Alexandria University Naval Architecture &

Faculty of Engineering Marine Engineering Department

TABLE OF CONTENTS

Resistance & Powering Prediction.................................................................................................. 2

Resistance Components .............................................................................................................. 2
Friction Resistance & Viscous Resistance .............................................................................. 3
Wave Making Resistance ........................................................................................................ 6
Model 1 resistance & Powering .................................................................................................. 9
Analysis Using Regression Based Methods ........................................................................... 9
Guldhammer And Harvald Method ........................................................................................ 9
Limits And Required Inputs ................................................................................................ 9
Equations........................................................................................................................... 10
Results ............................................................................................................................... 12
Shallow Water Correction ..................................................................................................... 15
Wave Pattern in Shallow Water ......................................................................................... 15
Schlichting Correction Method ......................................................................................... 17
Correction for Wave Retardation Effect ....................................................................... 17
Correction for Back Flow Effect ................................................................................... 18
Howe’s Shallow Water Formula ....................................................................................... 18
Model 1 Shallow Water Correction Results ...................................................................... 19
Model 1 Resistance Prediction Using CFD Analysis ........................................................... 21
Prerequisites ...................................................................................................................... 21
Computational Domain Size & Boundary Conditions.................................................. 21
Wave Damping Functions ............................................................................................. 22
Prism Layers Calculations ............................................................................................ 23
Mesh Generation ............................................................................................................... 24
Simulation Physical Setup ................................................................................................ 28
Results & Discussions....................................................................................................... 32
Total Resistance Result ................................................................................................. 36
Model 1 Resistance & Powering Summary .......................................................................... 37

Mohamed A. Abdelrazek,
Resistance & Powering 1
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

RESISTANCE & POWERING PREDICTION

When a ship of length 𝐿𝑊𝐿 moves forward in the vicinity of the free surface of water at a
constant velocity 𝑉𝑆, then its forward motion generates:
• Dynamic pressure on the immersed (wetted) surface producing a resultant force in the
longitudinal direction and opposite to the advancing direction.
• Tangential stresses on the immersed (wetted) surface due to the viscosity; their resultant
force is also opposite to the ship’s moving direction.
The total force opposite to the motion is called the resistance of the ship or drag. Each of the
resistance components most concerned arise from one of the two forces: namely normal dynamic
pressures and tangential stresses on the ship’s wetted surface.
When a body moves through a fluid it experience forces opposing the motion. As a ship moves
through water and air it experience both water and air forces. These will, in general, be of
different magnitudes and directions. The resistance is studied initially in still water with no wind.
The water resistance is the dominant factor in determining the speed achieved. Because 𝝆𝒂 ≪≪
𝝆𝒘, the air resistance of the conventional ships (merchant, service, pleasure, …) is usually much
smaller than the water resistance, except for those aero statically supported crafts.

RESISTANCE COMPONENTS
The fluid is viscous, and a deeply immersed body would suffer a frictional drag, in addition,
when the body approaches a free surface, the pressure variations around the body can manifest
themselves as elevations or depressions of the water surface. That is to say, waves are formed on
the surface. This process upsets the balance of pressure acting on the body which results in a
drag force. The magnitude of the drag force is related to the energy of the wave system created.
The total resistance of a ship moving on a calm water surface has several components. They are
skin friction resistance; residuary resistance; wave-making resistance; eddy-making resistance;
air resistance; appendage resistance.
The components with the most interest are the viscous resistance and the wave making
resistance.
𝑅𝑇 = 𝑅𝑓 + 𝑅𝑅
𝑅𝑇 = 𝑅𝑉 + 𝑅𝑤
The rest of the components can be taken as a percentage of the total resistance RT.
Each component can now be studied separately provided it is remembered that each will have
some interaction with the others.

Mohamed A. Abdelrazek,
Resistance & Powering 2
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 1 Resistance components and relations.

Friction Resistance & Viscous Resistance

Water is viscous and the conditions for dynamic similarity are geometric similarity and
constancy of Reynolds' number. Due to the viscosity the particles immediately adjacent to the
hull adhere to it and move at the speed of the ship. At a distance from the hull the water is at rest.
There is a velocity gradient which is greatest close to the hull. The volume of water which moves
with the body is known as the boundary layer. Its thickness is usually defined as the distance
from the hull at which the water velocity is 1 per cent of the ship speed.

Figure 2 Boundary layer demonstration.

Mohamed A. Abdelrazek,
Resistance & Powering 3
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Frictional resistance is associated with Reynolds. There are two distinct types of flow. In the
first, laminar flow, each fluid particle follows its own streamlined path with no mass transfer
between adjacent layers. This flow only occurs at relatively low Reynolds' numbers. At higher
numbers, the steady flow pattern breaks down and is replaced by a more confused flow pattern
called turbulent flow. There are two flow regimes, the change from one to the other depends on
the critical Reynolds' number and different resistance laws apply.
Frictional resistance 𝑅𝐹𝑠 is 80 ~ 85% of the total resistance 𝑅𝑇𝑠 in slow speed ships and as much
as 50% in high-speed ships.
Friction resistance has gone through various research and development until it reaches this
formula:
0.075
𝐶𝑓 =
(log(𝑅𝑒 ) − 2)2

Which is known as ITTC 57 Equation which considers turbulent flows.

