Ocean Engineering 262 (2022) 112227

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Experimental and statistical analysis of the hydrodynamic performance of

planing boats: A Comparative study
Erick Giraldo-Pérez ∗, Esteban Betancur, Gilberto Osorio-Gómez
Universidad EAFIT, Design Engineering Research Group (GRID), Cra. 49, No. 7 sur 50 blq. 19, piso 5, Medellín, 050022, Antioquia, Colombia

ARTICLE INFO ABSTRACT

Keywords: Hydrofoil-supported catamarans (Hysucats) have emerged in the last two decades as competitors against
Planing boats monohulls in terms of efficiency in high-speed planing regimes. Nonetheless, according to the literature, they
Hydrofoils have not been compared experimentally. In this work, a monohull, a catamaran and a Hysucat hull were
Hydrodynamic efficiency
compared experimentally in order to evaluate the geometry with the best hydrodynamic behavior. The hulls
Sustainable transport
were towed along a body of water with a test bench and variables such as drag, trim angle, wetted area,
wetted keel length, and sinkage were measured both with sensors, installed in the boat and in the test bench,
and multimedia data. These geometries had the same length, width, draft, submerged volume, and weight
distribution. The equations derived by Savitsky were used as a precursor for the experimental design and for
a deeper understanding of the planing phenomena. It was found that the Hysucat and the catamaran boats
present up to 50% higher drag in planing speeds than the monohull, basically due to their much higher wetted
area and lower trim angle. However, some potential alternatives for improving the hydrodynamic efficiency
of the catamaran and the Hysucat are proposed.

1. Introduction The energy consumed to move the boat depends on this resistance,
of which water causes around a 90% (Faltinsen, 2005). As a rule of
Fossil fuels are the most widely used to propel watercraft, with their thumb, reducing the operational speed, the submerged volume, the
consumption being much higher than land vehicles for similar perfor- wetted area, the waterplane area, or the area facing the flow will
mances. In comparison, a 7.5 meters long classic boat consumes around produce a reduction in the hydrodynamic resistance. However, the
15 times more than a family car, and the weight per usable energy of first strategy may not be an option for high-speed vessels. The four
batteries is at least 16 times higher than that of the gasoline (Candela latter ones can be reduced mainly by minimizing the vehicle weight,
speedboat, 2018). In addition, the cost of batteries compared to the creating hydrodynamic lift with hulls with flat bottoms or planing
total cost of an electric car is around 30% (Bullard, 2019), representing hulls (Savitsky et al., 1964), using hydrofoils (Vellinga, 2009), using
the highest cost for mobility electrification. The wide usage of fossil hard chines or hull steps to cause separation of the flow from the
fuels has transformed water transport into a main player of pollution hull (Savitsky and Morabito, 2010), bleeding air to generate air cavities
in air transport. According to the European Parliament, only maritime under the hull (Latorre, 1997), among other techniques.
transport was responsible for 4% of the global CO2 emissions in 2012,
Hydrofoils are appendages that can be attached to the bottom of
and it could grow up to 17% in 2050 if left unregulated (Cames et al.,
almost any vessel. This make them very versatile and a compelling way
2015). This is why most representative maritime organizations around
to increase efficiency of boats. However, one their biggest limitations
the world such as the International Maritime Organization (IMO) with
is that their proper operation is limited to depth waters in which they
its MARPOL convention, the European Commission (Comission, 2017),
can work far away the surface and the bottom, primarily because the
the UITP (Union Internationale des Transports Publics, 2013), among
foils reduce their performance when close to fluids or elements different
others, are fostering the development of efficient, environmentally
than water. Additionally, the foil could collide against a solid element
friendly and sustainable mobility solutions to reduce reliance on fossil
fuels and reduce pollutant emissions. at the bottom of the water body, causing an accident.
For increasing the performance and efficiency of vessels, it is vital Hydrofoils have been one of the most successful techniques for
to reduce the hydrodynamic resistance without reducing the safety and reducing hydrodynamic resistance on vessels and increasing perfor-
capacity of the vessel. This derives in reducing air and water pollution. mance. Multiple hydrofoil have broken speed records and have achieved

∗ Corresponding author.
E-mail address: [email protected] (E. Giraldo-Pérez).

https://doi.org/10.1016/j.oceaneng.2022.112227
Received 7 March 2022; Received in revised form 2 July 2022; Accepted 5 August 2022
Available online 24 August 2022
0029-8018/© 2023 Published by Elsevier Ltd.
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

high fuel savings since this invention (Vellinga, 2009; Yun and Bliault, 2. State of the art
2012). Nowadays, the best high-performance boats for competitions use
hydrofoils (solar boat race, 2018; Solar Splash, 2012). Electric boats Several studies have been carried out to determine the hydrody-
for leisure or passenger transportation with performances similar to namic performance of planing monohulls, multihulls, and hydrofoil-
combustion-engine boats are appearing more constantly around the supported catamarans. Some examples are Savitsky et al. (1964),
world. Clement and Blount (1963) and Fridsma (1969) for monohulls, Müller-
Another way of reducing hydrodynamic resistance, as mentioned Graf (1989) for catamarans and Hoppe (1989) for Hysucats. These
before, is by replacing traditional curved surfaces at the bottom of studies have become more important in the last decades due to a
the hull with flat surfaces to generate a hydrodynamic force pointing demand for more efficient transport. Three approach types to studying
upwards (i.e., lift). Planing boats take advantage of this principle. They planing hulls and hydrofoil-supported catamarans can be distinguished:
can travel at high speeds without requiring high power amounts and Experimental, statistical and numerical. Yousefi et al. (2013) mentioned
energy consumption; and without penalizing stability and maneuver- studies carried out with these techniques in their review article, but
ability. Like hydrofoils, planing boats take advantage of hydrodynamic more studies have come out since then, so they are reviewed next.
lift to reduce the boat sinkage and the wetted area (Savitsky et al.,
1964). 2.1. Experimental
Planing boats can be monohulls or multihulls. Historically, planing
monohulls have been more popular due to their ease of construction Even though towing tanks have been around for decades, they
and hydrodynamic modeling, allowing designers to accomplish design are still widely used for model testing monohulls, catamarans, and
objectives easier. However, planing catamarans have been taking a Hysucats. Some examples are Kim et al. (2013), Sukas et al. (2017),
bigger role in high-speed vessels for passenger transportation, leisure, Najafi and Nowruzi (2019) and De Luca and Pensa (2017), who ex-
or competition due to their high efficiency, roll stability at high speeds, perimentally measured the hydrodynamic performance of different
and smoother rides in rough water conditions (Yun et al., 2018). monohulls in calm and rough water. Najafi et al. (2019) and Hoppe
As mentioned before, if a hydrofoil is attached to a planing mono- (1989) evaluated the resistance reduction in a catamaran with hydro-
hull, the designer must assure that the monohull will operate in deep foils, and, Broglia et al. (2014), Morabito (2011), Souto-Iglesias et al.
regular-bottom waters. Commonly, monohull hydrofoils operate in (2012), Müller-Graf (1989) and Hu et al. (2020) tested symmetrical
and asymmetrical displacement and planing catamarans with different
channeled rivers, lakes, open seas, and bodies of water where a safe
geometry variations to determine interference effects, among several
depth can be guaranteed. Hopefully, attaching hydrofoils to planing
other authors. Despite the advantages of testing in towing tanks, these
catamarans may partially solve these inconveniences due to depth
installations require high investments and are not widely available.
restrictions. The benefit is twofold: the foils are protected by the
Some authors ran tests towing the models with motor boats (Day
demihull keels (at least against the bottom) without increasing the boat
et al., 2017; Ozden and Demir, 2009) or self-propelled models (Honar-
draft excessively.
yar et al., 2021) in bodies of water. They used GPS to measure the
This concept of hydrofoil catamarans is relatively new compared
speed and several types of supports to hold the model boat and measure
to the invention of planing boats and hydrofoils. The first report on
the drag. However, it is hard to control and measure the speed of a
this invention was published by Hoppe (1989). Since then, several
motorboat. Furthermore, conditions of calm water and no wind in field
successful applications of this concept have been carried out in leisure
experiments are hard to control. For the case of the self-propelled tests,
boats, patrol boats, and ferries with companies like Hysucat, Safehaven
it is not the resistance that is measured but the speed, engine power,
Marine, Icarus Marine, among others. Other applications include solar-
or propeller revolutions.
electric vessel competitions and high-speed sailing competitions such
as the Hydrocontest and the Red Bull Foiling Generation. In this sense,
2.2. Statistical methods
there is an opportunity of replacing classic planing monohulls for
hydrofoil-supported catamarans, seeking more efficient designs without
Statistical approaches, like Savitsky et al. (1964), Wyman (1998), Skene
incurring safety penalties. (1973), Sponberg (2011), and L. Blount and L. Fox (1976), have been
These efficient designs are nothing but vessels producing a rela- extensively implemented in the preliminary study of the hydrodynamic
tively low hydrodynamic drag, which implies less installed power for performance of planing vessels. Out of these methods, the Savitsky
traveling at a design speed and, hence, fewer batteries. This hydro- method is the one that best represents the planing phenomenon for
dynamic resistance can be measured or estimated mainly in 3 ways: monohull vessels compared to the other methods. It has been used in a
experimentally, using statistical methods, and with numerical methods the large number of studies as a benchmark for comparing experimental
(CFD). and numerical studies such as those of Khazaee et al. (2019), Sukas
In this research, a comparative experimental and statistical study et al. (2017), Ekman et al. (2016) and Yousefi et al. (2014), to mention
is carried out between a planing monohull, a planing catamaran, and a few.
a hydrofoil-supported catamaran, the so-called Hysucat, to determine The Savitsky method has a modification for catamarans. Here, the
the geometry with the best hydrodynamic behavior (Low hydrody- lift coefficient of the boat in planing is altered due to an interference
namic drag and appropriate dynamic positioning). The equation de- factor between the hulls (Liu and Wang, 1979; Lee, 1982). These coeffi-
rived by Savitsky et al. (1964) will be used as a tool for a deeper un- cients are used according to the experiments carried out by the authors
derstanding of the planing phenomena and as a precursor for designing with catamarans with different hull spacings at different speeds. This
the experiment. is one more advantage for Savitsky’s method compared with other
For the case study, the planing hulls were selected/designed ac- methods considered since none of them allows its application for
cording to the particular situation of Colombian rivers. These rivers planing catamarans. Unfortunately, it does not allow applications for
are characterized by low depth, high variability of depth, irregular hydrofoil-supported catamarans.
bottoms, and transportation of debris and silt in high concentrations. The Savitsky model is fed with the vessel speed, the boat weight,
This makes the operation of high-speed boats to be more complex and the boat center of gravity location, among others. It outputs
compared to other regions such as Europe. up to 10 variables, but the most important for this study are the
The results derived from this work will help in the implementation hydrodynamic drag, sinkage, and trim angle.
evaluation of Hysucats in water basins where the depth of the water is Some numerical studies, which do not have experimental results
restrictive. beforehand, are validated with the Savitsky model, at the expense of

