This article was downloaded by: [171.67.34.

205]
On: 22 June 2013, At: 19:35
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Biofouling: The Journal of Bioadhesion and Biofilm

Research
Publication details, including instructions for authors and subscription information:
[Link]

Effects of coating roughness and biofouling on ship

resistance and powering
a
Michael P. Schultz
a
Department of Naval Architecture and Ocean Engineering, United States Naval Academy,
Annapolis, Maryland, USA
Published online: 12 Sep 2007.

To cite this article: Michael P. Schultz (2007): Effects of coating roughness and biofouling on ship resistance and powering,
Biofouling: The Journal of Bioadhesion and Biofilm Research, 23:5, 331-341

To link to this article: [Link]

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: [Link]

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 Biofouling, 2007; 23(5): 331 – 341

Effects of coating roughness and biofouling on ship resistance

and powering

MICHAEL P. SCHULTZ

Department of Naval Architecture and Ocean Engineering, United States Naval Academy, Annapolis, Maryland, USA

(Received 21 March 2007; accepted 15 May 2007)

Abstract
Predictions of full-scale ship resistance and powering are made for antifouling coating systems with a range of roughness and
fouling conditions. The estimates are based on results from laboratory-scale drag measurements and boundary layer
similarity law analysis. In the present work, predictions are made for a mid-sized naval surface combatant at cruising speed
Downloaded by [[Link]] at 19:35 22 June 2013

and near maximum speed. The results indicate that slime films can lead to significant increases in resistance and powering,
and heavy calcareous fouling results in powering penalties up to 86% at cruising speed. The present estimates show good
agreement with results from full-scale ship power trials.

Keywords: Drag, powering, resistance, biofouling, roughness, antifouling coatings

worldwide ban in 2008 (Champ, 2003), and their

Introduction
primary replacement, copper-based paints, are not as
The primary role of a ship antifouling (AF) coating effective in controlling fouling (Abbott et al. 2000).
system is to limit the increase in frictional drag as a This has spurred active research for effective, non-
result of surface deterioration and biofouling accu- toxic alternatives (e.g. Swain & Schultz, 1996;
mulation. Failure in this pursuit results in either Carman et al. 2006; Genzer & Efimenko, 2006;
increased power to maintain a given speed or Statz et al. 2006).
reduced speed at a given input power, both of which The impact of fouling on drag has long been
have associated economic penalties (Townsin, noted, and much of the early work is reviewed by
2003). The effect of coating condition is crucial to Anon (1952). Subsequent hydrodynamic research
the performance of ships. Frictional drag alone can has included laboratory studies of the effects of
account for as much as 90% of the total drag on biofilms (e.g. Watanabe et al. 1969; Picologlou et al.
some hull types, even when the hull is relatively 1980; Schultz & Swain, 1999), filamentous algae
smooth and unfouled (Kempf, 1937). For this (e.g. Kouwen & Unny, 1973; Schultz, 2000), and
reason, a significant body of work has focused on calcareous macrofouling (e.g. Leer-Andersen &
better understanding the effect of coating roughness Larsson, 2003; Schultz, 2004). Full-scale ship testing
on frictional drag. A review of the early work in this to document the effect of fouling on the drag of
field is given by Lackenby (1962). More recently, the copper-based coatings has also been conducted (e.g.
research focused on assessing how changes in surface Lewthwaite et al. 1985; Haslbeck & Bohlander,
roughness affect the drag of the self-polishing 1992). The results of all these studies indicate that
copolymer (SPC) tributyl tin (TBT) systems (e.g. fouling leads to a significant increase in frictional drag,
Musker, 1980 – 1981; Townsin et al. 1981; Gran- although the magnitude of the increase depends on
ville, 1987; Medhurst, 1989; Grigson, 1992), with the fouling type and coverage. It should be noted that
less effort given to quantifying the effect of fouling. despite the large body of research focused on under-
This was likely due to the fact that the TBT systems standing the impact of fouling on ship resistance, one
provided long-term fouling control with minimal issue which has yet to be satisfactorily addressed is
settlement of fouling organisms (e.g. Jelic-Mrcelic how to define fouling coverage quantitatively so as to
et al. 2006). However, these systems are facing a allow simple predictions of resistance and powering.

Correspondence: M. P. Schultz, Department of Naval Architecture and Ocean Engineering, United States Naval Academy, Annapolis, MD 21402, USA.
Fax: þ1(410) 293 2219. E-mail: mschultz@[Link]
ISSN 0892-7014 print/ISSN 1029-2454 online 2007 Taylor & Francis
DOI: 10.1080/08927010701461974
 332 M. P. Schultz

In the present paper, predictions of full-scale ship model is a function of the Froude number,
resistance and powering are made for a mid-sized Frm ¼ Um(gLm)71/2, where g is the acceleration due
naval surface combatant coated with AF systems with to gravity and Lm is the length of the model.
a range of roughness and fouling conditions operating Likewise, the frictional resistance coefficient of the
at cruising speed and near maximum speed. The model is a function of the Reynolds number,
predictions are based on results from laboratory-scale Rem ¼ UmLmnm71, where nm is the kinematic viscosity
drag measurements and boundary layer similarity law of the fluid for the model tests. The total resistance
analysis. While the roughness and fouling conditions coefficient of the model is, therefore, given as
considered in this work are only representative (not (Gillmer & Johnson, 1982)1:
exhaustive) and the predictions are for a single hull
form, the methodology implemented here can be used CTm ¼ CRm ðFrm Þ þ CFm ðRem Þ ð2Þ
for other fouling conditions and ship hull forms for
which laboratory drag data are available. Future work In order to accurately predict the resistance of the
is planned in which the economic impact of hull full-scale ship with model tests, it is desirable to have
roughness and fouling is quantified for a range of ship dynamic similarity between the ship and the model.
types and operational scenarios using the methodol- Equation 2 illustrates that in order to accomplish
ogy described herein. this, both the Froude number and the Reynolds
number would have to be matched between the
model and the ship. In practice, this is impossible to
Downloaded by [[Link]] at 19:35 22 June 2013