Where:
𝑅
• 𝑐𝑓 (𝐹𝑟𝑖𝑐𝑡𝑖𝑜𝑛𝑎𝑙 𝑅𝑒𝑠𝑖𝑠𝑡𝑎𝑛𝑐𝑒 𝐶𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡) = 0.5𝜌 𝑓𝑆 𝑉 2
𝑆
• 𝜌: 𝑊𝑎𝑡𝑒𝑟 𝐷𝑒𝑛𝑠𝑖𝑡𝑦.
• 𝑆: Wetted Surface Area.
• 𝑉: Ship Velocity.
𝑉 𝐿
• 𝑅𝑒𝑦𝑛𝑜𝑙𝑑𝑠 𝑁𝑢𝑚𝑏𝑒𝑟 𝑅𝑒 = 𝑆𝜈

The main drawback here is that friction resistance only considers the skin friction only regardless
of the shape of the hull itself; two ships may have the same friction resistance, but their shape is
different.
This problem will lead to another form to define the resistance related to both skin friction and
hull form together which is viscous resistance.
For a ship moving in a viscous fluid, a turbulent boundary layer is created over its surface and is
also likely to separate at some point in the after body.

Figure 3 Hull shape effect on flow separation.

Mohamed A. Abdelrazek,
Resistance & Powering 4
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

The presence of the boundary layer and its growth along the surface of the hull modifies the
pressure distribution acting on the ship from that of the potential or inviscid case.
Due to the viscosity, the pressure change will cause flow separation from the surface and
generates eddies.
As a consequence, the net axial force in the direction of ship motion is not zero and a viscous
form drag arises for the 3D case of a ship hull moving steadily in a viscous fluid.
Energy is fed into eddies; this energy is consumed from the ship itself, and the resulting
resistance is called eddy-making resistance or viscous pressure resistance.
The curvature of a hull changes the local fluid velocity along its length, as the path along a
streamline running from bow to stern is longer on a ship-shaped body than on a flat plate so a
new coefficient is developed to represent the hull shape called the form factor (1+k).
Now viscous resistance can be represented as the product of friction resistance and the form
factor:
𝑅𝑉 = 𝑅𝑓 + 𝑅𝑉𝑃
∴ 𝑅𝑉 = 𝑅𝑓 + 𝑘𝑅𝑓
∴ 𝑅𝑉 = (1 + 𝑘)𝑅𝑓
∴ 𝑐𝑉 = (1 + 𝑘)𝑐𝑓
Form factor can increase accuracy where no bulbous bow and no immersed transom stern is
present.
Usually, the form factor is obtained from Prohaska test or using statistically based equation like
Holtrop form factor formula.
Holtrop Formula:
(1 + 𝑘𝑁) = 0.93 + 0.4871 ∙ 𝑐 ∙ 𝛼 1.0681 ∙ 𝛽 0.4611 ∙ 𝛾 0.1216 ∙ 𝛿 0.3649 ∙ (1 – 𝐶𝑃) −0.6042
1/𝛾 = 1 − 𝐶𝑃 + 0.06 𝐶𝑃 ∙ (𝐿𝐶𝐵 / (4 𝐶𝑃−1.0))
Where:
𝛼: breadth to length ratio, 𝛼 = 𝐵/𝐿.
𝛽: draft to length ratio, 𝛽 = 𝑇/𝐿.
𝛾: ship length to run length ratio, 𝛾 = 𝐿/𝐿𝑅.
𝛿: cubic length to displacement volume ratio, 𝛿 = 𝐿3 /𝛻.
𝑐: stern shape factor, 𝑐 = 1.00, 1.11, 0.89 for normal stern shape, U-shaped section, V-shaped
section, respectively.
𝐿𝐶𝐵: location of longitudinal center of buoyancy expressed as percentage of ship length.

Mohamed A. Abdelrazek,
Resistance & Powering 5
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Wave Making Resistance

A body moving on an otherwise undisturbed water surface creates a varying pressure field
which manifests itself as waves because the pressure at the surface must be constant and equal to
atmospheric pressure.
From observation when the body moves at a steady speed, the wave pattern seems to remain the
same and move with the body. With a ship the energy for creating and maintaining this wave
system must be provided by the ship's propulsive system. Put another way, the waves cause a
drag force on the ship which must be opposed by the propulsor if the ship is not to slow down.
This drag force is the wave-making resistance.
The nature of the wave system created by a ship is similar to that which Kelvin demonstrated for
a moving pressure point. Kelvin showed that the wave pattern had two key features: diverging
waves on each side of the pressure point with their crests inclined at an angle to the direction of
motion and transverse waves with curved crests intersecting the centerline at right angles. The
angle of the divergent waves to the centerline is sin"1/3, which is just under 20°.

Figure 4 wave system associated with moving pressure point.