2
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

only similar geometries considered for developing the Savitsky model. (1975) carried out several experimental resistance tests for catamarans
For instance, Giraldo-Pérez et al. (2020) compared their numerical and monohulls with hard chines at preplaning and planing speeds. Con-
results with Savitsky and other statistical methods such as Blount & sidered catamarans were symmetric, semi-symmetric, and asymmetric
Fox, Crouch, and Keith, Yengejeh et al. (2016) in monohulls, and with different demihull spacings. For the study of Müller-Graf (1989),
asymmetrical planing catamarans. Khazaee et al. (2019) compared the monohull presented the highest total resistance in the preplaning
Savitsky to numerical methods for planing monohulls. regime, followed by the catamarans with 20%–25% lower resistance.
In the planing regime, the behavior was inverted and, the monohull
2.3. Numerical methods presented the lowest total resistance, 16% lower than the symmetric
and asymmetric catamarans. In contrast, Sherman and Fisher (1975)
Numerical methods share the same idea of towing tanks. They are reported the monohull as having less resistance at preplaning and
virtual laboratories in which all the external variables are controlled planing speeds, with differences up to 11%.
via modeling and simulation. Computational Fluid Dynamics (CFD) Thomas et al. (2007) and Seif and Amini (2004) compared the hy-
simulations are the most successful numerical methods for planning drodynamic performance of monohulls and catamarans with statistical
hulls. However, simulating their behavior is complicated due to the methods. Thomas et al. (2007) developed a decision-making tool for
boat’s inherent tendency to change its position according to the speed. choosing between a catamaran and a monohull for a ferry covering a
This makes it necessary for the mesh to adapt to these movements, so route in Australia. After taking into account economical and technical
the simulations are very unstable. feasibilities, the catamaran was selected. The resistance of both hull
This difficulty is even more accentuated for vessels with appendages. types was calculated with Holtrop and Mennen (1982) for displacement
The appendages influence the position of the hull depending on the monohulls and Müller-Graf (1989) for the catamarans. Seif and Amini
speed and are small compared to the size of the hull. Hydrofoil crafts (2004), compared a planing monohull and a planing catamaran of
are a good example. equal displacement at different speeds using Savitsky for monohulls
Due to this, it is needed a high processing capacity. It is common and its modification for catamarans. The Monohull performed around
that researchers who do not have the available computational resources 30% better at planing speeds (Froude numbers from 2.5 to 5) until
to simplify or scale down the boats, or the simulation domain, which both resistance curves converged at a Froude number of 6. Davidson
may introduce errors because of unrealistic physics modeling. For et al. (2011) carried out a qualitative comparison between displace-
example, scales ranging from 1/5 to 1/11 were used in the simulations ment monohulls and catamarans for cargo duties. Here, hydrodynamic
and structural costs were considered, concluding that catamarans may
of Sukas et al. (2017), Fu et al. (2014), Yousefi et al. (2014), Kandasamy
have lower capital and operating costs. Even though in these studies
et al. (2011), and Frisk and Tegehall (2015). The latter three mentioned
the hydrodynamic performance played an important role in making
explicitly to have used a symmetric boundary condition at the center
decisions, these results were not validated against experimental or, at
plane of the hull.
least, numerical methods.
Modeling and simulating planing hulls and hydrofoil crafts require
Souto-Iglesias et al. (2012) and Broglia et al. (2014) compared,
a high level of expertise and a very high computational power to reach
experimentally and numerically, symmetrical catamarans with different
stability and results close to reality, so they are out of the scope of this
demihull spaces with a monohull in displacement regime (0.1 ≤ 𝐹 𝑟 ≤
study.
0.7). Both studies found that the monohull had around 20% less resis-
tance and less sinkage and trim angle at 𝐹 𝑟 ≤ 0.5. Beyond this Froude
2.4. Comparative studies between monohulls, catamarans and hydrofoil-
number, the hydrodynamic performance of the monohull started to be
supported catamarans
similar to those of the catamarans with the highest spaces between
the demihulls. Yun et al. (2018), in their complete review of high-
The aforementioned methods give different levels of precision com-
speed catamarans, argued that asymmetric planing catamarans present
pared to the real scale results and different levels of insights. Their
higher resistance than monohulls at Froude numbers starting at 2, with
application depends on the level of expertise and the available com-
catamarans having around 35% more resistance than monohulls. In
putational, financial, and time resources. However, when evaluating contrast, the trim angle is slightly higher for the monohull between
monohull or multihulls vessels with or without appendages, these Froude numbers of 2 and 3 with 4–5◦ , while for the catamarans is
methods are commonly combined to complement each other. Their around 3–4.5◦ .
main objective is to estimate the drag generated (and therefore the Some authors have carried out experimental comparisons between
propulsion power in demand) at a determined speed for the hull, in catamarans and Hysucats. Hoppe (1989) investigated the hydrody-
addition to other variables such as angle of trim, sinkage, and wetted namic drag performance showing a reduction of 41% for the Hysucat at
area. Froude numbers higher than 2, and up to 4.5. He published the patent
On this basis, authors like Faltinsen (2005), Yun et al. (2018) of the Hysucat in 1986 (Hoppe, 1986). Around one year after, Gerdsen
and Molland et al. (2017) agree that monohulls are more efficient and Jabbusch (1987) published a similar design. In the next decade,
at speeds beyond the preplaning regime compared to catamarans of several Hysucat concepts were developed, and a reduction in resistance
similar principal dimensions and displacement, mainly due to cata- of up to 40% was reached when installing the foils in previously
marans have an extra resistance component, which is the interference designed and operated catamarans (Hoppe, 2001). Lately, Najafi et al.
resistance due to the interaction of both demihulls. This interference re- (2019) carried out an experimental comparison between an asymmetric
sistance depends on both the speed range and the demihull separation. catamaran and a Hysucat at different speeds, installing foils with
It is the lowest for asymmetric catamarans traveling at speeds beyond three different profiles. They reported that the only regime with a
the preplaning regime, due to lower waves being generated because of significant reduction in hydrodynamic drag was at low and medium
the flat internal faces and the low entrance angle in the internal side planing speeds, with a reduction in drag for the Hysucat of 9 to 20%
of the demihulls (Tuck and Lazauskas, 1998). This geometry is usually depending on the used hydrofoil profile. For displacement, preplaning,
used for planing catamarans designed to travel at high speeds (Yun and high planing speeds there was not a significant reduction in drag.
et al., 2018). These variants in the shape of the catamaran make it The Hysucat traveled at trim angles 2–3◦ lower than the catamaran
difficult to determine drag benefits or losses compared to a monohull of during all the evaluated speeds.
similar dimensions, displacement, and weight distribution at different Kandasamy et al. (2011) carried out the only study with CFD in
speeds. Hysucats in the literature, which was based on previous experimen-
Some authors have compared catamarans with monohulls at differ- tal results. However, this study did not compare the hydrodynamic
ent speeds. For example, Müller-Graf (1989) and Sherman and Fisher performance against catamarans.

3
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Table 1
Reported resistance differences between different boat architectures. ‘‘X’’: not evaluated; ‘‘—’’: reference value; ‘‘NA’’: not
available info.
Reference Type of boat Speed regime
Displacement Preplaning Planing
Monohull X – –
Müller-Graf (1989)
Catamaran X Up to -25% Up to +16%

Monohull X – –
Sherman and Fisher (1975)
Catamaran X Up to +11% Up to +11%

Monohull – – X
Thomas et al. (2007)
Catamaran Up to -19% Up to -27% X

Monohull – – –
Seif and Amini (2004)
Catamaran X X Up to +30%

Monohull – X X
Souto-Iglesias et al. (2012)
Catamaran Up to + 18% X X

Monohull – X X
Broglia et al. (2014)
Catamaran Up to +20% X X

Monohull – – –
Yun et al. (2018)
Catamaran Lower Lower Up to +35%

Catamaran – – –
Hoppe (1989)
Hysucat NA NA Up to -45%

Catamaran – – –
Hoppe (2001)
Hysucat NA NA Up to -45%

Catamaran – – –
Najafi et al. (2019)
Hysucat Not significant Not significant Up to -20%

A summary of the reported resistance from studies referenced above

is presented in Table 1.
In summary, the hydrodynamic behavior of monohulls and catama-
rans is very similar during displacement speeds. During the preplaning
regime, authors differ on what geometry between a catamaran and
a monohull presents better hydrodynamic performance. At this speed
regime, foils start working, and a reduction in the hydrodynamic drag
is expected in Hysucats compared to catamarans.
At planing speed regimes, authors agree that monohulls have lower
drag than catamarans with differences ranging from 11% up to 35%,
While reductions in drag for Hysucats in comparison to catamarans
are between 9 and 41%. This hydrodynamic drag depends on the
dynamic positioning of the boat at different speeds, which is described
by variables such as trim angle, sinkage, and wetted area. The lower
the sinkage and the wetted area, the lower the drag.
Concerning the trim angle, the hull should be supported mostly
on the stern (rear) at the bottom, with the bow (forward) out of
the water. Savitsky et al. (1964) found that the optimum trim angle
for planing speeds for prismatic monohulls is between 4 and 5◦ , and
something similar should be for catamarans and Hysucats.
Despite the amount of research that has been carried out, it was not Fig. 1. Proposed methodology for experimental and statistical comparison analysis.
possible to find a study comparing the hydrodynamic performance of a
monohull, a catamaran, and a Hysucat with experimental and statistical
methods. According to this, it is reasonable to consider that a Hysucat
3.1. Selection of the hull shapes
will perform hydro-dynamically better than a monohull given the drag
reduction percentages reported above.
The objective is to determine the hull with the best hydrodynamic
performance out of a set of planing hulls. The methodology can be used
3. Proposed experimental and statistical analysis for comparing any type of planing hulls, but three requirements must
be fulfilled first for the hulls to be comparable between them:

The Fig. 1 depicts the proposed methodology for the experimental • They must have similar principal dimensions: Length, beam and
and statistical comparison analysis. height.
Next, each step of the methodology is described in detail. • They must have the same total weight or submerged volume.

