See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315507048

A simplified method to calculate trim and resistance of a two-stepped planing

hull

Article in Ships and Offshore Structures · March 2017

DOI: 10.1080/17445302.2016.1262809

CITATIONS READS

33 7,241

3 authors:

Abbas Dashtimanesh Sasan Tavakoli

KTH Royal Institute of Technology University of Melbourne
95 PUBLICATIONS 1,363 CITATIONS 94 PUBLICATIONS 1,197 CITATIONS

SEE PROFILE SEE PROFILE

Prasanta Sahoo
UNSW Canberra
81 PUBLICATIONS 660 CITATIONS

SEE PROFILE

All content following this page was uploaded by Prasanta Sahoo on 13 October 2017.

The user has requested enhancement of the downloaded file.

 Ships and Offshore Structures

ISSN: 1744-5302 (Print) 1754-212X (Online) Journal homepage: http://www.tandfonline.com/loi/tsos20

A simplified method to calculate trim and

resistance of a two-stepped planing hull

Abbas Dashtimanesh, Sasan Tavakoli & Prasanta Sahoo

To cite this article: Abbas Dashtimanesh, Sasan Tavakoli & Prasanta Sahoo (2017) A simplified
method to calculate trim and resistance of a two-stepped planing hull, Ships and Offshore
Structures, 12:sup1, S317-S329

To link to this article: http://dx.doi.org/10.1080/17445302.2016.1262809

Published online: 21 Mar 2017.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=tsos20

Download by: [Florida Institute of Technology] Date: 21 March 2017, At: 10:30
SHIPS AND OFFSHORE STRUCTURES, 
VOL. , NO. S, S–S
http://dx.doi.org/./..

A simplified method to calculate trim and resistance of a two-stepped planing hull

Abbas Dashtimanesha , Sasan Tavakolib and Prasanta Sahooc
a
Engineering Department, Persian Gulf University, Bushehr, Iran; b Department of Maritime Engineering, Amirkabir University of Technology, Tehran,
Iran; c Department of Ocean Engineering and Sciences, Florida Institute of Technology, Melbourne, FL, USA

ABSTRACT ARTICLE HISTORY

In a creative design process of planing craft, stepped hulls could fill a gap in the planing craft industries Received  September 
in response to low drag and high speed demands. However, there exists a need for new computational Accepted  November 
tools for performance prediction of such hulls. Therefore, in the current work, an attempt has been made to
KEYWORDS
develop a mathematical model for performance prediction of two-stepped hulls. Savitsky’s mathematical Stepped hulls; planing craft;
model has been modified, and in conjunction with linear wake theory, a new mathematical model has been mathematical model;
proposed, which would enable prediction of trim, resistance and other parameters related to planing hulls resistance; trim
with transverse steps. To validate the proposed model, existing experimental data have been used. The
obtained results are in good agreement with experimental data. As such, developed mathematical model
can be used in conceptual design phase of stepped hulls with transverse steps.