The waves move with the ship so the length of the transverse waves must correspond to this
speed, that is their length is 2π(V)^2/g, the pressure field around the ship can be approximated by
a moving pressure field close to the bow and a moving suction field near the stern.
Both the forward and after pressure fields create their own wave system as shown in the figure
below. The after field being a suction one creates a trough near the stern instead of a crest as is
created at the bow.
The angle the divergent waves to the centerline will not be exactly that of the Kelvin wave field,
the maximum crest heights of the divergent waves do lie on a line at an angle to the centerline
and the local crests at the maxima are at about twice this angle to the centerline. The stern
generated waves are less clear, partly because they are weaker, but mainly because of the
interference they suffer from the bow system.

Mohamed A. Abdelrazek,
Resistance & Powering 6
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 5 Bow and stern wave systems.

In some ships, the wave pattern may be made even more complex by the generation of other
wave system by local discontinuities in the ship’s form.
Since at most speed both the bow and stern systems are present aft of the ship, there is
interaction between the two transverse wave system:
• If the systems are so phased that the crest are coincident, the resulting system will have
increased wave height.
• If the crest of one system coincident with the trough of the other the resulting wave
height and energy will be less.
Froude studied the effect on the resistance of the length of the ship by towing models with the
same endings but with varying lengths of parallel middle body, the results are in line with what
could be expected from above general reasoning; The distance between bow and stern pressure
systems is typically 0.9L.
The condition that crests or troughs of the bow system should coincide with the first trough of
the stern system is:
𝑉2 𝑔
=
0.9𝐿 𝑁𝜋
Therefore:
For N=1,3,5,7, etc. the troughs will coincide and for N=2,4,6, etc. the crests from the bow system
coincide with the trough from the stern system as in figure below.
If there were no interaction between the bow and stern wave systems, the resistance would
increase steadily with speed as shown in fig.4. Because interaction occurs at speed discussed
above, the actual resistance curve will oscillate about the curve as indicated.
Mohamed A. Abdelrazek,
Resistance & Powering 7
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 6 Interaction of bow and stern wave systems.

Figure 7 Humps and Hollows in wave making resistance curve.

A hump occurs when N is an odd integer and a hollow when N is even integer. It is to be
expected that the most pronounced hump will be at N=1 because the speed is highest for this
condition and this hump is usually referred to as the main hump. The hump associated with N=3
is often called the prismatic hump.

Mohamed A. Abdelrazek,
Resistance & Powering 8
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

MODEL 1 RESISTANCE & POWERING

Analysis Using Regression Based Methods

Regression based methods will be used in this section to estimate the hull resistance and the
required thrust power.
Data on resistance and powering requirements are required in the ship design process long before
a model test can be conducted, therefore, methods have been developed which allow the naval
architect to estimate resistance and powering requirements based on ship speed and a few
principal geometric characteristics.
It must be noted that a margin of 25 to 30 percent can be added to the final result.

Guldhammer And Harvald Method

The method is based on calculating the residuary resistance coefficient CR then making suitable
corrections to achieve the final CR and CF is calculated using the ITTC57 equation.
CT = CR + CF + CA + CAA
Where:
𝐶𝐴: stands for a combined correlation and roughness allowance.
𝐶𝐴𝐴 is the air resistance coefficient.
The method was developed in the late 60s and early 70s when most ships did not yet have a
bulbous bow. In later publications, Harvald (1983) and Andersen and Guldhammer updated the
procedure and introduced a computation length 𝐿 which made the method applicable to ships
with bulbous bows. Andersen and Guldhammer (1986) provide regression formulas to replace
the traditional 𝐶𝑅-charts and extend the method to include powering estimates.

Limits And Required Inputs

The method is applicable to single and twin-screw displacement type vessels, based on the
regression formulas for the residuary resistance, limits of application for the method are shown
in the table below with Model 1 Characteristics given from MAXSURF Modeler:

Mohamed A. Abdelrazek,
Resistance & Powering 9
 Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Table 1 Limits & Inputs

Particulars Based on Calculations & Assumptions

Model 1 Guldhammer & Harvald
Notes
Data Standards
Lwl (m) 50 - -
Vol (m^3) 418 - -
Cb 0.836 0.55 to 0.85 Applicable
Ax (m^2) 9.8 - -
Cm 0.98 - -
Cp 0.853061224 -
Los (m) 50 - -
Fn 0.14 < 0.33 Applicable
L/B 5 5 to 8 Applicable
B/T 10 2.5 Correction is Needed
L/Vol^ (1/3) Also Called: M 6.687204078 4 to 6 Correction is Needed
S (m^2) 520 - -
Bulb No Bulb No Bulb No Correction is Needed
LCB (As Percentage) Assumed Optimum with Respect to Fn No Correction is Needed
Bow Form Moderate Moderate
Correction is Needed
Stern Form Extreme U Moderate

Equations
Andersen and Guldhammer (1986) stated the following regression formula for a standard
residuary resistance coefficient 𝐶𝑅std:
103 𝐶𝑅std = 𝐸 + 𝐺 + 𝐻 + 𝐾