Materials and methods

do. Instead incomplete dynamic similarity is used in
In this section, ship resistance and powering predic- which the Froude number of the model and the ship
tions based on scale model testing are discussed. are matched as shown in Equation 3:
Next, some basic characteristics of turbulent bound-
ary layer flows and the impact of surface roughness Frm ¼ Frs
on these flows are discussed. Finally, a method of
Um Us ð3Þ
relating laboratory measurements and observations pffiffiffiffiffiffiffiffiffi ¼ pffiffiffiffiffiffiffi
to the full-scale frictional drag of a ship using gLm gLs
boundary layer similarity laws is presented. (For
nomenclature, see Appendix). where, Us is the ship speed and Ls is the ship length.
Towing tank tests are carried out at a range of model
speeds corresponding to full-scale ship speeds in
Prediction of ship resistance and powering using
accordance with Equation 3, as shown below:
scale model tests
rffiffiffiffiffiffi
Estimates of ship resistance and power requirements Lm Us
Um ¼ Us ¼ pffiffiffi ð4Þ
are typically made through towing tank tests of a ship Ls l
model that is geometrically similar to the full-scale
ship. The total resistance (drag) of the ship model, where, the scale ratio, l ¼ Ls/Lm. The total resistance
RTm, is made up of two primary components. These coefficient, CTm, is measured at each speed. The
are the residuary resistance, RRm, and the frictional frictional resistance coefficient, CFm, is obtained using
resistance, RFm, as shown in Equation 1 (Gillmer & the ITTC-1957 formula (Woo et al. 1983) given as:
Johnson, 1982).
0:075
CFm ¼ ð5Þ
RTm ¼ RRm þ RFm ð1Þ ðlog10 Rem 2 Þ2

The residuary resistance is mainly due to wavemak- The residuary resistance coefficient of the model,
ing, while the frictional resistance is due to tangential CRm, is then found using Equation 2. Since the
shear stresses on the ship hull arising from the residuary resistance is a function of the Froude
viscosity of the fluid. It is customary to express number and Frm ¼ Frs in the model tests, it is
Equation 1 in its non-dimensional form. This is done understood that CRm ¼ CRs. The frictional resistance
by dividing by the reference dynamic pressure, coefficient of the ship, CFs, is also found using
1 2
2 rm Um , and the wetted area of the ship model hull, Equation 5 by substituting Res for Rem. The total
Sm, where rm is the fluid density for the model tests resistance coefficient of the ship, CTs, is given by
and Um is the ship model towing speed. The total Equation 6 (Gillmer & Johnson, 1982):
resistance coefficient of the model, CTm, is, therefore,
CTs ¼ CRs þ CFs þ CA ð6Þ
the sum of the residuary resistance coefficient of the
model, CRm, and the frictional resistance coefficient, where, CA is the correlation allowance. The correla-
CFm. The residuary resistance coefficient of the tion allowance is used to account for differences in
 Ship resistance and powering 333

actual ship resistance typically due to scale effect on where, PC ¼ propulsive coefficient which represents
propeller characteristics, scale effect on the ship the overall efficiency of the propeller and shafting.
wake, the effect of three-dimensionality on the
viscous resistance, and hull roughness (Woo et al.
Turbulent boundary layer background
1983). In the present work, the effect of hull
roughness will be dealt with separately and will be In order to better understand the influence of coating
accounted for in an added frictional resistance term, roughness and fouling on ship resistance and power-
DCFs so the total resistance coefficient of the ship, ing, some basic characteristics of the mean velocity
CTs, is: profile in a turbulent boundary layer and the effect of
surface roughness will be discussed here. For more
CTs ¼ CRs þ CFs þ CA þ DCFs ð7Þ thorough coverage, the reader is asked to refer to
Schlichting (1979) or Schultz and Swain (2000).
Figure 1 illustrates the relative contribution of A boundary layer forms in the region of fluid
resistance components for a typical naval ship over a adjacent to a ship’s hull when it is underway. In this
range of Froude numbers. At low to moderate speeds region, the motion of the fluid is impeded by its
(Frs 5 0.25), the frictional resistance is the largest viscosity. At the hull (or wall), there is no relative
component of the total drag. At higher speeds, the motion between the fluid and the wall. This is called
residuary resistance becomes dominant. The results the no-slip condition. At some distance away from
shown in Figure 1 are for a hydraulically smooth hull the wall, the mean velocity reaches its freestream
Downloaded by [[Link]] at 19:35 22 June 2013

(DCFs ¼ 0). It should also be noted that the condition value. A velocity gradient is generated due to these
of the hull (i.e. coating roughness and fouling) has a two boundary conditions. The region in which this
direct influence on the frictional resistance, while velocity gradient exists is called the boundary layer.
having a negligible effect on the residuary resistance. The boundary layer grows in thickness as one travels
An increase in hull roughness through coating condi- down the length of the hull.
tion or fouling would cause an increase in the frictional The mean flow in a turbulent boundary layer is
resistance leading to higher total drag. And, because of generally considered to consist of two layers, the
the aforementioned relative importance of the drag inner and outer layer, each with its own scaling law.
components, an increase in hull roughness would In the inner layer, the mean velocity, U, at a given
generally be expected to exert a greater influence at low distance from the wall, y, is determined by the
to moderate speeds than at high speed. friction velocity, Ut, the kinematic viscosity, n, and
The ship powering requirement in terms of shaft the roughness height, k. This is called the ‘law of the
power, SP, is related to the total resistance and ship wall’ and is given by Schubauer and Tchen (1961):
speed by Equation 8:
U yUt kUt
RTs US ¼f ;
SP ¼ ð8Þ Ut n n ð9Þ
PC þ þ þ
U ¼ f ðy ; k Þ:

For the outer layer, von Kármán (1930) showed that

the difference between the velocity at the outer edge
of the boundary layer, Ue, and the local mean
velocity, U, at a distance y from the wall is
determined by the boundary layer thickness, d and
Ut. This is called the ‘velocity-defect law’, given as:
Ue U y
¼g
Ut d ð10Þ
y
Ueþ U þ ¼ g :
d
If the inner and outer velocity profiles are matched
(Equations 9 and 10), a logarithmic velocity profile
results in the overlap region of the two layers (d
y n/Ut). This is called the ‘log-law’ and was proposed
Figure 1. Graph of the total resistance coefficient, CTs, of a typical by Millikan (1938) for smooth walls. For flows over
naval ship vs. Froude number, Frs. The dashed lines show the
smooth or rough walls, it can be expressed as:
relative contribution of the resistance components. The correlation
allowance, CA is taken to be 0.0004 as suggested by Gillmer and
1
Johnson (1982). The hull is assumed to be hydraulically smooth U þ ¼ lnðyþ Þ þ B DU þ ðkþ Þ ð11Þ
(i.e. DCFs ¼ 0). k
 334 M. P. Schultz

where, k ¼ von Kármán constant ¼ 0.41, B ¼ due to the pressure drag on the roughness elements.
smooth-wall log-law intercept ¼ 5.0, and DUþ ¼ This is called the fully rough flow regime.
roughness function (note that, for smooth walls, The roughness function results corresponding to
DUþ : 0). The effect of surface roughness, there- the mean velocity profiles shown in Figure 2 are
fore, is to cause a downward shift (DUþ) in the log- presented in Figure 3. Also shown are the results of
law velocity profile. This is indicative of an increased Shockling et al. (2006) for a similar roughness taken
momentum deficit due to the roughness and is in a fully developed pipe flow. The Figure shows that
directly related to the increase in frictional drag of the flow over this type of roughness is hydraulically
the surface. smooth (DUþ ¼ 0) for ksþ 3 and by the time
The effect of increasing roughness Reynolds ksþ 25, the flow is fully rough with the roughness
number, kþ, on the mean velocity profile can be function exhibiting a logarithmic dependence on ksþ.
seen in the recent results of Schultz and Flack (2007) The limits of the hydraulically smooth and fully
which were obtained in a zero pressure gradient, rough flow regimes depend, at least to some degree,
turbulent boundary layer using laser Doppler velo- on the roughness type. For example, Schlichting
cimetry on a rough surface consisting of diagonal (1979) suggests limits of ksþ 5 for hydraulically
scratches. The roughness height, k, is here taken to smooth flow and ksþ 70 for fully rough flow over
be the equivalent sand roughness height, ks. This is uniform, closely packed sand.
the height of uniform, closely-packed sand which The outer region of a turbulent boundary layer is
gives the same roughness function as the roughness generally thought to be independent of surface rough-
Downloaded by [[Link]] at 19:35 22 June 2013

of interest in the fully rough flow regime (Schlicht- ness except for the role that the roughness plays in
ing, 1979). The Figure shows that for small rough- setting the velocity (Ut) and length (d) scales. There-
ness Reynolds numbers (ksþ 5 *3), the flow is fore, the velocity-defect law (Equation 10) is usually
unaffected by the roughness and DUþ ¼ 0. This considered to be universal for both rough and smooth
condition is called the hydraulically smooth regime. surfaces. The mean velocity profiles of Schultz and
In this regime, the roughness elements are small Flack (2007) are presented using outer scaling in
enough so that any perturbations caused by them are Figure 4. Also included in the figure is a smooth wall
completely damped out by viscosity. As the rough- profile (ksþ 0). Excellent collapse of the mean
ness Reynolds number increases, the roughness velocity profiles is observed regardless of the rough-
elements begin to produce pressure drag leading to ness Reynolds number, thus supporting the concept of
an increase in the roughness function and frictional a universal velocity-defect law. Although the results
drag. This is called the transitionally rough flow shown in Figure 4 are for a single roughness type and
regime. As the roughness Reynolds number con- height, Connelly et al. (2006) also recently demon-
tinues to increase, at some point, the roughness is strated that a universal velocity-defect law applies for a
large enough so that the drag on the wall is entirely range of roughness types and heights.

Figure 2. Graph of turbulent boundary layer, mean velocity profiles using inner (law of the wall) scaling. The profiles are from Schultz and
Flack (2007) and illustrate the effect of increasing roughness Reynolds number, ksþ, on the law of the wall velocity profile.
 Ship resistance and powering 335

covered with the same roughness. In the present work,

Prediction of the change in frictional resistance
Granville’s similarity law scaling will be used to
at ship scale
predict the change in frictional drag of a plate of ship
The similarity between smooth and rough wall length in order to model the change in frictional
boundary layers which was shown in the previous resistance of the ship itself (DCFs). This is a reasonable
section was utilised by Granville (1958, 1987) to assumption since the effect of the pressure gradient
develop a similarity law scaling procedure. This resulting from the three dimensionality of the actual
procedure allows one to predict the impact of a given ship hull is embodied in the correlation allowance,
roughness on the frictional drag of a plate of arbitrary and the pressure drag of the hull form itself is not
length based on laboratory-scale measurements of the expected to be altered dramatically by the roughness.
frictional drag and roughness function of a plate Granville’s similarity law scaling procedure is
illustrated graphically in Figure 5. The procedure
consists of first plotting the frictional resistance
coefficient of a smooth plate as a function of log(Re).
The results of Schultz (2002) showed that smooth
plate laboratory data agreed well with the Kármán-
Schoenherr friction line (Schoenherr, 1932) given as:

0:242
pffiffiffiffiffiffi ¼ logðReCF Þ: ð12Þ
Downloaded by [[Link]] at 19:35 22 June 2013