4
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

• They must have the center of gravity located at the same point as
long as it greatly affects the planing behavior.

3.2. Application of Savitsky method

3.2.1. Selection of the speeds for towing tests

Drag is the most important variable on the hydrodynamic perfor-
mance. It is used here to select the speeds at which the models are
towed. An important consideration is to select enough speeds to achieve
a proper resolution of the curve of hydrodynamic drag vs speed. Higher
resolutions are needed near curve inflections and around the objective
planing speed. The first approximation of this drag curve is obtained
using the Savitsky method.
The input variables for the Savitsky model are given for the real-
scale prototype and are the boat weight, the longitudinal center of
gravity, some principal geometrical parameters of the hull, and the
speed.
The most important output variables in order to measure the hydro- Fig. 2. Outline of the collected variables. One arrow represents direct collection of the
dynamic performance are drag, trim angle, wetted area, wetted keel variable in the computer. Two arrows represent the variable is collected separately, and
then sent to the computer.
length and sinkage. These variables are to be measured at different
speeds. The conventional Savitsky method is used for the monohull
and a modified version is used for the catamaran. No adaptation of the
Savitsky method for Hysucats was found in the literature. The hull models are to be towed from zero speed to the selected
speed for the run, allowing the boat to stabilize and keeping the speed
3.2.2. Selection of the scale for towing tests the longest according to the available length of the body of water. This
Another important input is the scale to which the experimental tests process is to be executed at least 3 times per treatment at every speed
are carried out. The decision depends on the maximum speed to be to have significant statistical data.
reached with the scaled models and on the available infrastructure, Even though the data is to be taken throughout the run, it is
including the length of the body of water in which the models will be necessary to select only the time segment where the speed and drag
tested. are stable. A time segment of a run is considered stable if it lasts at
According to Froude’s law of similarity, the closer the scale to the least 5 s and if it achieves a determined standard deviation and a mean
real scale, the higher the speeds (closer to the real scale speeds). There- value. The standard deviation for each variable cannot exceed 2% of
fore, the higher the vibrations and load exerted over the infrastructure the maximum value to reach in the run for the respective variable.
with which the model is towed, implying a higher noise in the collected Also, the mean value for the speed cannot be more than a certain
value away from the intended speed for each specific run. Similarly to
data. Also, the lower the scale, the higher the errors to be encountered
the standard deviation, this value is set as 2% of the maximum speed
in the results considering that several properties of the fluids are not
to be reached in the experiment. The intended speeds are the speeds
scalable, such as the viscous properties.
selected with the Savitsky method. This allows us to assume that the
output variables are taken at the intended speeds.
3.3. Experimental design
With the application of the criteria, it should be expected a normal
distribution for all the variables, which allows for averaging the results
The experiment setup was inspired in the ITTC recommended pro-
for each run. With averaging results, pseudo-replication is avoided due
cedure 7.5-02-05-01 (Tank, 2008), so this procedure is followed where
to the points taken for a single run cannot be considered as different
possible.
samples because they are dependent on each other.

3.3.1. Data collection 3.3.2. Comparison between hull geometries

The main objective of the experiment is to tow the models at After obtaining the experimental results, statistical tools can be
different speeds and measure the variables of interest. Fig. 2 shows the used to determine if the experimental setup must be changed or if
variables of interest and some proposed sensors/tools to collect each there is insufficient data considering the quality of the experimental
of them. It has to be ensured that the experiments were carried out in measurements according to the restrictions mentioned above. Also,
similar environmental conditions. Variables of temperature and wind other metrics such as the statistical power or the 𝑝-value are considered.
speed are measured to check a constant behavior. Ripples and small The statistical hypotheses to be tested for each intended speed
waves in the water are to be avoided as possible. with the multiple comparison tests are determined according to the
It must be ensured that the models are at least two lengths away number of planing boats and according to previous information from
from any wall sideways, and at least one length away from any wall the Savitsky method on the hydrodynamic behavior.
downwards for them to not interfere in the experiments. The models As the output variables are assumed to be taken at the intended
are to be weighted after every 15–20 runs, also checking the center of speeds, it is possible to carry out the ANOVA analysis and multiple
gravity is in the same place. comparison tests for each speed between the treatments. The three
The experiment is a complete randomized design experiment for the ANOVA assumptions (normality, homogeneity of variance between the
external effects to be homogeneous between the evaluated treatments. treatments, and independence between runs) are to be checked with
Here, it is assumed that there are only two sources of variability, the both statistics and charts. For the assumption of normal distribution
treatments themselves and the variability due to external sources. of residuals, Shapiro tests are carried out. Charts of distribution of the
The experiment has two input variables, but for convenience, the residuals and QQ-plots are also used. For asserting the assumption of
comparisons for drag and trim are made at the same speed. This homogeneity of variances, Barlett and Levene’s tests are to be carried
transforms the experiment into a single-factor experiment, in which out. Residuals are plotted according to the treatment, in order to
there are three treatments, which are the selected hull geometries. visualize their range of variation. Lastly, for testing the assumption

5
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 4. Proposed hulls for the experiments.

Table 2
Principal dimensions of the catamaran hull. These parameters are the
Fig. 3. Monohull boat for the transportation of passengers operating in Colombian same for the monohull, excepting there are not demihulls.
rivers. Total length 7.9 m
Total beam 2.12 m
Wetted length 7.5 m
Distance between demihulls 1 m
of independence of errors, Durbin & Watson tests are carried out and Freeboard height (Max. load) 0.4 m
residuals are plotted according to the order of the runs to be taken. Draft (Max. load) 0.7 m
For the multiple comparison test, the Tukey-HSD test is selected. The Total height 1.1 m
Displacement 4.3 m3
chosen significance level for evaluating all the statistical hypotheses is
5%.

4. Case study 4.1.1. Description of the monohull

The monohull has high usability in the context due to its simplicity
in construction and operational versatility. A disadvantage of this shape
The case study is framed in a research project aimed to design an
is that for low-depth bodies of water, hydrofoils cannot be attached to
electric boat that replaces the current boats operating in Colombian
it because the draft of the boat is greatly increased.
rivers for the public transportation of passengers. Therefore, the de-
In addition, it is not convenient for the hull to emerge completely
signed boat must have similar characteristics to these selected vessels.
out of the water in this context because of the low depth and irregular
These vessels are monohull planing boats propelled with fossil fuel
bottom of the river, also the hydrofoil would not be protected in a
outboard engines with an output power of 200 HP that travel at high
hypothetical crash against the river bottom.
speeds, concretely at around 55 km/h on average, and are able to
convey between 20 and 25 passengers. Fig. 3 shows a reference image
4.1.2. Description of the catamaran hull
of the monohull boat.
The catamaran is selected because of its high efficiency at high
Colombian rivers where these monohull vessels operate have several
speeds if properly designed, its higher roll stability, and because it
characteristics that make the operation of boats more complex, in
allows installing a hydrofoil between the demihulls without the need
comparison to the operation in other bodies of water in other regions.
of increasing the hull draft.
These Colombian rivers are low-depth rivers that greatly variate their
The catamaran hull was designed based on the requirements of
flow rate along the year due to the seasonal rains and droughts, which
the zone of interest. The restrictions and requirements were derived
makes the depth also variate (Arcadis Nederland and Jesyca, 2015;
from a preliminary study carried out in the works of Mira et al.
Calle, 2016). This high variation in depth, combined with the fact
(2020) and Giraldo-Pérez et al. (2020). Some important restrictions and
that the bottom of these rivers has very irregular shapes, makes them
requirements related to the form of the hull were: width, draft, required
very dangerous to navigate with motorboats during some seasons in
power at cruise speed, the total weight of the boat at maximum load,
the year. Furthermore, according to some authors like Restrepo et al. passenger capacity, the available area on deck, and available volume
(2018) and Higgins et al. (2016), these rivers transport silt at very high for batteries for a fully electric boat.
concentration rates making it unfeasible to install channels, as in some Between the demihulls, the more turbulent the water the more
European rivers. affected the performance of the hydrofoils will be, so the demihulls
In this chapter, results are shown according to the steps in the have flat inner walls to minimally disturb the flow inside the tunnel.
methodology presented in Fig. 1. These flat walls also help reduce the interference resistance of the
catamaran (Yun et al., 2018).
4.1. Selection of the hull shapes The upper part of the tunnel has a V-shaped geometry to facilitate
the breaking of the water flow facing the vessel and to induce the
According to the context, and the reviewed literature, that high- planing regime more easily. A significant portion of the tunnel remains
lights the hydrodynamic advantages of the Hysucat, two hypothe- under the water surface to provide greater buoyancy to the vessel, but
ses arise, which are intended to be answered using the proposed at cruising speed it raises above the water surface due to the combined
methodology: planing effect and the eventual assistance of the hydrofoils. The hull has
a deadrise angle of 22◦ near the bow and ends at 5◦ towards the stern
1. The proposed catamaran presents a higher hydrodynamic drag of the boat. This change of angle allows that in a displacement regime,
and therefore a worse hydrodynamic behavior compared to the the deadrise angle facing the water is sharper and facilitates the task of
monohull at planing speeds. cutting through the water waves. Additionally, a lower deadrise angle
2. When installing a hydrofoil in the proposed catamaran, the re- at the stern of the boat helps to concentrate most of the hull lift to enter
duction in the hydrodynamic drag is enough to be less than that in planing regime and keep the front of the boat out of the water, thus
of the monohull thanks to the better hydrodynamic behavior, at reducing the drag.
the same nominal planing speed. The geometrical parameters were chosen according to a market
study, in which Hysucat vessels with similar dimensions and capacities
That is why in this work, three planing hulls were chosen. A in terms of number of passengers and operating speeds were taken into
monohull, a catamaran, and, a catamaran with hydrofoils or Hysucat. account. Tables 2 and 3 contain the geometrical parameters of the hull
Fig. 4 shows the front views of the selected three boats. and the dimensionless hull numbers, respectively.