The four contributions depend on Froude number 𝐹𝑟, length–displacement ratio 𝑀, and prismatic
coefficient 𝐶𝑃:
𝑵 𝟐. 𝟓
𝑬 = (𝑨𝟎 + 𝟏. 𝟓 𝑭𝟏.𝟖 𝟏
𝒓 + 𝑨𝟏 𝑭𝒓 ) . [𝟎. 𝟗𝟖 + ] + (𝑴 − 𝟓)𝟒 . (𝑭𝒓 − 𝟎. 𝟏)𝟒
(𝑴 − 𝟐)𝟒
With Variables:
𝐴0 = 1.35 − 0.23 𝑀 + 0.012 𝑀2
𝐴1 = 0.0011 𝑀9.1
𝑁1 = 2 𝑀 − 3.7
𝐿
𝑀= 3
√𝛻

Mohamed A. Abdelrazek,
Resistance & Powering 10
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

(𝟕 − 𝟎. 𝟎𝟗 𝑴𝟐 ) (𝟓 𝑪𝑷 − 𝟐. 𝟓)𝟐
𝑮=
[𝟔𝟎𝟎 (𝑭𝒓 − 𝟎. 𝟑𝟏𝟓)𝟐 + 𝟏]𝟏.𝟓

𝑯 = ⅇ𝑯𝟏 Where: 𝐻1 = 80 [𝐹𝑟 − (0.04 + 0.59 𝐶𝑃 ) − 0.015 (𝑀 − 5)]

(𝟐𝟎𝑪𝒑 −𝟏𝟔)
𝑲 = 𝟏𝟖𝟎𝑭𝟑.𝟕
𝒓 ⅇ

The residuary resistance coefficient𝐶𝑅std is valid only for what was considered a standard hull
form in the 1960s:
I. Beam–draft ratio of 𝐵∕𝑇 = 2.5.
II. Location of 𝐿𝐶𝐵 at its optimal position.
III. No bulbous bow.
IV. No appendages like bossing’s, struts.
For ship hulls that deviate from the standard shape, five corrections have been introduced:
I. Δ𝐶𝑅(𝐵∕𝑇) for beam–draft ratios other than 2.5.
II. Δ𝐶𝑅(𝐿𝐶𝐵) for 𝐿𝐶𝐵 positions deviating from optimum.
III. Δ𝐶𝑅(form) for section shapes at bow and stern that deviate from normal hull forms.
IV. Δ𝐶𝑅(bulb) for bulbous bows.
V. Δ𝐶𝑅(app) for appendages other than the typical single rudder and bilge keels.
The adjusted residuary resistance coefficient will be:
𝐶𝑅 = 𝐶𝑅std + Δ𝐶𝑅(𝐵∕𝑇) + Δ𝐶𝑅(form) + Δ𝐶𝑅(bulb) + Δ𝐶𝑅(𝐿𝐶𝐵) + Δ𝐶𝑅(app)

And in case of Model 1, with B/T > 2.5, modified stations shape, no bulb, optimum LCB and no
other appendages, only two corrections are needed: Δ𝐶𝑅(𝐵∕𝑇) and Δ𝐶𝑅(form):
For vessels with 𝐵∕𝑇 ≠ 2.5 a simple speed independent correction is applied:
103 Δ𝐶𝑅(𝐵∕𝑇) = 0.16 (𝐵/𝑇 − 2.5)

A standard hull form has neither pronounced V- nor pronounced U-shaped stations in entrance
and run of the hull, Guldhammer and Harvald (1974) assigned fore and aft body shape factors 𝐹𝐹
and 𝐹𝐴 between −3 and +3, respectively; the value −3 is for extreme V-shaped stations, 0 for
normal stations, and +3 for extreme U-shaped stations.

Mohamed A. Abdelrazek,
Resistance & Powering 11
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Once values have been assigned to 𝐹𝐹 and 𝐹𝐴, a correction to residuary resistance coefficient is
computed from:
103Δ𝐶𝑅(form) = − 0.1 (𝐹𝐹 − 𝐹𝐴) /3

Correlation allowance CA is based on Holtrop 78 formula:

For 𝑇𝐹/𝐿𝑊𝐿 < 0.04 with no bulb:
𝐶𝐴 = 0.006 ∙ (𝐿𝑊𝐿 + 100) −0.16 − 0.00205 = 0.00064139 For Model 1

Air resistance coefficient 𝐶𝐴𝐴 is assumed to be constant in Andersen and Guldhammer (1986),
the suggested value is:
103 𝐶𝐴𝐴 = 0.07

Results
All these equations are implemented in an Excel spreadsheet with Model 1 inputs and results are
shown in the next page.

Mohamed A. Abdelrazek,
Resistance & Powering 12
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Total Resistance VS Speed

200

180

160

140

120
Rt (KN)

100

80

60

40

20

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Speed (Knot)

Deep Water

Figure 8 Model 1 Deep Water Resistance Speed Curve.

Mohamed A. Abdelrazek,
Resistance & Powering 13
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Power VS Speed
800

700

600

500
Power (KW)

400

300

200

100

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Speed (Knot)

Deep Water

Figure 9 Model 1 Deep Water Power Speed Curve.