CF

The frictional resistance coefficient for the rough

plate laboratory measurement is then plotted on the
graph of CF versus log(Re). This point coincides
with the intersection of two loci. The first is the
smooth friction line (Equation 12) displaced by a
distance of DUþk[ln(10)]71 in the positive log(Re)
direction. The second locus is line of constant
Lplateþ ¼ LplateUtn71 which satisfies the following
Figure 3. Graph of the roughness function, DUþ, vs. roughness relationship:
Reynolds number, ksþ. Data are from the boundary layer
measurements of Schultz and Flack (2007) and fully developed
Lplate þ
pipe flow experiments of Shockling et al. (2006). The dashed line Re ¼ qffiffiffiffiffi qffiffiffiffiffi : ð13Þ
CF
shows the asymptotic behaviour of the roughness function in the
2 1 1k C2F
fully rough flow regime.

Figure 4. Graph of turbulent boundary layer, mean velocity profiles using outer (velocity-defect law) scaling. The profiles are from Schultz
and Flack (2007) and illustrate the insensitivity of the outer layer to the surface roughness.
 336 M. P. Schultz

In order to determine CFs, the line of constant Lplateþ condition ranging from 3 – 7% compared to a
is shifted a distance of log(Ls/Lplate) in the positive hydraulically smooth surface. In the fouled condi-
log(Re) direction. The intersection of this line and tion, the increase in frictional resistance depended
the first locus identifies CFs. The increase in frictional strongly on the fouling type with increases 4100%
resistance, DCFs, due to the roughness is the for surfaces covered with calcareous fouling. When
difference between CFs and the smooth value given the surfaces were cleaned, the frictional drag
by Equation 12 for a plate of length, Ls. returned to near, but slightly above, the unfouled
The present ship-scale frictional drag predictions condition due to coating deterioration and damage.
are based on hydrodynamic measurements carried From these measurements and further testing by the
out at the US Naval Academy Hydromechanics author over several more fouling and cleaning cycles,
Laboratory on 1.52 m long plates coated with AF the hydrodynamic properties for a range of coating
coatings in the unfouled, fouled, and cleaned roughness and fouling conditions were generated in
conditions. This work is detailed in Schultz (2004). terms of ks. These are shown in Table I.
In that study, both fouling-release and biocide-based The reader should note that these are representa-
AF systems were tested in the unfouled, fouled, and tive conditions based on the measurements and
cleaned conditions. All the coatings tested exhibited observations given in Schultz (2004). They cannot,
an increase in frictional resistance in the unfouled therefore, be expected to characterise all possible
Downloaded by [[Link]] at 19:35 22 June 2013

Figure 5. Graphical representation of Granville’s similarity law scaling (Granville, 1958; 1987). The graph illustrates how the roughness
function, DUþ, determined from the laboratory hydrodynamic tests is related to the increase in the frictional resistance coefficient at ship
scale, DCFs.

Table I. A range of representative coating and fouling conditions. The Naval Ships’ Technical Manual (NSTM) rating is a fouling index
used by the US Navy based on Naval Ships’ Technical Manual (2002). The values of equivalent sand roughness height (ks) and average
coating roughness (Rt50) are based on the measurements of Schultz (2004).

Description of condition NSTM rating* ks (mm) Rt50 (mm)

Hydraulically smooth surface 0 0 0

Typical as applied AF coating 0 30 150
Deteriorated coating or light slime 10 – 20 100 300
Heavy slime 30 300 600
Small calcareous fouling or weed 40 – 60 1000 1000
Medium calcareous fouling 70 – 80 3000 3000
Heavy calcareous fouling 90 – 100 10,000 10,000

*NSTM (2002).
 Ship resistance and powering 337

coating or fouling conditions. For example, spatial to be 0.0004 as suggested by Gillmer and Johnson
heterogeneity in coating roughness and fouling and (1982) for a similar hull form. The change in the
differences in fouling community composition all frictional resistance coefficient (DCFs) resulting from
could have a significant effect on ks and the resulting the hull roughness was calculated using similarity law
DCFs. Also shown in Table I are the estimated scaling procedure (Granville, 1958; 1987) for
average hull roughness (Rt50) for the different Rt50 ¼ 150 mm (ks ¼ 30 mm). It was assumed for this
conditions and the US Navy fouling rating according calculation, as well as the subsequent ones presented
to Naval Ships’ Technical Manual (NSTM 2002). in this paper, that the roughness function follows the
Rt50 is the maximum peak to trough roughness results of Schultz and Flack (2007) and Shockling
height measured over a sampling length of 50 mm et al. (2006) shown in Figure 3. While it is known
which is a standard measure of coating roughness that the roughness function is not universal for all
based on hull surveys using the BMT Hull Rough- roughness types in the transitionally rough flow
ness Analyzer (Medhurst, 1990; Howell & Behrends, regime (Grigson, 1992), the results of Schultz
2006). Rt50 is not generally used to characterise a (2004) agree reasonably well with this roughness
fouled surface; therefore, the values given for these function for a range of coating roughness and fouling
conditions in Table I represent the likely maximum conditions. The procedure for calculating DCFs is as
peak to trough height of the fouling in each case. follows: i) The value of ks for a given hull condition is
Chapter 081 of the NSTM (2002) is the document selected based on Table I; ii) an initial value of ksþ is
that the US Navy uses for guidance on underwater assumed; iii) the corresponding roughness function
Downloaded by [[Link]] at 19:35 22 June 2013