6
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 5. Target weight reduction with foils at cruise speed.

Table 3
Adimensional numbers and geometrical parameters of the hull. The adimensional
numbers are calculated at the cruise speed, which is 15 m/s. These parameters are
the same for the monohull, except there are no demihulls.
Parameter/Variable Formula Value
𝑉
Longitudinal Froude number 𝐹𝑙 = √ 1.7
𝑔⋅𝐿
𝑉
Volumetric Froude number 𝐹𝑙 = √ 1
3.8
𝑔⋅𝑉 3

Length to beam ratio 𝐿∕𝐵 3.6

𝐿𝑊 𝐿
Slenderness ratio 1 4.7
∇ 3
∇
Static load coefficient 𝐵3
0.46
𝐵
Demihull separation ratio 𝐶𝑠 = 8.3
2⋅𝑏 Fig. 6. Longitudinal position of the hydrofoils on the model.
Half angle of entrance 𝑖𝑒 60◦
𝐿𝐶𝐺
LCG to length ratio measured from the transom 𝐿
0.4
Deadrise angle at amidship 𝛽 17◦ Table 4
∇
Block coefficient 𝐶𝑏 = 0.33 Geometric parameters of the prototype and model for hydrofoil design.
𝐿.𝐵.𝐷
Parameter/Variable Prototype Model
Estimated weight (kg) 4300 8.5
Tunel width (m) 1 0.13
4.1.3. Description of the hydrofoil system Cruise speed (m/s) 15 5.3
The hydrofoils are intended to partially push the hull out of the Foilborne speed (m/s) 7.5 2.7
water. Even though this does not provide a drag reduction as high as in Weight Reduction objective (%) 72
Weight supported by the front foil (%) 86
the case where the hull completely emerges out of the water surface, it
Supported weight by the rear foil (%) 14
may provide a substantial reduction in drag of up to 41% in catamarans
as reported in the literature. The objective of the hydrofoils is to reduce
the submerged volume of the craft at the design planing speed and
in consequence, reduce the hydrodynamic drag. The idea behind this equal size and weight capacity, except that for the catamaran for which
target is that the tunnel’s upper part raises out of the water and travels the interference factor and separation ratio is different from 1. For
only supported by the hydrofoils and the demihulls. both hull geometries, it is assumed that the thrust line passes through
The demihulls, in addition to supporting the vessel by buoyancy the center of gravity of the vessel and that its inclination is zero
forces and supporting the boat in planing, protect the hydrofoils from
with respect to the horizontal. Also, it is assumed that the boats are
possible collisions with the river bottom and reduce the likelihood of
prismatic, given that Savitsky does not take into account variations in
collisions with floating elements in the river. The demihulls help the
the transverse geometry along the boat.
vessel stability, mainly in pitch and roll at rest and in planing regime.
Seven speeds are selected to build the whole curve of drag vs speed
Due to this, it is not necessary to have control surfaces on the hydrofoils
or variable angles to keep stable the hull. The profile for the hydrofoils with an acceptable resolution, especially at the speeds where inflections
and their general geometric parameters are selected according to the occur in the curve obtained from Savitsky. Fig. 7 shows the graphical
recommendations found in the literature (Vellinga, 2009; Najafi and representation of the selected speeds according to Savitsky results.
Nowruzi, 2019). The speeds were selected according to the monohull as the baseline
Fig. 5 shows the objective reduction for the catamaran with the geometry.
hydrofoils for the upper part of the tunnel to completely emerge out The selected scale is 1/8 due to limitations in the maximum speed
of the water. For the crest located in the middle of the upper part that can be reached with the equipment available to tow the model. If
of the tunnel not to touch the water surface, the boat must travel as the maximum scale to be reached at the real scale is 20 m/s, the speed
horizontally as possible. for a scale of 1/8 according to the law of Froude similitude is 7.1 m/s.
The hydrofoils are arranged in a traditional configuration, similar For obtaining the speeds to tow the models, the real scale intended
to that of an airplane. The main foil is located slightly forward of speeds (See Fig. 7) are divided by 80.5 .
the center of gravity, supporting around 86% of the weight carried
by both foils. The secondary foil, whose job is to provide stability and 4.3. Experimental setup
counteract the moment caused by the main foil lift, is located near the
transom and supports the 14% left. Fig. 6 shows the arrangement of
After an iterative refinement process according to the quality of the
the foils with the distances relative to the transom. Table 4 shows the
collected data, the selected experiment setup is described next.
cruise, foilborne speed of the hydrofoils and the percentage of weight
According to the scale, a run is considered stable if it fulfills three
to be carried out by each of them in the model and the real-scale
prototype. conditions: it lasts at least 5 s and if both the standard deviation of
velocity and force are maximum 0.1 m/s and 0.5 N, respectively, also
4.2. Application of the Savitsky method if its average velocity is no more than 0.3 m/s away from the intended
velocity for each run. These values are determined as they are less
Table 5 shows the input parameters for the monohull and the than 2% of the maximum expected values of speed and drag in the
catamaran. The parameters for the two vessels are similar, given their experiments. The maximum value expected for the drag is 30 N.

7
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Table 5
Input parameters and variables for the Savitsky model.
Parameter/variable Monohull Catamaran
Weight (kg) 4300 4300
Mean chine beam (m) 1.84 1.84
Mean deadrise angle (◦ ) 13 17
LCG (m) 3.16 3.16
𝜌 (𝑘𝑔∕𝑚3 ) 1000 1000
𝜇 (𝑚2 ∕𝑠) 1.19e−6 1.19e−6
Perpendicular distance between CG and the action line of the thrust (m) 0 0
Tilt angle of the thrust line(◦ ) 0 0
√
Interference factor (-) N/A 2
Separation ratio between demihulls (-) N/A 0.5
Speed range (m/s) 1–18 1-18
Safety factor (-) 1.3 1.3

Fig. 7. Selected speeds for the towing test.

The boat is towed in a 100 meters long lake with a test bench that is Table 6
Characteristics of the scale models.
fixed to the ground and whose traction is generated by a Golden Motor
Parameter/variable Monohull Catamaran
Magicpie BLT-800 electric hub motor. A rim is connected to the engine
through its axis. The rim rolls up nylon, which in turn is attached co- Scale 1/8 1/8
Weight (kg) 8.5 8.5
linearly to the model’s center of gravity. This tension generated in the
LCG/LOA (-) 0.4 0.4
nylon is transmitted by a pulley system, which is arranged in such a LOA (m) 1 1
way that this tension can be measured with a load cell. In this pulley Beam (m) 0.3 0.27
system, a fixed pulley and a guide ring are placed in such a way the Height (m) 0.14 0.12
Cruise speed (m/s) 5.5 5.5
nylon touched 180◦ of the main pulley, for the force over the cell to
be 2 times the tension in the nylon. The rotational speed of the rim is
measured with an encoder and is converted to the speed of the boat
assuming rigid nylon. on the surface facilitates the detachment of the flow, which generates
A Pololux inertial measurement unit (IMU) sensor Mini IMU-9 v5 premature turbulence.
is positioned in the center of gravity of the models to measure the According to the selected boats and the hypotheses expressed in
trim angle (pitch) of the models. Other variables describing the position Section 3.1, the statistical hypotheses to be tested are the following for
of the boat, like the wetted surface, the sinkage, and the wetted keel the drag:
length are measured by recording videos of the boat, from one side of
the lake to capture the position of the boat with respect to the water 𝐻0 ∶ 𝜇𝑐𝑎𝑡𝑎𝑚𝑎𝑟𝑎𝑛 = 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (1)
surface with the aid of waterlines that were marked in the hulls. The
layout of the experiment is presented in Fig. 8. 𝐻𝑎 ∶ 𝜇𝑐𝑎𝑡𝑎𝑚𝑎𝑟𝑎𝑛 > 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (2)
The models were manufactured in polylactic acid (PLA) by additive
manufacturing. The scaling procedure of the boat was done with the and
number of Froude. The specifications of the models are shown in 𝐻0 ∶ 𝜇𝐻𝑦𝑠𝑢𝑐𝑎𝑡 = 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (3)
Table 6.
The Fig. 9 shows the built models of the (a) monohull, (b) catamaran
𝐻𝑎 ∶ 𝜇𝐻𝑦𝑠𝑢𝑐𝑎𝑡 < 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (4)
and (c) Hysucat.
The complementary weights within the model were carefully posi- and the following for the dynamic positioning variables (trim angle,
tioned in order to keep the center of gravity at 40% of the boat length wetted area, wetted keel length, and sinkage):
measured from the transom. In the case of the Hysucat, the foils were
manufactured in aluminum and were machined in a CNC machine. 𝐻0 ∶ 𝜇𝑐𝑎𝑡𝑎𝑚𝑎𝑟𝑎𝑛 = 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (5)
Both, the boats and the foils were sanded in order to smooth the surface.
This was especially important for the hydrofoils since any roughness 𝐻𝑎 ∶ 𝜇𝑐𝑎𝑡𝑎𝑚𝑎𝑟𝑎𝑛 ≠ 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (6)

8
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 8. Diagram of the experimental setup.

Fig. 9. Manufacture of the models for the experimental tests. (a) Monohull, (b) Catamaran, (c) Hysucat.

and treatments at the same speeds, constitutes an unbalanced experiment.