Mohamed A. Abdelrazek,
Resistance & Powering 14
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Shallow Water Correction

The resistance of a ship is quite sensitive to effects of shallow water. The potential flow around
the hull changes appreciably. If the ship is considered as being at rest in a flowing stream of
restricted depth, but unrestricted width, the water passing below it must speed up more than in
deep water, with a consequent greater reduction in pressure and increased sinkage, trim and
resistance. Furthermore, if it is restricted laterally, as in a river or canal, these effects are
exaggerated. The sinkage and trim in very shallow water can set an upper limit to the speed at
which ships can operate without grounding. Also, the wave pattern created by the moving ship
changes in shallow water, leading to a change in the wave making resistance.

As the ship enters water of restricted depth, termed shallow water, a number of changes occur
due to the interaction between the ship and the seabed:
1. Due to constriction between the hull bottom and sea/riverbed, potential flow velocity increases
between hull bottom and sea/riverbed. The consequent reduction in pressure around the hull
leads to reduction in buoyancy and change in center of buoyancy. The ship is then subjected to
sinkage and change in trim.
2. Increase in total water resistance. Increase in potential flow velocity causes increase in
frictional resistance. Change in wave pattern causes increase in wave making resistance at sub
critical speeds.
If in addition the water is restricted laterally, as in a river or canal, these effects are further
exaggerated. The sinkage and trim in very shallow water may set an upper limit to the speed at
which the ships can operate without grounding.

Wave Pattern in Shallow Water

Categorization of water as deep, intermediate, or shallow is based on the ratio of water depth to
wavelength h/λ, where h is the water depth and λ is the wavelength. In shallow water as h/λ is
small, the velocity of propagation of the wave, commonly called celerity, is given by:

𝐶 = √𝑔ℎ
Where: g is the acceleration due to gravity, h is the water depth.
The velocity now depends only on the water depth and waves of different wavelength propagate
at the same speed called Critical Speed. Since the waves travel in the same velocity C as the ship
speed V, speed ranges can be taken in terms of depth Froude number. The depth Froude number
is defined as:
𝑉
𝐹𝑟ℎ =
√𝑔ℎ
The speeds below and above C are known as the sub critical speeds and super critical speeds,
respectively.

Mohamed A. Abdelrazek,
Resistance & Powering 15
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

At speeds well below Frh =1.0 i.e., critical speed, the wave system is as shown in figure below.
with a transverse wave system and a divergent wave system propagating away from the ship at
an angle of about 35°.
As the ship speed approaches the critical speed, Frh=1.0, the wave angle approaches 0 °, or
perpendicular to the track of ship as shown in below and at speeds greater than the critical speed,
the diverging wave system returns to a wave propagation angle of about cos-1 (1/Frh) and
transverse waves do not exist.

Figure 10 Critical Froude Number Effect on Generated Wave Pattern.

Mohamed A. Abdelrazek,
Resistance & Powering 16
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Schlichting Correction Method

Schlichting analyzed, theoretically and experimentally, the increased ship resistance due to the
effect of shallow water in the subcritical speed range 𝑭𝒏𝒉 < 𝟏, not the decreased resistance in the
supercritical speed range 𝑭𝒏𝒉 > 𝟏, in water of an unlimited lateral extents.
At any particular deep-water speed 𝑉∞, the resistance components are 𝑅𝑉∞, 𝑅𝑊∞ and 𝑅 𝑇∞:
𝑅𝑇∞ = 𝑅𝑉∞ + 𝑅𝑊∞
At this speed 𝑉∞, the wave pattern generated by the ship has a wavelength: 𝐿𝑤𝑣 = 2𝜋 ∙ 𝑉∞2 /𝑔.

Correction for Wave Retardation Effect

In water of depth ℎ the same wavelength 𝐿𝑤𝑣 would be generated at some lower intermediate
speed 𝑉i:
𝑉𝑖 = 𝑉∞ − 𝛿𝑉𝑐
∴ 𝛿𝑉𝐶 = 𝑉∞ (1 − 𝑉𝑖 ∕ 𝑉∞ )

𝑉𝑖 ∕ 𝑉∞ = √𝑡𝑎𝑛ℎ(2𝜋ℎ ∕ 𝐿𝑤𝑣 )
𝐿𝑤𝑣 = 2𝜋 𝑉∞2 ∕ 𝑔

∴ 𝑉𝑖 ∕ 𝑉∞ = √𝑡𝑎𝑛ℎ(𝑔ℎ ∕ 𝑉∞2 )

2 )
∴ 𝑉𝑖 ∕ 𝑉∞ = √𝑡𝑎𝑛ℎ(1 ∕ 𝐹𝑟ℎ

2 ))
∴ 𝛿𝑉𝑐 = 𝑉∞ (1 − √𝑡𝑎𝑛ℎ(1 ∕ 𝐹𝑟ℎ

Schlichting assumed that the wave making resistance in shallow water at speed 𝑉𝑖 would be the
same as that at the speed 𝑉∞ in deep water.
The total resistance at the intermediate speed 𝑉𝑖 in water of depth ℎ would then be found by:
𝑅𝑇𝑖 = 𝑅𝑉𝑖 + 𝑅𝑊∞
𝑅𝑉𝑖 = (1 + 𝑘)𝑖 𝑅𝑓𝑖
(1 + 𝑘) 𝑠ℎ𝑎𝑙𝑙o𝑤 = (1 + 𝑘) 𝑑𝑒𝑒𝑝 + 0.644 ∙ (Τ/ℎ)1.72