hull cleaning. The NSTM ratings shown in Table I (DUþ) is determined from Figure 3; iv) the viscous
are used by divers to evaluate the fouling condition of length scale (n/Ut) is determined as ks/ksþ; v) the
the hull in situ. Ratings of 30 or less are used to denote value of Lplateþ is determined as Lplate divided by the
coverage of soft fouling only. A rating of 0 denotes a viscous length scale (Lplate ¼ 1.52 m in the study of
surface free of fouling, while a rating of 30 represents Schultz, 2004); vi) the two loci of Granville’s
heavy coverage of soft fouling. A rating of 40 indicates procedure are located. (a) The first is a line of Lplateþ
a surface with incipient calcareous fouling. The rating from step v satisfying Equation 13; (b) the second is
increases with the severity of the calcareous fouling, Equation 12 offset by DUþk[ln(10)]71 in the positive
with a rating of 100 denoting heavy coverage of large log(Re) direction; vii) the ship scale frictional
calcareous fouling of multiple types. Some distinction resistance coefficient (CFs) is located by the intersec-
should also be drawn between light and heavy slime tion of the second locus and the first locus shifted a
layers. Heavy slime would be a condition where the distance of log(Ls/Lplate) in the positive log(Re)
underlying paint colour is difficult or impossible to direction; viii) the corresponding value of Res is
determine. In the light slime condition, the underlying obtained from the graph. If Res is the value desired,
paint colour is visible. the procedure is complete. If it is too low, a larger ksþ
The paper of Woo et al. (1983) provides both full is used and steps iii – viii are repeated. If Res is too
scale trial powering data as well as model drag and high, a smaller ksþ is used and steps iii – viii are
propeller results for the US Navy Oliver Hazard Perry repeated; ix) the increase in frictional resistance,
class frigate (FFG-7). These data were used in the DCFs, due to the roughness is the difference between
present work in order to assess the impact of coating CFs and the smooth value given by Equation 12 for a
roughness and fouling on the shaft power require- plate of length, Ls; x) DCFs is used in Equations 7 and
ments (SP) of an FFG-7 hull form. The FFG-7 8 to predict the resulting SP.
has a waterline length of 124.4 m with a beam of
14.3 m and displaces 3779 metric tonnes. The
Results and discussion
full-scale trials were carried out in seawater with
r ¼ 1022.3 kg m73 and n ¼ 8.97 6 1077 m2 s71. The predictions of the full-scale powering of the FFG-7
The full-scale predictions herein were also made from model test results were made for ship speeds
for these conditions. The ship was recently out of dry of 7.7 m s71 (15 knots) and 15.4 m s71 (30 knots),
dock when the full-scale trials were conducted, and which correspond to a typical cruising speed and nearly
the average hull roughness (Rt50) was estimated to be full speed, respectively. These predictions along with
150 mm. The ship model tests were conducted on a the results of the full-scale trials of Woo et al. (1983)
1:20.82 scale model at the Naval Surface Warfare are presented in Figure 6. The agreement between
Center, Carderock, MD, USA. Further details are the predictions and the full-scale trials results is
given in Woo et al. (1983). excellent (within 1.5% at both speeds).
The data from the model tests were used to predict Predictions of the change in total resistance were
the full-scale shaft powering requirements of the carried out for the same two ship speeds for all of the
FFG-7 using the procedures previously outlined in coating roughness and fouling conditions shown in
this paper. The correlation allowance (CA) was taken Table I. These are shown in Table II for a ship speed
 338 M. P. Schultz

Table III. Predictions of the change in total resistance (DRT) for an

Oliver Hazard Perry class frigate (FFG-7) with a range of
representative coating and fouling conditions at a speed of
15.4 m s71 (30 knots). Also presented is the percentage decrease
in speed for a fixed shaft power.

Description of DRT @ % DRT @

condition Us ¼ 15.4 m s71 (kN) Us ¼ 15.4 m s71

Hydraulically – –
smooth surface
Typical as applied 46 4%
AF coating
Deteriorated coating 118 10%
or light slime
Heavy slime 192 16%
Small calcareous 305 25%
Figure 6. Graph of required shaft power (SP) vs ship speed (Us) for fouling or weed
an FFG-7 frigate. The solid line shows the results of the full-scale Medium calcareous 447 36%
trials of Woo et al. (1983). The filled circles are the predictions fouling
based on the present method using the model test data of Woo Heavy calcareous 677 55%
et al. (1983) with CA ¼ 0.004. The prediction used similarity law fouling
scaling for a typical as applied AF coating (Rt50 ¼ 150 mm;
Downloaded by [[Link]] at 19:35 22 June 2013

ks ¼ 30 mm).

In general, the percentage increases at this speed are

Table II. Predictions of the change in total resistance (DRT) for an lower than at cruising speed. This is due to the
Oliver Hazard Perry class frigate (FFG-7) with a range of
smaller relative contribution of frictional resistance
representative coating and fouling conditions at a speed of
7.7 m s71 (15 knots).
compared to the residuary resistance at high speed.
The lone exception is the typical, as applied AF
Description of DRT @ % DRT @ coating. This surface goes from operating nearly in
condition Us ¼ 7.7 m s71 (kN) Us ¼ 7.7 m s71 the hydraulically smooth flow regime at cruise
Hydraulically – – (ksþ *7) to well into the transitionally rough flow
smooth surface regime at high speed (ksþ * 14). This explains the
Typical as applied 4.6 2% rise in total resistance penalty at high speed for this
AF coating particular condition. It should be noted that the
Deteriorated coating 23 11%
maximum equivalent sand roughness height (ks) to
or light slime
Heavy slime 41 20% maintain a hydraulically smooth surface does de-
Small calcareous 69 34% pend on the type of ship and its speed. For the
fouling or weed FFG-7, the maximum value of ks for hydraulically
Medium calcareous 105 52% smooth conditions is *13.5 mm at 15 knots and is
fouling
*7 mm at 30 knots. This value tends to increase
Heavy calcareous 162 80%
fouling with ship length, and it decreases with ship speed.
Predictions of the change in required shaft power
were also carried out. These are shown in Table IV
for a ship speed of 7.7 m s71 (15 knots). The reader
of 7.7 m s71 (15 knots). The results indicate that should note that the change in required shaft power
only a small resistance penalty (*2%) is expected to presented here accounts not only for the increase in
result from a typical, as applied AF coating total hull resistance but also for the increased
compared to a hydraulically smooth hull. With propeller loading at a given speed due to the hull
coating deterioration or a light slime layer, the fouling, which can lead to a significant reduction in
impact on resistance begins to become significant the propulsive coefficient for the heavier fouling
(*11% increase in RT). The resistance penalty conditions. These changes in propulsive efficiency
continues to increase with the severity of fouling were calculated based on the open water propeller
until, for a heavy layer of calcareous fouling, the total test results presented in Woo et al. (1983). The
resistance is nearly doubled at cruising speed (*89% powering results again indicate that only a small
increase in RT). penalty (*2%) is expected to result from a typical, as
The predictions of the change in total resistance for applied AF coating compared to a hydraulically
a ship speed of 15.4 m s71 (30 knots) are shown in smooth hull. With coating deterioration or a light
Table III. The percentage increase in resistance due to slime layer, the impact on powering begins to
hull roughness and fouling ranges from 4 – 55%. become significant (*11% increase in SP). The
 Ship resistance and powering 339