This was handled by the tests of multiple comparisons and the tests for
𝐻0 ∶ 𝜇𝐻𝑦𝑠𝑢𝑐𝑎𝑡 = 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (7)
the ANOVA assumptions as long as there were more than two points
for the treatment containing the least repetitions.
𝐻𝑎 ∶ 𝜇𝐻𝑦𝑠𝑢𝑐𝑎𝑡 ≠ 𝜇𝑚𝑜𝑛𝑜ℎ𝑢𝑙𝑙 (8)
5.1. Data cleaning
Where 𝜇 is the mean for each variable. Note that the monohull is
always the reference. When it comes to finding what treatment has a Fig. 10 shows an example of the cleaning process for a run of the
greater or lower output than the other, the Tukey-HSD test p-values do monohull boat at an intended speed of 6 m/s. It is observed that much
not provide this information by their own. This information is obtained of the data during the run is discarded and only the part where the two
from the plot of the confidence intervals for the mean of each treatment curves reach a constant value is taken.
or ‘mean plot’. Figs. 11 and 12 illustrates the results of performing this procedure
for all the runs and the scaling up of the results to the real scale, for
5. Results the drag and the trim angle, respectively.
Fig. 13 shows the distributions for speed, drag, and trim angle for
For the speed, drag, and trim, at least 3 repetitions per treatment are all the runs at real scale at an intended speed of 15 m/s. It can be
measured. The fact that not the same repetitions could be measured for observed that not all the runs have a distribution similar to a normal

9
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

The flattened normal distributions occur more often for the trim
angle (pitch), which may be presented due to high oscillations in the
speed, which in turn causes oscillations in the boat position. These
oscillations are detected by the IMU, but not by the load cell. Recall the
IMU was located in the boat, while the load cell was on the ground, far
away from the boat so both sensors were subjected to different sources
of variation.
When increasing the speed, the variation is higher for all the vari-
ables. This variation is more accentuated for the trim angle and can be
observed in Fig. 12. This behavior was expected to occur beforehand
considering the electric engine and the towing test bench tend to
be more unstable because of the loads and vibrations. Even though
this high dispersion, in most cases the distributions are approximately
symmetrical and their means lie near each other for the same treatment
at the same speed.
Runs for the intended speed of 20 m/s present a special behavior,
as can be seen in Fig. 11. Due to a restriction in the maximum power of
the electric engine with which the models were towed, the measured
speed presents very high oscillations for the catamaran and Hysucat.
Runs for catamaran and Hysucat do not fulfill the criteria for discarding
runs explained in the previous chapter. The means of the measured
runs are farther away than 0.2 m/s (or 0.6 m/s at the real scale) than
the intended speed of 20 m/s. That may give an advantage to those
two models (or treatments) with respect to the monohull due to the
drag is expected to be underestimated. However, ANOVA and multiple
comparison tests are carried out bearing this in mind. Means for all
Fig. 10. Filtering of the data for a run with the monohull for an intended speed of
the distributions are used for carrying out the ANOVA and multiple
6 m/s.
comparisons tests for the drag and the trim angle, assuming the speeds
to which the runs were carried out are the intended speeds, not the
measured speeds.

5.2. Experimental comparisons for drag and trim

Table 7 shows a summary of the statistics for all the ANOVAs

(and their respective assumptions) carried out for trim and drag at
the intended speeds. It can be observed that, according to the p-values
smaller than 0.05, the drag changes according to the hull shape starting
from 10 m/s up to 20 m/s. The same occurs for trim, but starting from
15 m/s.
In the cases where the p-values of the tests of normality and inde-
pendence of errors are smaller than 0.05, the corresponding QQ-plot,
or the histogram of the residuals, shows that there is not necessarily
a violation of the normality assumption. Something similar occurs for
the assumption of independence of errors.
Fig. 11. Filtered drag vs. speed results for the three hulls. Real-scale results. Fig. 14 shows the confidence intervals for the mean of the drag for
the three hulls for all the intended speeds.
Contrary to what had been observed with ANOVAs p-values for the
drag, in the Table 7, for 4 m/s, the Tukey-HSD test shows that there
is a difference between the treatments. This contradiction between the
ANOVA and the Tukey-HSD test may be due to low statistical power,
especially for the monohull and the Hysucat, whose number of repeats
are tree, while for the catamaran there are six. ANOVA and Tukey-
HSD results show that the p-values are very close to 0.05, so the most
conservative is to assume that there are no differences at all in the drag
between the treatments at this speed.
For a significance of 5%, the monohull statistically presents less drag
than the catamaran and the Hysucat for speeds of 10 m/s and higher,
except for 20 m/s in which the monohull is statistically equal to the
Hysucat. This result for 20 m/s may not be completely correct, due
to the fact that the measured speeds reached by the Hysucat and the
catamaran were lower than for the monohull.
Fig. 12. Filtered trim angle vs. speed results for the three hulls. Real-scale results. The catamaran and the Hysucat drags are statistically always the
same, except for a speed of 10 m/s, in which the drag of the catamaran
is about 1 kN lower. This shows that the hydrofoils brought no drag
distribution. Some of them present multimodal distributions or very reduction at any speed different than 10 m/s that could be detected
flattened normal distributions with outliers. with the present designed experiment. This result is surprising for

10
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 13. Violin plots for speed, drag, and trim angle (pitch) for the three boats at an intended speed of 15 m/s.

speeds of more than 10 m/s, in which the Hysucat enters in planing over 15 m/s. However, rejecting the second null hypothesis is due to
regime and the hydrofoils are supposed to generate enough lift to make the base geometry being a catamaran, not to the installed hydrofoils.
the boat partially emerge from the water surface.
The mean differences of drag between the monohull and both the 5.3. Regression analysis for drag and trim
catamaran and the Hysucat according to the Tukey-HSD test ranges
from 1 kN at speeds of 10 m/s to 5 kN at speeds of 20 m/s, which A polynomial linear regression analysis is performed for the drag
corresponds to 12% and 36% higher average drags in comparison to and the trim angle, versus the speed. This helps in comparing the hydro-
the average drag of the monohull at those corresponding speeds. This dynamic behavior between the three boats according to experimental
shows that the monohull has better planing capabilities in terms of results and also comparing against the results obtained with the Savit-
the hydrodynamic drag than the catamaran and the Hysucat. This sky model. When carrying out comparisons between the experimental
allows to reject the first null hypothesis stated in Section 3.3.2, Eq. (1), results in the regression analysis, it is important to take into account the
(catamaran-monohull) for speeds over 10 m/s, but does not allow to speeds from which the output variables start being statistically different
reject the null hypothesis in Eq. (3) (Hysucat-monohull). between treatments according to the results in the previous subsection.
Fig. 15 shows that, according to the ANOVAs (see Table 7), the Fig. 16 shows the linear regression analysis carried out for the drag
trim angle starts being statistically significantly different (higher) for with the mean values for all the runs. As expected, the regression lines
the monohull for speeds greater than 13 m/s. for the catamaran and the Hysucat are very similar for all the speeds.
The trim angles for the catamaran and the Hysucat are statistically It can be noted that the drag for the monohull starts growing at a
the same for all the speeds, which shows that the hydrofoils did not slower rate than the drag for the other two hulls at speeds beyond
cause a change in the travel position of the catamaran, not even at 10 m/s, which is near the transition threshold between the preplaning
planing speeds. Interestingly, the widest intervals of the mean for the and planing regime.
trim angle are presented for the three boats at speeds of 10 and 13 m/s The experimental results give much higher values than those pre-
even though they are not the highest speeds. This does not allow to dicted by the Savitsky method, except for speeds around 4 m/s where
find different statistical groups despite the drag differences that can be the values are similar. This difference between the Savitsky and ex-
identified starting from 10 m/s. perimental results grows as the speed increases. For the monohull, the
The mean difference between the monohull and both the Hysucat relative error ranges from 14% to 133%. For the case of the catamaran,
and catamaran ranges according to the Tukey-HSD tests from about 3 to the relative error ranges from 29% to 217%. Assuming the Savitsky
3.5◦ . This indicates that the monohull travels more inclined at planing model was properly implemented, this shows there were some error
speeds in a range of average trim angles between 4 and 6.5◦ , which is sources that affected the experiment, giving as a result, increased values
the recommended range by Savitsky et al. (1964). This allows to reject of hydrodynamic resistance.
both null hypothesis for the positional variables, stated in Section 3.3.2, Fig. 17 shows the linear regression analysis carried out for the trim
Eqs. (5) (catamaran-monohull) and (7) (Hysucat-monohull) for speeds angle with the mean values for all the runs. The trim angle for the

11
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Table 7
Summary of the statistics and p-values from the ANOVAs carried out and their respective assumptions for drag and trim
angle for intended speeds from 4 to 20 m/s. Only p-values are presented here, except for Durbin–Watson, whose statistic is
presented instead of the 𝑝-value.
Speed (m/s) Test name Drag Trim angle
Statistic p value Statistic p value
Shapiro - 0.08 - 0.52
Barlett - 0.07 - ( Not enough data)
4 Levene - 0.49 - 0.44
Durbin–Watson 1.97 – 2.71 –
ANOVA - 0.052 - 0.51
Shapiro - 0.44 - 0.74
Barlett - 0.36 - 0.68
6 Levene - 0.49 - 0.09
Durbin–Watson 2.83 – 1.83 –
ANOVA - 0.43 - 0.29
Shapiro - 0.19 - 0.003
Barlett - 0.52 - 0.17
10 Levene - 0.7 - 0.77
Durbin–Watson 2.06 – 2.27 –
ANOVA - 0 - 0.86
Shapiro - 0.79 - 0.08
Barlett - 0.46 - 0.17
13 Levene - 0.56 - 0.54
Durbin–Watson 2.31 – 2.16 –
ANOVA - 0.01 - 0.89
Shapiro - 0.33 - 0.73
Barlett - 0.007 - 0.5
15 Levene - 0.02 - 0.3
Durbin–Watson 2.07 – 2.14 –
ANOVA - 0.001 - 0.007
Shapiro - 0.052 - 0.26
Barlett - 0.01 - 0.03
17 Levene - 0.34 - 0.15
Durbin–Watson 1.28 – 2.6 –
ANOVA - 0 - 0
Shapiro - 0.53 - 0.93
Barlett - 0.1 - 0.31
20 Levene - 0.45 - 0.01
Durbin–Watson 2.61 – 1.81 –
ANOVA - 0.03 - 0

catamaran and Hysucat is almost identical for all the speeds. The trim Table 8
angle of the monohull starts diverging from the other two boats at a Results of the regression analysis for the experimental results of drag and trim angle.