Mohamed A. Abdelrazek,
Resistance & Powering 17
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Correction for Back Flow Effect

There is further loss in speed 𝛿𝑉𝑝 because of the increase in potential flow around the hull due to
the restriction of area by the proximity of the bottom, giving as the final speed:
𝑉ℎ = 𝑉𝑖 − 𝛿𝑉𝑝
𝛿𝑉𝑝 = 𝑉𝑖 (1 − 𝑉ℎ ∕ 𝑉𝑖 )
1
𝑉ℎ ∕ 𝑉𝑖 =
0.9956 + 0.1358 𝐴𝑥 ∕ ℎ2

Schlichting’s work is not theoretically rigorous but can be looked upon as satisfactory solution to
the complicated problem. The method is effective in obtaining the shallow water resistance at
subcritical speeds.

Howe’s Shallow Water Formula

This method is used directly to predict the resistance of a barge shaped vessels operating in
restricted water ways.
The formula given by Howe (1976) is obtained based on the test data of five configurations, the
Anglo-American system is used in the original form of Howe's formula, but the formula
rewritten in the international system is shown as follows:
15.24 0.445 0.38 1.19 2
0.6+
𝑅𝑇 = 1.0684 (3.2808 𝑇) 𝐵𝑐 −𝐵 𝑒𝑥𝑝 ( )𝐿 𝐵 𝑉
ℎ−𝑇

Where:
RT: Total resistance in kilograms.
T: Vessel draft in meter.
BC: Breadth of the restricted water way, for River Nile ~ 1000 m.
B: Vessel breadth in meter.
h: Depth of water in meter.
L: Vessel length in meter.
V: The ship speed in meters per second.

Mohamed A. Abdelrazek,
Resistance & Powering 18
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Model 1 Shallow Water Correction Results

An Excel spreadsheet is also used to tabulate the equations and results are shown below:

Total Resistance VS Speed

200

180

160

140

120
Rt (KN)

100

80

60

40

20

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Speed (Knot)

Deep Water Shallow Water Howe Shallow Water

Figure 11 Total Resistance Speed Curve for Deep and Shallow Waters.

Mohamed A. Abdelrazek,
Resistance & Powering 19
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Power VS Speed
800

700

600

500
Power (KW)

400

300

200

100

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Speed (Knot)

Deep Water Shallow Water Howe Shallow Water

Figure 12 Total Power Speed Curve for Deep and Shallow Waters.
Mohamed A. Abdelrazek,
Resistance & Powering 20
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Model 1 Resistance Prediction Using CFD Analysis

The CFD Simulation will be conducted using Siemens Simcenter STAR-CCM+ software with
Parasolid model derived previously from SOLIDWORKS in lines derivation section.

Prerequisites
A geosim will be used instead of the full scale with down scale factor λ = 10 at full scale
velocity of 6 knots (3.084 m/s).
Due to hull symmetry, only half of the geosim will be simulated.
To satisfy all the similarities conditions:

𝐿𝑆 𝑆 3 𝛻
= √𝑆 𝑆 = √𝛻 𝑆 = 𝜆 (Geometric Similarity)
𝐿𝑀 𝑀 𝑀

𝑉𝑆
= √𝜆 (Kinematic Similarity)
𝑉𝑀

𝐹𝑆 𝑅𝑇𝑆
=𝑅 = 𝜆3 (Dynamic Similarity)
𝐹𝑀 𝑇𝑀

Table 2 Half Geosim Characteristics.

Full Scale Geosim Half Geosim

L (m) 50 5 5
S (m2) 520 5.2 2.6 λ
Vol (m )
3
418 0.418 0.209 10
V (m/s) 3.084 0.97524643 0.97524643
ρ (m3/s) 1000 1000 1000
υ (m2/s) 0.00001 0.00001 0.00001
Re 154200000 4876232.152 4876232.152
Fn 0.139249891 0.139249891 0.139249891

Computational Domain Size & Boundary Conditions

Table 3 Domain Limits.

Aft End Fore End Top Bottom Side

Length (m) 15 10 5 0.23 7
From From
Reference From Free surface From Free surface From Symmetry Plane
Stern Bow

Mohamed A. Abdelrazek,
Resistance & Powering 21
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 13 Model 1 Computational Domain (Virtual Towing Tank).

The following boundary conditions are applied:

• Symmetry plane at the hull midplane.
• Symmetry plane at the side of the tank.
• Hydrostatic pressure corresponding to the undisturbed water surface at the downstream
boundary of the tank.
• Inlets with prescribed velocity and volume fraction at the upstream and top boundaries.
• Wall for hull and bottom boundary to simulate the shallow water effect.

Wave Damping Functions

Unless appropriate measures are applied, wave reflections can occur during a flow resistance
simulation. These wave reflections must be avoided, because they interact with the true wave
field and can lead to invalid results:
𝑉2
𝐿𝑤𝑣 = 2𝜋 𝑐𝑜𝑠(𝜃)
𝑔
𝐿𝑤𝑣
∴ = 2𝜋 𝐹𝑛2 𝑐𝑜𝑠 2 (𝜃)
𝐿
∴ 𝐿𝑤𝑣𝑀𝑎𝑥 = 0.6157 𝑚
Which is the damping length.