powering penalty continues to increase with the Hundley and Tate (1980) conducted a series of ship
severity of fouling until, for a heavy layer of powering trials on USS Harold E. Holt, a Knox class
calcareous fouling, the required power is increased frigate. When operating with a clean propeller and a
by 86%. hull fouled with ‘light grass’ and 10 – 20% coverage
The predictions of the change in required shaft of ‘incipient tube worm growth’, the required shaft
power for a ship speed of 15.4 m s71 (30 knots) are power (SP) at 7.7 m s71 (15 knots) was 4525 kW.
shown in Table V. The percentage powering penalty The SP after the hull was cleaned was 3650 kW.
due to hull roughness and fouling ranges from 4 – Therefore, the increase in required power that can be
59%. In general, the percentage increases at this attributed to the fouling is 24%. The fouling
speed are lower than at cruising speed for the same condition from the present paper that best coincides
reason as was discussed in the resistance results. Also with the pre-cleaning Holt power trial is the ‘small
presented in Table V is the percentage reduction in calcareous fouling or weed’ condition. In this
speed at a fixed input shaft power of 2.7 6 104 kW, condition, the SP is predicted to be 2989 kW at
the required shaft power for the hydraulically smooth 7.7 m s71 (15 knots). It is not as clear what the post-
hull at 15.4 m s71 (30 knots). The analysis shows cleaning hull roughness was on the Holt, but it would
that the percentage reduction in speed ranges from be expected to have been between the ‘typical as
0.9 – 10.7%. This corresponds to a speed loss of applied AF coating’ and ‘deteriorated coating’
0.14 m s71 to 1.6 m s71 (0.28 – 3.2 knots). conditions. The predicted power requirements in
Although there are relatively few full-scale trial these conditions are 2258 kW and 2457 kW, respec-
Downloaded by [[Link]] at 19:35 22 June 2013

results in the literature for hull forms similar to the tively. Therefore, a 22 – 32% increase in SP due to
FFG-7, some comparisons between the ones that the fouling is predicted by present method, which is
exist and the present predictions will be made. in good agreement with the measured increase of
24% on a similar hull form.
Haslbeck and Bohlander (1992) conducted full-
Table IV. Predictions of the change in required shaft power (DSP) scale sea trials in order to quantify the effect of slime
for an Oliver Hazard Perry class frigate (FFG-7) with a range of layers on ship powering. In their study, power trials
representative coating and fouling conditions at a speed of
7.7 m s71 (15 knots).
were conducted on the USS Brewton, another Knox
class frigate. The ship was coated with an ablative AF
Description of DSP @ % DSP @ paint containing both cuprous oxide and TBT and
condition Us ¼ 7.7 m s71 (kW) Us ¼ 7.7 m s71 had been subject to fouling in Pearl Harbor, Hawaii
Hydraulically – –
for 22 months. An initial hull inspection indicated
smooth surface the presence of a fairly heavy slime film with little to
Typical as applied 50 2% no calcareous fouling. Power trials were conducted
AF coating in which the shaft power and ship speed were
Deteriorated coating 250 11% measured. Subsequently, the ship was cleaned and
or light slime
Heavy slime 458 21%
power trials were performed again. After cleaning,
Small calcareous 781 35% the average hull roughness (Rt50) was measured to be
fouling or weed 264 mm using the BMT Hull Roughness Analyzer. A
Medium calcareous 1200 54% difference in required shaft power of 9% at a ship
fouling speed of 8.2 m s71 (16 knots) was measured be-
Heavy calcareous 1908 86%
fouling
tween the two power trials. This is most appro-
priately represented by the difference between the

Table V. Predictions of the change in required shaft power (DSP) for an Oliver Hazard Perry class frigate (FFG-7) with a range of
representative coating and fouling conditions at a speed of 15.4 m s71 (30 knots). Also presented is the percentage decrease in speed for a
fixed shaft power.