speed of 10 m/s, giving support to the affirmations made in the multiple Hull shape Drag Trim angle
comparison tests for the trim. At planing speeds, both the catamaran r2 coef. p val. r2 coef. p val.
and Hysucat travel almost horizontally, with trim angles under 3.5◦ , 0 – 0 –
while the monohull travels more inclined with trim angles between 4 1.2809 0 0.3099 0.031
Monohull 0.992 0.972
and 5◦ . −0.093 0.017 0.0289 0.14
Similarly, for the case of the drag, trim angles derived from the 0.0014 0.081 −0.0017 0.017

regression analysis also present high deviations from the results of 0 – 0 –

the Savitsky model. For the monohull, there is 60% of relative error 1.277 0 0.7664 0.091
Catamaran 0.988 0.787
−0.0428 0.18 −0.0253 0.676
at lower speeds. These errors are lowered to around 18% at speeds
0.0015 0.196 −0.0007 0.731
over 12 m/s. For the case of the catamaran, the relative error is the
0 – 0 –
maximum at 20 m/s with a value of 100%; and the least at 10 m/s
0.965 0.002 0.7278 0.083
with a value of 22%. It is worth noting that even the shapes of the Hysucat 0.989 0.726
0.0227 0.592 −0.0258 0.679
experimental and Savitsky curves for the catamaran are very different −0.0012 0.43 −0.0006 0.805
from each other, even with apparently opposite behaviors in some
cases.
In contrast to the high coefficients of determination obtained for the
drag for the catamaran and the Hysucat, the coefficients of determina- from implementing a hypothesis test, where it is evaluated whether the
tion for the trim angle for these two are much lower with values of absence of the coefficient causes any effect in the equation. A 𝑝-value
0.787 and 0.726 respectively, which may be due to the much higher value of under 0.05 implies that the absence of the coefficient does
dispersion of the data points in comparison to the monohull. cause a change in the equation and is therefore important to describe
According to the regression analysis, the polynomials giving the the phenomenon. A 𝑝-value over this number does not necessarily mean
highest coefficients of determination for both drag and trim are degree
the opposite and may also indicate there is a low statistical power due
three. The polynomials are of the type:
to a lack of data.
𝑌 = 𝛽0 + 𝛽1 𝑋1 + 𝛽2 𝑋1 2 + 𝛽3 𝑋1 3 (9) It is observed that most of the coefficients present p-values over
Table 8 shows the coefficient of determination and the value of 0.05. However, if the regression is carried out for the drag without
each of these coefficients with their respective 𝑝-value, which results averaging for each run, all the p-values are reduced to almost zero as

12
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 14. Confidence intervals for the mean according to the Tukey-HSD test for the drag for the three hulls. Intended speeds from 4 to 20 m/s. The blue line is for the baseline
treatment or hull shape. A gray line means no statistical difference with the baseline treatment and the red line means the opposite. (For interpretation of the references to color
in this figure legend, the reader is referred to the web version of this article.)

can be seen in the Table 9, which means that there is a lack of data for similar behavior, with differences of around 10 cm at displacement
the case where the runs are averaged. and preplaning speeds. The curve for the catamaran follows a different
shape for most of the speeds. The three hulls have in common their
5.4. Results for sinkage, wetted area and wetted keel sinkage is reduced by around 40% due to the hydrodynamic lift caused
at planing speeds.
This is the opposite of what the Savitsky model shows for the
Fig. 18 shows side photos for the three models at cruising speed,
catamaran, which says its sinkage should slightly increase as the speed
which is at the intended speed of 5.3 m/s for the scale models. This is
increases. The differences for the catamaran between the experiments
equivalent to 15 m/s in the full-scale prototypes. There, it is observed and Savitsky are 10% at the cruise speed, 40% at the highest speeds,
that the monohull emerges completely from the water at the front, and 54% at the lowest speeds. For the case of the monohull, the
thanks to the fact that it travels with a higher backward inclination. differences range between 20% at the highest speeds to 70% at the
Additionally, it is observed that in the case of the monohull, the water lowest speeds and 92% at 10 m/s.
is deflected towards the sides, while for the catamaran and Hysucat the For scaling up the drag force from the scale models to the full-scale
water spray covers almost all the sides of the models. prototypes it was necessary to measure the wetted areas and wetted
Fig. 19 shows the experimental sinkage for the three hulls and the keel lengths of the models. Fig. 20 depicts the wetted areas at cruise
Savitsky model. Here, the monohull and the Hysucat seem to have speed for the CAD models of the full-scale monohull and catamaran.

13
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 15. Confidence intervals for the mean according to the Tukey-HSD test for the trim angle for the three hulls. Intended speeds from 4 to 20 m/s.

Fig. 16. Averaged drag vs. speed results for the three full-scale prototypes.

14
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 17. Averaged trim vs. speed results for the filtered values of the three full-scale prototypes.

Fig. 18. Lateral photograph of the three models at cruising speed (5.3 m/s).

Fig. 19. Sinkage vs speed for the three boats.

Fig. 21 summarizes the results for the wetted area for the three have on average twice the wetted area of the monohull, near the 40%
hulls. mentioned by Yun et al. (2018).
The wetted area for the Hysucat and the catamaran is almost the With respect to Savitsky, experimental wetted areas are very differ-
same for all the speeds, while for the monohull is greatly reduced ent, except for the monohull at speeds beyond 15 m/s. The differences
beyond planing speeds. The monohull has on average 18% less wetted for the catamaran are between 75% and 85%, while for the monohull
area than the other two hulls at displacement and preplaning speeds. range between 5% at 17 m/s to 32% at the lowest speeds, and 75% at
At planing speeds beyond 12 m/s, the catamaran and the Hysucat 10 m/s.

15
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 20. CAD representation of the wetted areas for the three vessels. Only one image is included for the catamaran and Hysucat because the shape of the wetted area is relatively
similar. ‘cat’ and ‘hys’ stand for catamaran and Hysucat respectively.

Fig. 21. Wetted area vs speed for the three models.

Table 9 The larger wetted area can be attributed to the fact that the water
Results of the regression analysis for the experimental results of drag has a contact at all times with the inner walls of the demihulls and
and trim angle. Runs without averaging.
the top of the tunnel. This contact with the top of the tunnel is an
Hull shape Drag (kN)
undesirable effect since from the conceptual design it was planned that
r2 coef. p val. this area would not have contact with water at cruising speed due to the
0 – high drag it could add up. This would be achieved by implementing the
1.1551 0
Monohull 0.986 hydrofoils. Additionally, the monohull has hard chines combined with
−0.0410 0
0.0008 0
spray deflectors that cause a separation of the flow from the hull and
guide it to the sides. This prevents downward hull suction from being
0 –
1.2597 0 caused (No Frills Sailing, 2021) and prevents the sides of the boat from
Catamaran 0.985 being in contact with water.
−0.0357 0
0.0011 0 In contrast, the catamaran and Hysucat, by not having sharp corners
0 – but rounded chines, allow the water to follow the contour by deflecting
Hysucat 0.987
0.8277 0 the water upwards, as it is presented in Fig. 23.
0.0427 0 This deflection may be causing a resultant downward force. Possi-
−0.0020 0
bly, this causes the hulls to be sucked in the same direction, eliminating
the hydrofoil effect in the case of the Hysucat. Additionally, this upward
deflection of the flow causes almost the entire lateral area of the
Similar to what happens for drag and trim, the wetted keel length of catamaran to remain wet according to Fig. 18.
the monohull starts being very different beyond 10 m/s in comparison This large additional wetted area for the catamaran and Hysucat
with the other two boats because it starts to decrease significantly, compared to the monohull may be the cause of the large difference in
while for the catamaran and the Hysucat it stays almost constant. The consumption between these vessels. One way to reduce this gap could
catamaran and the Hysucat travel almost horizontally, which makes the be by implementing spray deflectors (Russell, 2015) on the catamaran
wetted keel length to be almost equal to the total length of the boat for and varying the area and angle of attack of the hydrofoils, looking for
all the speeds. In contrast, the monohull reaches at planing speeds a the vessel to further reduce its sinkage at high speeds. These solutions
wetted keel length of almost half of its total length, as is observed in may reduce the drag of the catamaran and Hysucat, but it is uncertain
Fig. 22. how much this reduction might be.
This significant reduction in wetted area and wetted keel length due It could also be thought that the catamaran and Hysucat are trav-
to the higher trim angle at planing speeds in the case of the monohull, eling at pre-planing regimes instead of planing, which is causing the
partially explains that the drag force for the monohull is increased at a higher increase of the drag/speed rate beyond 6 m/s in comparison
lower rate in comparison to the other boats at planing speeds. to the monohull as is seen in Fig. 16. However, Froude numbers were

16
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

Fig. 22. Wetted keel length vs speed for the three boats.