Mohamed A. Abdelrazek,
Resistance & Powering 22
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

2𝜋
𝜔 = √𝑔 = 10.005 𝑟𝑎𝑑/𝑠
𝐿𝑤𝑣

For damping functions:

𝑛=2
𝑓1 = 𝜋 𝜔 = 31.433
𝑓2 = 0

Prism Layers Calculations

According to White (2006):
𝑋 (1 − 𝑟 𝑚 )
𝛿=
1−𝑟
𝛿
𝑙𝑛 (1 − (1 − 𝑟) 𝑋)
𝑚=
𝑙𝑛(𝑟)
𝛿 0.16
=
𝐿 𝑅1∕7
𝑒

𝑌+ 0.487
= 𝑉
𝑋 𝜈 𝑙𝑛(0.06 𝑅𝑒)
Where:
δ: Prism layers thickness (m).
X: First layer thickness (m).
r: Stretching factor.
m: Number of layers.
By setting Y+ =100, r =1.2 and substituting values from above:

Table 4 Prism Layers Characteristics.

δHull δBottom XHull mHull mBottom

0.0886 0.0686 2.65E-03 12 10

Mohamed A. Abdelrazek,
Resistance & Powering 23
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Mesh Generation
In order to achieve accurate resistance predictions, it is recommended to use trimmed
hexahedral meshes with local refinements and prism layers along the walls of the wetted surface.
In order to refine the computational mesh only where it is necessary, volumetric controls with
particular shapes are used and a particular cell size is then assigned to these volumetric controls.
Using trimmed cells means that the mesh is aligned with the undisturbed free surface.
A medium to fine global mesh is used with the following parameters:

Table 5 Global Mesh Settings.

Enhanced Quality Triangle

Surface Remesher
Create Aligned Mesh
Base Size 1 m
5%
Target Surface Size
0.05 m
1%
Minimum Surface Size
0.01 m
Number of Prism Layers 5
Prism Layer Stretching 1.2
Prism Layer Total Thickness 0.0886 m
Volume Growth Rate Very Slow
100%
Maximum Cell Size
1 m
(0,0,0)
Mesh Alignment Location
Free Surface

In addition to the globally applied mesh settings, a local mesh setting for the boundary surfaces
is applied:

Table 6 Local Hull Boundary Mesh Settings.

Hull Boundary
Base Size 1 m
2%
Target Surface Size
0.02 m
1%
Minimum Surface Size
0.01 m
Rest values are parent values

Mohamed A. Abdelrazek,
Resistance & Powering 24
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Table 7 Local Bottom Boundary Mesh Settings.

Bottom Boundary
Base Size 1 m
2%
Target Surface Size
0.02 m
1%
Minimum Surface Size
0.01 m
Number of Prism Layers 5
Prism Layer Stretching 1.2
Prism Layer Total Thickness 0.0686 m
Rest values are parent values

Table 8 Local Rest of the Boundaries Mesh Settings.

Rest of the Tank Boundaries

Base Size 1 m
200%
Target Surface Size
2 m
180%
Minimum Surface Size
1.8 m
Prism Layers Disabled
Rest values are parent values

It must be noted that the number of prism layers is reduced from the calculated values to satisfy
the available computational resources, this will affect the Y+ value as discussed in the results
section.
To reduce wave reflections due to abrupt mesh transitions, the volume growth rate is modified to
be very slow.
In order to refine sharp corners or edges of the geometry, for example at the bow, or to improve
the resolution of flow features, for example flow separation or the Kelvin wake, a volumetric
refinements to volume mesh is applied, it is particularly important to refine the free water surface
area, where there is an interface area between water and air.
The volumetric refinement is conducted using the Macro file partShapes.java from the software
library, the macro executes and creates the required part shapes and volumetric controls as shown
in the next page.
The generated mesh is shown in the next page, with the following characteristics:
6206969 cells, 18572800 faces, 6610455 verts.

Mohamed A. Abdelrazek,
Resistance & Powering 25
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 14 Volumetric Refinements.

Figure 15 Generated Mesh.

Mohamed A. Abdelrazek,
Resistance & Powering 26
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 16 Near Hull Mesh.

To check the skewness of the mesh a histogram is shown below:

Figure 17 Cell Skewness.

As shown above most of the cells have a skewness about zero which is an indication of a good
mesh.
A cell quality histogram is also shown below, most cells have a quality very close to one which
indicates a high mesh quality:

Mohamed A. Abdelrazek,
Resistance & Powering 27
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 18 Cell Quality Histogram.

Simulation Physical Setup

Flow resistance simulation is a complex transient multi-physics problem.
The physics models used are shown below:

Figure 19 Simulation Physical Models.

Mohamed A. Abdelrazek,
Resistance & Powering 28
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

The selected turbulence model is the K-Omega SST model by Menter, which is the standard
turbulence model for ship resistance simulations.
Eulerian continuous-continuous VOF multiphase model is used with flat wave to simulate both
water and air and their interaction:

Figure 20 Wave Parameters.