DSP @ % DSP @ % reduction in speed

Description of condition Us ¼ 15.4 m s71 (kW) Us ¼ 15.4 m s71 for fixed SP ¼ 2.7 6 104 kW

Hydraulically smooth surface – – –

Typical as applied AF coating 1004 4% 0.9%
Deteriorated coating or light slime 2618 10% 2.7%
Heavy slime 4311 16% 4.0%
Small calcareous fouling or weed 6934 26% 5.8%
Medium calcareous fouling 10,329 38% 7.5%
Heavy calcareous fouling 16,043 59% 10.7%
 340 M. P. Schultz

‘heavy slime’ and the ‘deteriorated coating’ condi-

Acknowledgements
tions in the present work. At cruising speed
(7.7 m s71 or 15 knots), the predicted SP in the The author gratefully acknowledges the financial
‘heavy slime’ condition is 2666 kW. Given the support of this research by the Office of Naval
aforementioned prediction for the ‘deteriorated coat- Research. Thanks are also due to Mr Bill Beaver and
ing’, the increase in required power resulting from Mr Zseleczky and to the referees for their many
the fouling is 8%. This is in very good agreement helpful comments and suggestions.
with the full-scale trials.
Full-scale trials documenting the effect of heavy
Note
calcareous fouling are presented in Marine Fouling
and Its Prevention (Anon, 1952). Unfortunately, the 1. An alternative approach to ship resistance and powering
prediction from scale model tests is to assume that the total
hull forms reported on (two 1920 – 1930s era
resistance is comprised of wavemaking and viscous compo-
destroyers and a World War II era battleship) differ nents (International Towing Tank Conference [ITTC],
significantly from the present frigate. This makes 1978). This is generally referred to as the ITTC-78 method.
direct comparisons impractical. However, at cruising In this method, the viscous component is assumed to be due
speed an increase of 60 – 110% in required shaft to both frictional resistance and pressure or form resistance.
The ITTC-78 procedure explicitly accounts for the form
power was measured for these hull forms. A
resistance as a form factor multiplied by the frictional
reduction in maximum speed at a fixed input power resistance. In the present procedure employed by the US
of 9% was observed for destroyer with a maximum Navy (e.g. Gillmer & Johnson, 1982; Woo et al. 1983), the
Downloaded by [[Link]] at 19:35 22 June 2013

speed similar to the FFG-7. These results, therefore, form resistance is implicitly accounted for in the residuary
are similar to the present predictions. resistance, and no form factor is used. It should be noted that
the resistance and powering predictions from the two methods
are generally comparable.
Conclusions
A method of predicting the impact of coating rough-
ness and fouling on the frictional resistance and References
powering of a full-scale ship has been presented. The Abbott A, Abel PD, Arnold DW, Milne A. 2000. Cost-benefit
prediction relies on results from laboratory-scale ship analysis of the use of TBT: the case for a treatment approach.
model tests along with boundary layer similarity law Sci Total Environ 258:5 – 19.
Anon. 1952. Marine fouling and its prevention. Annapolis, MD:
scaling to assess the impact of the surface condition on US Naval Institute.
the frictional resistance. The method was illustrated Carman ML, Estes TG, Feinberg AW, Schumacher JF, Wilkerson
for an Oliver Hazard Perry class frigate (FFG-7). The W, Wilson LH, Callow ME, Callow JA, Brennan AB. 2006.
results show that a typical, as applied AF coating Engineered antifouling microtopographies – correlating wett-
suffers fairly small penalties in required shaft power at ability with cell attachment. Biofouling 22:11 – 21.
Champ MA. 2003. Economic and environmental impacts on port
both cruising speed and near top speed. A deterio- and harbors from the convention to ban harmful marine anti-
rated coating or light slime film is predicted to cause a fouling systems. Mar Pollut Bull 46:935 – 940.
significant increase in required shaft power for this Connelly JS, Schultz MP, Flack KA. 2006. Velocity-defect scaling
hull form. Heavy calcareous fouling is predicted to for turbulent boundary layers with a range of relative roughness.
increase required shaft power by 86% as compared to Exp Fluids 40:188 – 195.
Genzer J, Efimenko K. 2006. Recent developments in super-
a hydraulically smooth hull at cruising speed. Com- hydrophobic surfaces and their relevance to marine fouling: a
parisons of the present results with full-scale trials of review. Biofouling 22:339 – 360.
similar hull forms reported in the literature show good Gillmer TC, Johnson B. 1982. Introduction to naval architecture.
agreement. Future plans are to utilise this methodol- Annapolis, MD: US Naval Institute.
ogy to assess the economic impact of hull roughness Granville PS. 1958. The frictional resistance and turbulent
boundary layer of rough surfaces. J Ship Res 2:52 – 74.
and fouling to the entire US Navy fleet. Another area Granville PS. 1987. Three indirect methods for the drag
for future work is to better address the role that spatial characterization of arbitrarily rough surfaces on flat plates.
heterogeneity in coating roughness and fouling and J Ship Res 31:70 – 77.
differences in fouling community composition have Grigson CWB. 1992. Drag losses of new ships caused by hull
on the resulting drag and powering penalties. In the finish. J Ship Res 36:182 – 196.
Haslbeck EG, Bohlander G. 1992. Microbial biofilm effects on
present work, the roughness and fouling was assumed drag – lab and field. Proc SNAME Ship Production Symp.
to have a homogeneous distribution over the entire Paper No. 3A-1. 7 p.
hull. This is typically not the case, and significant Howell D, Behrends B. 2006. A review of surface roughness in
variation is often observed. At present, it is not antifouling coatings illustrating the importance of cutoff length.
Biofouling 22:401 – 410.
altogether clear the influence this will have. Hydro-
Hundley LL, Tate CW. 1980. Hull-fouling studies and ship
dynamic measurements on surfaces with significant powering trial results on seven FF 1052 class ships. DW Taylor
heterogeneity in roughness and fouling may serve to Naval Ship Research & Development Center Report #
further elucidate this. DTNSRDC-80/027. 111 p.
 Ship resistance and powering 341