Fig. 23. Graphic representation of the flow deflection sideways for the monohull and upwards for the catamaran and the Hysucat.

calculated from the wetted keels shown in Fig. 22, which shows all the boats at the intended speeds. The filtered data show, as expected,
three hulls go well beyond the planing threshold, which is around 1 normal distributions for most of the runs, especially for the speed and
and slightly below 10 m/s for the three hulls. drag. The trim angle, which was collected from the IMU, shows a more
sensible response to the instantaneous changes in speed during the
6. Conclusions runs, especially at speeds over 12 m/s for the real scale. Runs at the
maximum intended speed, which is 20 m/s for the real scale, show
Hydrofoil-supported catamarans are presented in this work as a high variations, especially for the speed and the trim angle, which
candidate to compete against monohulls in terms of hydrodynamic per- may be due to the limitation of power of the electric engine and
formance in contexts where installing hydrofoils that increases the draft the high vibrational loads exerted over the bench test. However, the
of the monohull is not technically feasible. Although planing boats and runs keep a symmetrical behavior that allows averaging the results to
hydrofoils have been widely studied through statistical, experimental, carry out the ANOVA and multiple comparison tests without incurring
and numerical approaches, the literature review allows concluding that pseudoreplication.
no studies comparing these geometries with experimental, statistical From the ANOVA and multiple comparison tests for the drag and
models or numerical methods have been at least published. trim it can be observed that there is no statistical difference between
The proposed methodology allows comparing experimentally any the catamaran and the Hysucat at any speed, which shows the hydrofoil
hull geometries at different speeds as long as dimensions, weight, and did not produce any change in the catamaran behavior, not at least
weight distribution are kept as similar as possible for all the geometries. for the level of variability presented. In contrast, the average drag for
The methodology allows to find statistical differences between the the catamaran and Hysucat starts being higher than for the monohull
hydrodynamic performance of multiple geometries at the same time, at speeds over 10 m/s, which is the speed of transition between the
so the reliability of the results heavily depends on the number of repe- displacement and planing regimes. The differences in average drag for
titions carried out per treatment or hull geometry at each speed. Here, the catamaran and Hysucat in comparison to the monohull start being
the Savitsky statistical model helps with the design of the experiment 12% higher at 10 m/s in comparison with the monohull and ends
and as a baseline for the experimental results. up being up to 36% higher at 20 m/s. These differences in average
As a case study for applying the methodology, a comparison be- drag between the catamaran and the monohull agree with what was
tween a monohull, a catamaran, and a Hysucat is carried out. The found in the literature review. However, the drag reduction objective
geometries were selected according to the context of Colombian rivers for the Hysucat was not achieved. These behaviors are reinforced by
and their particular characteristics. Seven intended speeds and a scale the obtained results with the linear regression analyses.
of 1/8 of the models in relation to the real scale prototypes were For the case of the positional variables, the monohull travels with a
selected. These speeds were selected in order to obtain acceptable reso- higher backward inclination at speeds higher than 15 m/s with average
lutions of the curves for the variables vs the speed. Limiting the compar- trim angles between 4 and 6.5◦ , which is the optimum according to the
ison of the hulls at the same speeds allows for reducing the complexity literature. In contrast, the catamaran and the Hysucat travel almost
of the design of experiments to a single-factorial experiment. horizontally with average trim angles between 1.5 and 2.5◦ at the
A bench test was developed for towing the model boats and col- same speeds. This higher backward inclination for the monohull allows
lecting the data. The experiment allowed for collecting variables such it to have around half of the wetted area at speeds over 15 m/s in
as speed, hydrodynamic drag, trim angle, sinkage, wetted area, and comparison to the catamaran and the Hysucat.
wetted keel area. Inferential statistical analyses were carried out for Furthermore, the wetted keel length for the monohull goes down
speed, drag, and trim angle. However, the other variables measured to 60% of its total length over 15 m/s, while the wetted keel lengths
provided valuable information about the dynamic positioning of the for the catamaran and the Hysucat have been kept to 90% of the

17
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

total length of the catamaran/Hysucat at speeds over 10 m/s. Other According to the literature, Hysucats have the potential to, at least,
factors that contribute to the higher wetted area for the catamaran and match the performance of monohulls. Hydrofoils could be implemented
the Hysucat are the inherent feature of the catamarans for having a in planing catamarans aimed to be operated in waters with harder
higher wetted area than the monohulls according to the literature (even operating conditions, such as Colombian rivers. Future work should
though this should have been reduced by installing the hydrofoils); be centered on improving the experiment setup and accuracy, carrying
and the fact that the round chines allow the water sprays to follow out experiments with boats at larger scales, and improving the cata-
the hull contour, which derives in the sides of the catamaran/Hysucat maran geometrical parameters so undesired effects that increase the
hull to be completely wetted as well. This much higher wetted area hydrodynamic resistance are avoided.
for the catamaran/Hysucat is, according to the theory, one of the
main responsible for the higher drag and general worse hydrodynamic CRediT authorship contribution statement
performance in comparison to the monohull at planing speeds.
With respect to the Savitsky results, very high differences were Erick Giraldo-Pérez: Conceptualization, Methodology, Software,
obtained in comparison to the experimental results especially for the Validation, Formal analysis, Investigation, Visualization, Resources,
drag, with relative errors ranging from 14 to 133% for the monohull Data curation, Writing – original draft. Esteban Betancur: Conceptu-
and 29 to 217% for the catamaran. For the trim angle, relative errors alization, Methodology, Software, Validation, Investigation, Resources,
range from 18 to 60% for the monohull and 22 to 100% for the cata- Writing – review & editing, Supervision. Gilberto Osorio-Gómez: Con-
maran. For sinkage, wetted area, and wetted keel length no different ceptualization, Methodology, Validation, Resources, Writing – review &
results are obtained, except for the wetted area for the monohull at editing, Supervision, Project administration, Funding acquisition.
speeds beyond 15 m/s, where absolute errors fall to less than 10%.
Assuming the Savitsky model was correctly implemented, the high Declaration of competing interest
differences in the experimental results in comparison to the Savitsky
results show some error sources affected the experiment, giving as The authors declare the following financial interests/personal rela-
a result increased values of hydrodynamic drag and different values tionships which may be considered as potential competing interests:
in dynamic positioning variables. However, these large discrepancies Erick Giraldo-Pérez reports financial support was provided by The
do not invalidate the experimental results in the sense they are a World Bank, which had no role in the study design (Collection, analysis
comparison exercise and do not plan to give ground-truth values for and/or interpretation of data), writing of the report or in the decision
the hydrodynamic performance. to submit the article for publication.

7. Future work Data availability

The data that has been used is confidential.

Some signals of low statistical power due to few runs carried out for
each treatment were identified, such as contradictions between the tests
Acknowledgments
to check the fulfillment of the ANOVA assumptions; a disagreement
between the ANOVA results and the Tukey-HSD results for one speed;
This research has been developed in the framework of the ‘‘ENER-
and the p-values of the coefficients for the linear regression being
GETICA 2030’’ Research Program, with code 58667 in the ‘‘Scientific
higher than 0.05. Carrying more runs (at least 10 per treatment or
Colombia’’ initiative, funded by The World Bank through the call ‘‘778-
hull geometry at each speed) would allow us to determine with a
201 Scientific Ecosystems’’, managed by the Colombian Ministry of
much higher statistical confidence whether the hydrofoils do cause a
Science, Technology and Innovation (Minciencias), with contract No.
change in the behavior of the catamaran at the expense of increasing
FP44842-210-2018.
the experiment cost and time.
Nonetheless, this would not change the fact that the monohull has
References
a much better hydrodynamic performance at planing speeds than the
Hysucat. For that to happen, the hydrofoils would have to reduce Arcadis Nederland, B., Jesyca, S., 2015. Plan maestro fluvial de Colombia 2015. Bogotá:
at least 36% of the catamaran hydrodynamic drag at speeds beyond Ministerio de Transporte de Colombia.
10 m/s. Another way to increase the statistical confidence and improve Broglia, R., Jacob, B., Zaghi, S., Stern, F., Olivieri, A., 2014. Experimental investigation
of interference effects for high-speed catamarans. Ocean Eng. 76, 75–85.
the quality of collected data would be to use a more robust test
Bullard, N., 2019. Electric car price tag shrinks along with battery cost.
infrastructure that resists the better the vibrational loads and where In: Bloomerg. URL: https://www.bloomberg.com/opinion/articles/2019-04-12/
external factors can be more certainly blocked. electric-vehicle-battery-shrinks-and-so-does-the-total-cost.
For solving the problem of the water wetting the laterals of the Calle, G.M., 2016. Un río difícil. El magdalena: historia ambiental, navegabilidad y
catamaran/Hysucat hull, unnecessarily increasing the wetted area of desarrollo. In: Memorias. Revista Digital de Historia Y ArqueologíA Desde El Caribe.
(28), Universidad del Norte, pp. 29–60.
these geometries, it could be installed spray deflectors or hard chines Cames, M., Graichen, J., Siemons, A., Cook, V., 2015. Emission Reduction Targets for
similar to the monohull in order to deflect these water streams towards International Aviation and Shipping. Policy Department A: Economic and Scientific
the laterals. The wetted area could be further decreased by tuning some Policy, European Parliament, B-1047 Brussels.
hydrofoil parameters such as the angle of attack and plan area. These Candela speedboat, 2018. Candela speedboat: the journey. URL: https:
//candelaspeedboat.com/company/.
solutions might decrease the Hysucat drag at planing speeds, but it is
Clement, E.P., Blount, D.L., 1963. Resistance Tests of a Systematic Series of Planing
uncertain how much this reduction might be, and, again, the reduction Hull Forms. David Taylor Model Basin, Department of the Navy, Research and
should be more than 36% for it to be lower than the drag of the Development Report, Report 1314, Hydromechanics Laboratory, Washington DC,
monohull. USA, Published By: The Society of Naval Architects and Marine Engineers, Paper
No. 10.
Another parameter that may be affecting the result of the exper-
Comission, T.E., 2017. Work programme 20182020: 11. In: Smart, Green and Integrated
iment is the boat scale. Hydrodynamic behaviors such as the water Transport, LC-MG-1-11-2019: Structuring R&I Towards Zero Emission Waterborne
viscosity are hard if not impossible to scale, and the smaller the scale, Transport. Vol. 27. pp. 2018–2020.
the harder to achieve kinetic and dynamic similitudes. The performance Davidson, G., Roberts, T., Friezer, S., Davis, M., Bose, N., Thomas, G., Binns, J.,
of Hydrofoils (and lifting surfaces in general) is highly dependent on the Verbeek, R., 2011. Maximising efficiency and minimising cost in high speed craft.
In: International Conference on Fast Sea Transportation. Vol. 11. pp. 727–734.
viscous properties of the fluid involved. That is why the small scale may Day, A., Cameron, P., Nixon, E., 2017. Moderate-cost approaches for hydrodynamic
have played a key role in the hydrofoils not causing a difference in the testing of high performance sailing vessels. In: INNOVSAIL International Conference
hydrodynamic behavior of the Hysucat with respect to the catamaran. on Innovation in High Performance Sailing Yachts. pp. 173–182.