Damping Parameters are added:

Figure 21 Wave Damping Parameters.

For simplicity, the simulation will not solve for ship motions; sinkage and trim so the DFBI
module is disabled.

Mohamed A. Abdelrazek,
Resistance & Powering 29
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

To avoid the back flow problem, an adaptive time step with explicit multistep is used to ensure
that CFL number is within range for this type of simulations; CFL is kept between 5 and 10,
minimum time step is setted to be 0.001 s, and initial time step is 0.01 s.

Figure 22 Multistep Configuration.

Under relaxation is used to increase the stability of the simulation, but a stricter convergence
criterion is applied to avoid pseudo or false convergence:

Table 9 Under Relaxation Factors.

Under Relaxation Factor

Velocity 0.7
Pressure 0.4
Explicit Multistep 0.95
K-Omega Turbulence 0.8
Turbulent Viscosity 1

For convergence criteria, criteria are applied on mean total drag with asymptotic option, a strict
limit of 0.02 is applied:

Figure 23 Convergence Setup.

Mohamed A. Abdelrazek,
Resistance & Powering 30
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

High Resolution Interference Capturing HRIC Scheme is used to model the free surface ship
generated waves with upper and lower limit of Courant number as follow:

Figure 24 HRIC Settings.

Initial conditions are defined according to the VOF wave field functions parameters:

Table 10 Initial Conditions Settings.

Initial Conditions Field Functions

Pressure Hydrostatic Pressure of Heavy Fluid of Flat Vof Wave 1
Velocity Velocity of Flat Vof Wave 1
Volume Fraction of Water Volume Fraction of Heavy Fluid of Flat Vof Wave 1
Volume Fraction of Air Volume Fraction of Light Fluid of Flat Vof Wave 1

For stopping criteria, the following settings are applied:

Table 11 Stopping Criteria.

Mohamed A. Abdelrazek,
Resistance & Powering 31
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Results & Discussions

The graphs listed below shows the parameters derived from the simulation.
The graph below shows the variation of Courant number across the domain, the abrupt change is
due to the change of time step during the solution, the value decreases and comes below 15 by
the end of the simulation.

Figure 25 CFL Max Across the Domain.

For the hull itself, CFL varies as shown below but it does not exceed 10, which is a good
indication:

Figure 26 Hull Courant Number.

Mohamed A. Abdelrazek,
Resistance & Powering 32
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

For Y+, the following figure shows the variation of the value across both the hull and the bottom
with maximum value of 248:

Figure 27 Y+ Value Across the Hull & the Bottom.

The reason of the variation between the value from solution 248 and the assumed value 100 is
due to the reduction of prism layers, the value of 248 can still be good as it lies between 30 and
300 (Log Law Region).
The below plot shows the residuals of the simulation:

Figure 28 Simulation Residuals.

Mohamed A. Abdelrazek,
Resistance & Powering 33
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

The following figures show the free surface without any back flow, and the ship generated wave
pattern:

Figure 29 Free Surface Representation.

Figure 30 Ship Generated Waves.

The wave pattern complies with the fact that for shallow water, the angle of waves becomes
larger when vessel approaches critical speed.

Mohamed A. Abdelrazek,
Resistance & Powering 34
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Resistance data are plotted below:

Figure 31 Log Scaled Values of Resistance Components.

Figure 32 Log Scaled Values of Total Resistance.

Mohamed A. Abdelrazek,
Resistance & Powering 35
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Figure 33 Log Scaled Values of Mean Total Resistance.

Total Resistance Result

By running the mean total drag report, the following results are obtained:

Results are for half geosim, for full geosim:

𝑅𝑇 = 25.94 𝑁
To get the full-scale hull result:
𝐹𝑆 𝑅𝑇
= 𝑆 = 𝜆3
𝐹𝑀 𝑅𝑇𝑀

∴ 𝑅𝑇𝑆 = 25.94 𝐾𝑁

Mohamed A. Abdelrazek,
Resistance & Powering 36
Alexandria University Naval Architecture &
Faculty of Engineering Marine Engineering Department

Model 1 Resistance & Powering Summary

To sum up all the methods used, the following table collects all the results needed at 6 knots:

Table 12 Collected Resistance & Power Data.

With 5%
Rt (KN) P (KW)
Margin P (KW)
Schlichting 22 67.848 71.2404
Howe 20 61.68 64.764
Star CCM 25.94 79.99896 83.998908

A margin of 5% is added to compensate for superstructure air resistance and appendages attached
to the hull.
Simulation results will be considered as it is the maximum one.
Note that the power given from the table above is the thrust power not the delivered power.
Delivered power can be estimated by assuming a propeller behind hull efficiency of 30%,
electric and mechanical efficiency of 95% as follows:
𝑃𝑇
𝑃𝐷 =
𝜂𝐵𝐻 𝜂𝑀𝐸

𝑃𝐷 ≈ 295 𝐾𝑊

The delivered power does not count for the fuel cell efficiency.

Mohamed A. Abdelrazek,
Resistance & Powering 37