Nomenclature 1. Introduction
bi Wetted beam of the ith body Stepped hulls were first introduced by Rev Ramus of Sussex
Bchine Chine beam England in 1872 (Froude 1872). Subsequently, powerboat pio-
Cl0i Flat planing plate lift coefficient for the ith body neers used stepped hulls on hydroplanes between the two world
Clβ i Lift coefficient of the ith body wars (Garland 2011). Implementing steps in the bottom of
Cfi Frictional coefficient of the ith body hydroplanes and planing hulls led to a considerable reduction
Cvi Speed coefficient of the ith body in planing wetted area. One could presume that stepped hulls
di Lift arm moment of the ith body were a good choice to achieve higher speed. Moreover, it was
dcf Roughness factor observed that by utilising stepped hulls, hydrodynamic lift dis-
DFi Frictional drag acting on ith body tributes over two or three small wetted areas of the hull bottom
Ti Draft of ith body which could lead to more stable craft in rough weather condi-
g Gravitational constant tions. However, pioneers observed that fast stepped boats were
Hi Height of ith step unpredictable and hard to handle at very high speeds. Therefore,
lcg Longitudinal centre of gravity with respect to transom stepped hulls did not see any further development.
Lifti Lift force acting on ith body From 1960 to 1990, few authors such as Clement and Pope
Li Difference between chine wetted length and keel wetted (1961), Moore (1967) and Clement and Desty (1980) contin-
length in ith body ued to contribute towards performance prediction of stepped
Ldryi Dry length of ith step from the transom hulls. They believed that with some refinements, stepped bot-
Lstepi Distance of ith step from the transom toms can be used on planing hulls. Clement and Koelbel
Lweti Wetted length of ith body (1991) studied the effects of step design on performance of
m Mass of hull planing powerboats. They also suggested optimised designs
Rni Reynolds number of the ith body for stepped planing monohulls and catamarans (Clement and
Rpi Pressure resistance of the ith body Koelbel 1991). Clement and Koelbel (1992b) in an article dis-
Rti Total resistance of the vessel cussed about development of stepped hulls during the past
β Deadrise angle of the boat century. Subsequently, various researchers have tried to under-
β i Local deadrise angle of the boat stand physical behaviour of stepped hulls using experimen-
 Displacement tal and numerical studies. In this regard, Clement (2003) and
γ i Spray angle in the ith body Clement and Koelbel (1992a) published two reports in which
λi Mean wetted length of the ith body efficient configuration of one stepped hulls has been suggested
τ i Local trim angle of the ith body and some shortcomings such as porpoising, bow steering, etc.
τ o Overall trim angle of the hull have been described. The reattachment of the spray root to
the after-body is an important issue in design of stepped
hulls which was also investigated by Clement (2003, 2006).

CONTACT Abbas Dashtimanesh [email protected]

©  Informa UK Limited, trading as Taylor & Francis Group
S318 A. DASHTIMANESH ET AL.

In this regard, Savitsky and Morabito (2010) performed an there was a lack of complete and accurate validation for their
extensive set of experiments to derive mathematical formula- model.
tions for flow separation from steps. It may however be noted In the current article, an attempt has been made to develop a
that measurements of transom waves behind several un-stepped novel mathematical model based on Danielsson and Stromquist
prismatic hulls were carried out and no direct measurements (2012) studies. As mentioned in literature reviews, there has
of step flows were carried out. In an article, Danielsson and been no direct measurement or empirical formula for wake
Stromquist (2012) questioned the applicability of derived for- profile beneath the two-stepped hulls. Therefore, researchers,
mulas by Savitsky and Morabito (2010) for two-stepped hulls Danielsson and Stromquist (2012), observed that linear wake
and they were unsuccessful in applying these relations for two- profile (LWP) may be a good assumption for flow separation