18
E. Giraldo-Pérez et al. Ocean Engineering 262 (2022) 112227

De Luca, F., Pensa, C., 2017. The naples warped hard chine hulls systematic series. Najafi, A., Aliakbari, T., Hashemi, S.A., 2019. Experimental optimization of hydrody-
Ocean Eng. 139, 205–236. namic performance of catamarans using hydrofoil element. Proc. Inst. Mech. Eng.
Ekman, F., Ribas, C.N., Rydelius, F., 2016. Model for Predicting Resistance and M J. Eng. Marit. Environ. 233 (2), 488–501.
Running Attitude of High-Speed Craft Equipped with Interceptors. Royal Institute Najafi, A., Nowruzi, H., 2019. On hydrodynamic analysis of stepped planing crafts. J.
of Technology, Stockholm. Ocean Eng. Sci. http://dx.doi.org/10.1016/j.joes.2019.04.007, URL: http://www.
Faltinsen, O.M., 2005. Hydrodynamics of High-Speed Marine Vehicles. Cambridge sciencedirect.com/science/article/pii/S2468013319300737.
University Press. No Frills Sailing, 2021. Hard chines in sailboat design. URL: http://no-frills-sailing.
Fridsma, G., 1969. A Systematic Study of the Rough-Water Performance of Planing com/hard-chines-in-sailboat-design/.
Boats. Technical Report, Stevens Inst Of Tech Hoboken Nj Davidson Lab. Ozden, M.C., Demir, E., 2009. The successful design and construction of solar/electric
Frisk, D., Tegehall, L., 2015. Prediction of high-speed planing hull resistance and boats nusrat and muavenet: An overview. In: Ever Monaco 2009 Conferences on
running attitude-a numerical study using computational fluid dynamics. Ecological Vehicles and Renewable Energies. Citeseer, pp. 27–29.
Fu, T., Brucker, K., Mousaviraad, S., Ikeda, C., Lee, E., O’shea, T., Wang, Z., Stern, F., solar boat race, T., 2018. A brief history of solar boat racing. URL: http://
Judge, C., 2014. An assessment of computational fluid dynamics predictions of torontosolarboat.com/solar-boat-racing/.
the hydrodynamics of high-speed planing craft in calm water and waves. In: 30th Restrepo, J.D., Escobar, R., Tosic, M., 2018. Fluvial fluxes from the magdalena river
Symposium on Naval Hydrodynamics. pp. 2–7. into cartagena bay, caribbean Colombia: Trends, future scenarios, and connections
Gerdsen, H.G., Jabbusch, W., 1987. Foil arrangement for a planning craft. Google with upstream human impacts. Geomorphology 302, 92–105.
Patents, US Patent 4, 665, 853. Russell, J., 2015. Spray rails and lifting strakes. URL: https://aeromarineresearch.com/
Giraldo-Pérez, E., Gaviria, G., Betancur, E., Osorio-Gómez, G., Mejía-Gutiérrez, R., 2020. tbdp6/strakes_and_sprayrails.html.
Influence of energy consumption on battery sizing of electric fluvial vessels: a Savitsky, D., Morabito, M., 2010. Surface wave contours associated with the forebody
Colombian case study. In: 2020 Fifteenth International Conference on Ecological wake of stepped planing hulls. Mar. Technol. 47 (1), 1–16.
Vehicles and Renewable Energies. EVER, IEEE, pp. 1–8. Savitsky, D., et al., 1964. Hydrodynamic design of planing hulls. Mar. Technol. SNAME
Higgins, A., Restrepo, J.C., Ortiz, J.C., Pierini, J., Otero, L., 2016. Suspended sediment News 1 (04), 71–95.
transport in the magdalena river (Colombia, south America): Hydrologic regime, Seif, M.S., Amini, E., 2004. Performance comparision between planing monohull and
rating parameters and effective discharge variability. Int. J. Sediment Res. 31 (1), catamaran at high froude numbers. Citeseer.
25–35. Sherman, T., Fisher, P.A., 1975. A Study of Planing Catamaran Hull and Tunnel
Holtrop, J., Mennen, G., 1982. An approximate power prediction method. Int. Shipbuild. Interactions. Technical Report, Michigan Univ Ann Arbor Ship Hydrodynamics Lab.
Prog. 29 (335), 166–170. Skene, N.L., 1973. Skene’s Elements of Yacht Design, eighth ed. Dodd, Mead.
Honaryar, A., Ghiasi, M., Liu, P., Honaryar, A., 2021. A new phenomenon in Solar Splash, 2012. The 2012 DONG enery solar challenge. URL: https:
interference effect on catamaran dynamic response. Int. J. Mech. Sci. 190, 106041. //solarsplash.com/introduction-and-information/history/solar-splashs-dong-
Hoppe, K.-G.W., 1986. Catamaran with hydrofoils. Google Patents, US Patent 4, 606, energy-solar-challenge-updates/.
291. Souto-Iglesias, A., Fernández-Gutiérrez, D., Pérez-Rojas, L., 2012. Experimental as-
Hoppe, K.-G., 1989. The HYSUCAT Development. Department of Mechanical sessment of interference resistance for a series 60 catamaran in free and fixed
Engineering, University of Stellenbosch. trim-sinkage conditions. Ocean Eng. 53, 38–47.
Hoppe, K., 2001. Recent applications of hydrofoil-supported-catamarans. In: Fast Ferry Sponberg, E.W., 2011. The design ratios. A Naval Architect’s Dozen (Or Thereabouts).
International. Vol. 36. Citeseer. Revised.
Hu, J., Zhang, Y., Wang, P., Qin, F., 2020. Numerical and experimental study on Sukas, O.F., Kinaci, O.K., Cakici, F., Gokce, M.K., 2017. Hydrodynamic assessment of
resistance of asymmetric catamaran with different layouts. Brodogradnja 71 (2), planing hulls using overset grids. Appl. Ocean Res. 65, 35–46.
91–110. Tank, I.T., 2008. Recommended procedures and guidelines 7.5-02-05-01: Testing and
Kandasamy, M., Ooi, S.K., Carrica, P., Stern, F., Campana, E.F., Peri, D., Osborne, P., extrapolation methods high speed marine vessels resistance test. URL: https://ittc.
Cote, J., Macdonald, N., de Waal, N., 2011. CFD validation studies for a high-speed info/media/4114/75-02-05-01.pdf.
foil-assisted semi-planing catamaran. J. Mar. Sci. Technol. 16 (2), 157–167. Thomas, G., Tomic, P., Tuite, A., 2007. High-speed catamaran or monohull? How do
Khazaee, R., Rahmansetayesh, M.A., Hajizadeh, S., 2019. Hydrodynamic evaluation of you choose? Ships Offshore Struct. 2 (2), 137–147.
a planing hull in calm water using RANS and Savitsky’s method. Ocean Eng. 187, Tuck, E.O., Lazauskas, L., 1998. Optimum hull spacing of a family of multihulls. Ship
106221. Technol. Res. Schiffstechnik 45 (4), 180.
Kim, D.J., Kim, S.Y., You, Y.J., Rhee, K.P., Kim, S.H., Kim, Y.G., 2013. Design of high- Union Internationale des Transports Publics, 2013. Waterborne Transport, A Unique
speed planing hulls for the improvement of resistance and seakeeping performance. Contribution to Enhancing Mobility for Cities on Water. Union Internationale
Int. J. Nav. Archit. Ocean Eng. 5 (1), 161–177. des Transports Publics, Brussels, URL: https://www.uitp.org/sites/default/files/cck-
L. Blount, D., L. Fox, D., 1976. Small-craft power prediction. Mar. Technol. 13, 14–45. focus-papers-files/fp{_}waterborne-en.pdf.
Latorre, R., 1997. Ship hull drag reduction using bottom air injection. Ocean Eng. 24 Vellinga, R., 2009. Hydrofoils: Design, Build, Fly. Peacock Hill Publishing, URL: https:
(2), 161–175. //books.google.com.co/books?id=UhQsQQAACAAJ.
Lee, T.-S., 1982. Interference factor for catamaran planing hulls. AIAA J. 20 (10), Wyman, D.B., 1998. Wyman’s formula. In: Professional Boatbuilder. URL: https://
1461–1462. pbbackissues.advanced-pub.com/?issueID=54{&}pageID=99.
Liu, C.Y., Wang, C.T., 1979. Interference effect of catamaran planing hulls. J. Yengejeh, M.A., Amiri, M.M., Mehdigholi, H., Seif, M.S., Yaakob, O., 2016. Numerical
Hydronaut. 13 (1), 31–32. study on interference effects and wetted area pattern of asymmetric planing
Mira, J.-D., Valderrama, S., Londoño, M.-J., Giraldo-Pérez, E., Betancur, E., Osorio- catamarans. Proc. Inst. Mech. Eng. M J. Eng. Marit. Environ. 230 (2), 417–433.
Gómez, G., Mejía-Gutiérrez, R., 2020. Preliminary design tools applied to a Yousefi, R., Shafaghat, R., Shakeri, M., 2013. Hydrodynamic analysis techniques for
solar powered vessel design: a south American river analysis. In: 2020 Fifteenth high-speed planing hulls. Appl. Ocean Res. 42, 105–113. http://dx.doi.org/10.
International Conference on Ecological Vehicles and Renewable Energies. EVER, 1016/j.apor.2013.05.004, URL: http://www.sciencedirect.com/science/article/pii/
IEEE, pp. 1–9. S0141118713000394.
Molland, A.F., Turnock, S.R., Hudson, D.A., 2017. Ship Resistance and Propulsion. Yousefi, R., Shafaghat, R., Shakeri, M., 2014. High-speed planing hull drag reduction
Cambridge University Press. using tunnels. Ocean Eng. 84, 54–60.
Morabito, M.G., 2011. Experimental investigation of the lift and interference of Yun, L., Bliault, A., 2012. Hydrofoil craft. In: High Performance Marine Vessels.
asymmetric planing catamaran demi-hulls. In: Proceedings of the 11th International Springer, pp. 161–202.
Conference on FAST2011. Yun, L., Bliault, A., Rong, H.Z., 2018. High Speed Catamarans and Multihulls:
Müller-Graf, B., 1989. Resistance and Propulsion Characteristics of the VWS Hard Chine Technology, Performance, and Applications. Springer.
Catamaran Hull Series 89. Versuchsanstalt FÜR Wasserbau Und Schiffbau, Berlin.

19