International Towing Tank Conference (ITTC) 1978. Proc 15th Townsin RL, Byrne D, Svensen TE, Milne A. 1981. Estimating
International Towing Tank Conference: report of performance the technical and economic penalties of hull and propeller
committee. The Hague. roughness. Trans SNAME 89:295 – 318.
Jelic-Mrcelic G, Sliskovic M, Antolic B. 2006. Biofouling von Kármán T. 1930. Mechanishe aehnlichkeit and turbulenz.
communities on test panels coated with TBT and TBT-free Nachr Ges Wiss Göttingen 58 – 76.
copper-based antifouling paints. Biofouling 22:293 – 302. Watanabe S, Nagamatsu N, Yokoo K, Kawakami Y. 1969. The
Kempf G. 1937. On the effect of roughness on the resistance of augmentation in frictional resistance due to slime. J Kansai Soc
ships. Trans INA 79:109 – 119. Naval Architects 131:45 – 51.
Kouwen N, Unny TE. 1973. Flexible roughness in open channels. Woo EL, Karafiath G, Borda G. 1983. Ship-model correlation of
ASCE J Hydraulics Division 99:713 – 728. powering performance on USS Oliver Hazard Perry, FFG-7
Lackenby H. 1962. Resistance of ships, with special reference to class. Mar Technol 20:35 – 52.
skin friction and hull surface condition. Proc Inst Mechl Eng
176:981 – 1014.
Leer-Andersen M, Larsson L. 2003. An experimental/numerical
approach for evaluating skin friction on full-scale ships with
Appendix
surface roughness. J Mar Sci Technol 8:26 – 36.
Lewthwaite JC, Molland AF, Thomas KW. 1985. An investigation CA ¼ correlation allowance
into the variation of ship skin frictional resistance with fouling. RF
CF ¼ frictional resistance coefficient ¼ 1rU 2S
Trans R Inst Naval Architects 127:269 – 284. 2
RR
Millikan CM. 1938. A critical discussion of turbulent flows in CR ¼ residuary resistance coefficient ¼ 1rU 2S
2
channels and circular tubes. Proc 5th Int Congr Appl RT
CT ¼ total resistance coefficient ¼ 1rU 2S
Mechanics, Cambridge, MA. pp 386 – 392. 2
Medhurst JS. 1989. The systematic measurement and correlation DCF ¼ change in the frictional resistance coefficient
Downloaded by [[Link]] at 19:35 22 June 2013

of the frictional resistance and topography of ship hull coatings, due to hull roughness/fouling
with particular reference to ablative antifoulings. PhD Thesis, Fr ¼ Froude number ¼ pUffiffiffiffi
University of Newcastle-upon-Tyne, UK. gL
Medhurst JS. 1990. Outline of a draft international standard for
g ¼ gravitational acceleration
the measurement and characterisation of roughness topography k ¼ roughness height
in fluid flow. London: Royal Institute of Naval Architects. ks ¼ equivalent sand roughness height
March 1990. 9 p. k ¼ roughness height
Musker AJ. 1980 – 1981. Universal roughness functions for L ¼ waterline length
naturally-occurring surfaces. Trans Can Soc Mech Eng 1:1 – 6. Lplate ¼ length of flat plate used in laboratory
Naval Ships’ Technical Manual. 2002. Waterbourne underwater experiment
hull cleaning of Navy ships. Publication # S9086-CQ-STM- RF ¼ frictional resistance
010/CH-081R4. Naval Sea Systems Command. 31 p.
RR ¼ residuary resistance
Picologlou BF, Zelver N, Characklis WG. 1980. Biofilm growth
and hydraulic performance. ASCE J Hydraulics Div 106:733 –
RT ¼ total resistance
746. DRT ¼ change in total resistance
Schlichting H. 1979. Boundary-layer theory. 7th ed. New York: Re ¼ Reynolds number ¼ UL n
McGraw-Hill. Rt50 ¼ maximum peak to trough height over a
Schoenherr KE. 1932. Resistances of flat surfaces moving through 50 mm sampling length
a fluid. Trans SNAME 40:279 – 313. PC ¼ propulsive coefficient
Schubauer GB, Tchen CM. 1961. Turbulent flow. New Jersey, S ¼ wetted surface area of the hull
USA: Princeton University Press. SP ¼ shaft power ¼ RPC TU
Schultz MP. 2000. Turbulent boundary layers on surfaces covered
DSP ¼ SPrough 7 SPsmooth
with filamentous algae. ASME J Fluids Eng 122:357 – 363.
Schultz MP. 2002. The relationship between frictional resistance
U ¼ velocity
and roughness for surfaces smoothed by sanding. ASME J Ue ¼ velocity at outer edge
qffiffiffiffi of the boundary layer
Fluids Eng 124:492 – 499. Ut ¼ friction velocity ¼ trw
Schultz MP. 2004. Frictional resistance of antifouling coating
DUþ ¼ roughness function
systems. ASME J Fluids Eng 126:1039 – 1047.
Schultz MP, Swain GW. 1999. The effect of biofilms on turbulent
y ¼ distance from the wall
boundary layers. ASME J Fluids Eng 121:733 – 746. d ¼ boundary layer thickness
Schultz MP, Swain GW. 2000. The influence of biofilms on skin l ¼ scale ratio ¼ LLms
friction drag. Biofouling 15:129 – 139. n ¼ fluid kinematic viscosity
Schultz MP, Flack KA. 2007. The rough-wall turbulent boundary r ¼ fluid density
layer from the hydraulically smooth to the fully rough regime. tw ¼ wall shear stress
J Fluid Mech 580:381 – 405.
Shockling MA, Allen JJ, Smits AJ. 2006. Roughness effects in Superscript
turbulent pipe flow. J Fluid Mech 564:267 – 285.
Statz A, Finlay J, Dalsin J, Callow M, Callow JA, Messersmith PB. þ ¼ normalised by Ut or n/Ut
2006. Algal antifouling and fouling-release properties of metal
surfaces coated with a polymer inspired by marine mussels. Subscripts
Biofouling 22:391 – 399.
Swain GW, Schultz MP. 1996. The testing and evaluation of non- m ¼ model
toxic antifouling coatings. Biofouling 10:187 – 197. rough ¼ for actual surface condition
Townsin RL. 2003. The ship hull fouling penalty. Biofouling s ¼ ship
19(Suppl.):9 – 16. smooth ¼ for hydraulically smooth surface condition