stepped hulls. Furthermore, Taunton et al. (2010, 2011) were one from steps. Therefore, the present study attempts to take into
of the pioneers who have carried out experimental work on two- account this suggestion. Simple relations from existing litera-
stepped hulls. They conducted a series of experiments on models ture study have also been implemented in a sequential man-
of 2 m in length of two-stepped hulls in calm water and waves. ner to provide a robust computational algorithm for perfor-
They provided a set of towing tank data which were suitable for mance prediction of two-stepped hulls with transverse steps.
validation of numerical and mathematical models. Vitiello et al. Moreover, an extensive set of validation have been performed
(2012) also performed model experiments and sea trial tests on by using various towing tank test data acquired in previous
a two-stepped hull in a towing tank of University of Federico II studies and a detailed discussion on the results have been
in Napoli. However, their dilemma was on developments of an presented.
accurate experimental set-up rather than presentation of physi-
cal details of two-stepped hulls. White et al. (2012) have also per-
formed some experiments on two-stepped hulls and concluded 2. Problem definition
that stepped hull may improve the powering performance of General method to achieve a very high speed at sea is imple-
planing boat only under certain conditions. Recently, Lee et al. mentation of various tools for total drag reduction. Removal of
(2014) have studied two-stepped hulls by systematic variation of some part of hull’s bottom leads to flow separation, and conse-
step configuration and displacement. They have considered 84 quently, reduction of wetted area. In this regard, naval architects
test cases and observed that in all cases, two-stepped hulls lead strive to design planing hulls using various transverse disconti-
to resistant reduction in comparison with equivalent un-stepped nuities in the bottom, better known as steps. Generally, three dif-
hulls. ferent forms of stepped hulls may be considered in a planing hull
Although there exists an urgent need for conducting more identified as (1) transverse step, (2) step pointed aft and (3) re-
experimental tests on stepped hulls, some researchers have tried entrant V step as shown in Figure 1. In the present paper, trans-
to modify the well-known Savitsky (1964) method for stepped verse step has been incorporated. In addition to step geometry,
hulls. Svahn (2009) used presented formulas by Savitsky and step height and longitudinal position of step are also important
Morabito (2010) and extended Savitsky’s (1964) method using factors which influence the wetted area.
wake theory for performance prediction of one stepped hull. Wetted area of each portion of two-stepped hull may be dif-
Kaidy (2013) also developed a computer program to estimate ferent. Generally, a two-stepped hull can be considered as three
trim and resistance of one stepped planning hulls. Danielsson tandem planing surfaces (TPS) which have their own planing
and Stromquist (2012) tried to develop Svahn’s (2009) model characteristics such as wetted length, wetted surface, local trim
for two-stepped hulls. However, they failed in implementa- angle, local deadrise angle, Reynolds number, Froude number,
tion of wake theory of Savitsky and Morabito (2010) for two- etc. Therefore, it is vital to consider each planing surface, sepa-
stepped hull because of range of applicability and other issues. rately.
Therefore, they assumed that flow separates from step and reat- However, to compute planing characteristics of each indi-
taches to after-body portion of hull by a linear trend. This vidual planing surface, development of an appropriate model
assumption was reasonable in very high-speed situation. How- for prediction of flow separation from each step is essential.
ever, Danielsson and Stromquist (2012) had some difficulties in Unfortunately, there is no appropriate wake model as yet which
definition of wetted length and Reynolds numbers. Moreover, can be used for two-stepped hulls. In the present study, it is

