species have different cell surface properties (i.e. hydrophobicity or hydrophilicity), that may influence how the structures interact with the flow within the viscous and turbulent boundary layer above the film (de Beer and Kühl 2001). Biofilm thicknesses range from micrometers to centimeters, and the structure of biofilms is highly heterogeneous, often composed of bulbous cell clusters between which are voids that permit fluid flow (de Beer at al. 1996). When grown under shear, biofilms form thin, flexible streamers that protrude from the surface (Taherzadeh et al. 2009). Eddies are shed off of the cell clusters, causing three-dimensional flapping of the streamers (Stoodley et al. 1998). Biofilms found on ship hulls are often primarily composed of diatoms, and are referred to as diatomaceous slimes (Schultz et al. 2015). Fouling-release and antifouling hull coatings can be ineffective at preventing diatomaceous slime fouling (Molino and Wetherbee 2008). These slimes are also common on marine sediments, where they stabilize the sediment and may alter transport between porewater and the water column (Tolhurst et al. 2008). Though biofilms typically have low vertical relief and the roughness elements are compliant, biofilm fouling induces a steep drag penalty on fouled surfaces, increasing the skin friction on a plate by up to 70% of that of a smooth surface (Schultz et al. 2015). Field and laboratory trials indicate that slime on ship hulls significantly increases the resistance and power requirements of the vessel (Schultz 2007; Haslbeck and Bohlander 1992). In most cases, studies of the effects of roughness on the turbulent boundary layer focus on rigid roughness elements, often with regular spacing (Krogstad and Antonia 1999; Flack et al. 2005; Flack and Schultz 2010). However, in biological systems, compliance and irregularity are the norm. Direct measurements show that biofilms increase skin friction on fouled surfaces, and analysis of the mean velocity profile shows that the effective roughness (ks) of a biofilm is greater than the physical height of 10 International Symposium on Turbulence and Shear Flow Phenomena (TSFP10), Chicago, USA, July, 2017 2 3D-4 the biofilm itself (Walker et al. 2013b). However, under some wall boundary conditions a compliant surface can decrease skin friction due to turbulence by lessening the intensity of turbulence near the wall and reducing the amount of energy carried in streamwise vortices (Xu et al. 2003). Some studies of biofilms and other types of algae growth on already-rough surfaces such as coral reefs or pebbles show a reduction of surface roughness as well as a decrease in bed shear stresses compared to the bare roughness elements because the biofilm growth effectively smooths out the surface (Graba et al. 2010; Nikora et al. 2002; Stocking et al. 2016). It is generally accepted that the effects of roughness on the mean velocity profile in the boundary layer, at high Reynolds numbers relevant to ships, are limited to the inner portion of the boundary layer when the height of the roughness elements are significantly smaller than the boundary layer thickness (Castro 2007, Flack and Schultz 2014, Wu and Christensen 2007). This is referred to as outer layer similarity, where that the outer layer of the turbulent boundary layer over rough and smooth walls is unaffected by the surface roughness (when flow characteristics are normalized by the wall shear velocity). This assumption of outer layer similarity holds for biofilms (Walker et al. 2013b) and forms the basis of scaling techniques that aim to model the effects of surface roughness on vessel performance (Schultz 2007). Different types of roughness can have similar effects on the mean velocity profile (e.g. mesh and rods (Krogstad and Antonia 1999)), but different effects on turbulence generation and turbulent stresses; i.e. roughness alters the structure of the turbulence itself within the boundary layer, altering the size and coherence of vortices and the generation of turbulence at the wall (Volino et al. 2009; Volino et al. 2011, Wu and Christensen 2010, Mejia-Alvarez and Christensen 2011). However, little is known about how compliant roughness alters this turbulence structure. Detailed planar flow measurements are presented over a large biofilm-fouled plate. In order to assess both the average velocity field over a biofilm as well as the heterogeneous nature of turbulence parameters over a natural living surface, high resolution 2-D particle image velocimetry (PIV) was used on the boundary layer in moderate Reynolds number flow. The results presented are for a uniformly-fouled plate with relatively thick biofilm fouling. Therefore, the methodology used, where both the velocity field throughout the boundary layer, and the spatiallyresolved generation of turbulent and shear stresses are measured, provides insights into the mechanisms of the effects of biofilm on boundary layer flow. Given that biofilms can show a large increase in skin friction despite a small physical roughness height, this study examines the spatially explicit effects of a biofilm on the shear velocity, turbulent kinetic energy, instantaneous momentum transport, rotational motion, and coherent structures within the turbulent boundary layer. MATERIALS AND METHODS Biofilm and Facilities A dynamic slime exposure facility, described in Schultz et al. 2015, was used to grow biofilm on large (200 mm x 1.52 m) acrylic plates affixed to the outside of a rotating drum submerged in brackish water with a salinity of 18 ppt. The drum rotates at 60 rpm, creating a peripheral velocity of 1.9 ms, so that biofilm growth occurred under shear. The biofilm consisted of four genera of diatoms (Amphora, Achnanthes, Entomoneis and Navicula) that are commonly found on ships, and are also found on antifouling and fouling-release coatings that have been exposed to the marine environment under dynamic conditions (Schultz et al. 2015). The fouled plate tested was exposed in the dynamic slime facility for 10 weeks and had a uniform layer of biofilm that averaged 1.7 ±0.5 mm thick with a mean peak-totrough distance of 0.5 mm. Testing was performed in a recirculating tunnel facility in the United States Naval Academy Hydromechanics Laboratory. The flow enters the test section through several flow-conditioning devices: a contraction, mesh screens and a honeycomb flow straightener. The freestream turbulence in this facility is less than 0.5% (Volino et al. 2007). The test section of the tunnel is 0.2 m x 0.1 m, with a length of 2 m. The adjustable top wall of the tunnel was set to provide a zero-pressure gradient flow during testing. The free stream velocity was 1.1 ms. Particle image velocimetry (PIV) was used to capture the flow field in the streamwise wall-normal x − y plane. The system consisted of one 6.6k×4.4k pixels 12 bit frame straddle CCD camera (TSI 29MP) coupled with a 190 mJ per pulse, dualcavity pulsed Nd:YAG laser (Quantel). A 0.3 mm thick laser lightsheet was formed by a spherical-cylindrical lens configuration. The flow was seeded with 2 μm silver coated glass-sphere particles, and all measurements were performed ~1.22 m downstream of the boundary layer trip, and ~0.42 m downstream of the leading edge of the fouled plate. Image pairs were processed using a recursive Nyquist grid with 50% overlap ending in a 32 pixel window, resulting in a velocity resolution of 176μm and a field of view of 72 x 42 mm ( 2.4δ×1.4δ ). Vectors statistically very different from their neighbours were removed and replaced with interpolated vectors. More details of PIV processing are given in Barros et al. 2016. Smooth wall boundary layer data is used for comparison in this study. Data were taken in the same facilities as the biofilm data over a smooth acrylic plate. Spatially explicit data are from the PIV analysis as described above, with a 157.27 x 51.47 mm window. The spatial resolution of the smooth wall PIV vector data is 144 μm. Additionally, a smooth wall mean velocity profile was taken using Laser Doppler Velocimetry (LDV) at the same PIV measurement location for comparison purpose. The LDV setup was similar to that described in Schultz and Flack 2007. Table 1. Roughness parameters of the biofilm-fouled plate and the smooth p