Figure . Different forms of step. (This figure is available in colour online.)

 SHIPS AND OFFSHORE STRUCTURES S319

Figure . Various possible wetting conditions for planing surfaces. (This figure is available in colour online.)

assumed that flow separates from steps and reattaches to after- A stepped hull with n steps can be considered as n + 1 plan-
body portion by a linear trend. Consequently, various condi- ing surfaces with individual planing characteristics (Figure 3).
tions can come into play on planing surfaces which are shown in Equations of motion for equilibrium condition have been writ-
Figure 2. Various wetted surfaces have been depicted in Figure 2 ten as follows:
and described as follows:

(1) Chine wetting condition for fore-body portion of hull. Heave direction:
+1
n
(2) Chine drying situation of fore-body portion of hull.  
− mg + Lifti cos(τo + τi ) − DFi sin(τo + τi ) = 0
(3) Chine drying situation of middle-body portion of hull.
i=1
(4) Chine drying situation of aft-body portion of hull.
Pitch direction:
(5) Chine wetting condition for middle-body portion of +1
n
hull. − mg.lcg cos(τo ) + [Lifti di − DFi Ti /2] = 0 (1)
(6) Chine wetting condition for aft-body portion of hull. i=1

Various combinations of depicted situations (a–f) in Figure 2

where m is hull mass (kg), g is gravitational constant (m/s2 ), lcg is
can occur at different Froude numbers. Due to division of two-
longitudinal position of centre of gravity and τo is overall trim
stepped hull as three TPS, hydrodynamic lift also distributes in
angle of planing hull. Moreover, Lifti , DFi , τi , di and Ti are
three different parts. To recognise how lift distributes over hull,
hydrodynamic lift, frictional resistance, local trim angle, lift arm
equilibrium equations which are presented in the next section
moment and draft related to ith planing surface, respectively. It
need to be solved.
should be noted that origin of coordinate system is located on
lower edge of the transom and pitch moments are calculated rel-
3. Mathematical model
ative to this point. In the next part, the computational details of
In this section, equations of motions related to two-stepped
planing characteristics have been presented.
planing hulls have been derived. For this purpose, two-stepped
hull is divided into fore-body, middle-body and after-body parts
and various planing characteristics of each individual planing 3.2. Planing equations and computational procedure
surface are calculated. Moreover, it is assumed that each planing To obtain various forces and planing parameters in Equation (1),
body is related to next planing surface by a LWP. Complete fea- two assumptions have been made which are as follows:
tures of mathematical model including motion equations, char-
acteristics of each planing surface and computational procedure (1) LWP: As mentioned in Danielsson and Stromquist
have been presented in the following subsections. (2012), there is no reliable formula for fluid flow sepa-
ration from steps, especially in the case of two-stepped
3.1. Motion equations in equilibrium condition hulls. Therefore, LWP may be considered as a reasonable
To derive equilibrium equations of a stepped planing hull, it is assumption of two-stepped hulls operating in the high-
convenient to divide stepped hulls to several planing surfaces. speed regime.

Figure . Forces acting on each planing surface in equilibrium condition. (This figure is available in colour online.)
S320 A. DASHTIMANESH ET AL.

(2) TPS: A planing hull with n steps can be considered as n To calculate various planing characteristics of TPS and solve
+ 1 planing surfaces with individual planing character- equilibrium equations, well-known formulas and method of
istics. These TPS will be associated with each other using Savitsky (1964) will be extended to two-stepped hulls. However,
LWP assumption. it must be noted that these relations should be utilised in proper
order which have been discussed in the next sub-section.

Lwo
... Planing characteristics of fore-body
At first, overall wetted length, Lweto, and general trim angle, τo,
Γ3 Γ2 Γ1 are guessed (Figure 4). It is then possible to calculate the wetted
length of various body parts as follows:

⎧
Lw3 Lw2 Lw1 ⎨Lwet1 = Lweto − Lstep1 Fore − body
Lwet2 = Lstep1 − Lstep2 − Ldry2 Middle − body . (2)
Figure . Wetted surface of stepped planing hulls. (This figure is available in colour ⎩
Lwet3 = Lstep2 − Ldry3 Aft − body
online.)

where Lstep j is the distance of jth step from transom. To obtain

Β
wetted length of middle-body and aft-bodies, one should recog-
nise how the fluid flow separates from steps. Based on LWP
assumption, dry length of middle- and aft-bodies can be esti-
mated as follows:
Β ΒΙ
Wake profile ⎧
Horizontal line ⎪
⎪
H2
⎨ Ldry 2 = Middle − body
Figure . Deadrise angle of stepped hull. (This figure is available in colour online.) tan(τo + τ1 )
(3)
⎪
⎪ H3
⎩ Ldry 3 = Aft − body
tan(τo + τ2 )

Input: displacement, LCG,

Find the local deadrise angle
body characteriscs
of each surface

Guess a trim angle Compute li force of each

surface

Guess an overall weed length Compute resistance force of

each hull

No
Determine local trim angle of
each planing surface Does Heave and pitch
equaons sasfy?

Yes
Determine weed surface of
each planing surface

Output: trim angle, weed

surface, resistance

Figure . Computational procedure algorithm. (This figure is available in colour online.)

 SHIPS AND OFFSHORE STRUCTURES S321

Figure . Body plan and profile of model C (Taunton et al. ).

Experimental Results of Taunton et al (2010) Proposed Method Experimental Results of Taunton et al (2010) Proposed Method

5 100

80
4
60
3 R (N)
Τ (deg)

40
2
20
1
0
3 5 7 9 11 13
0
3 5 7 9 11 13
V (m/s)
V (m/s)
Figure . Comparison between predicted and measured total resistance. (This fig-
Figure . Comparison of obtained trim with experimental data of Taunton et al. ure is available in colour online.)
(). (This figure is available in colour online.)

where βi is the deadrise angle of the ith surface. Using

Equation (4), wetted beam of each body parts obtained as fol-
Experimental Results of Taunton et al (2010) Proposed Method
lows:
0.6

0.5 bi = 2Lweti tan(γi ) for i = 1, 2, 3. (5)

0.4
S (m2)

0.3 However, it may be noted that calculated wetted beam may be

0.2 larger than chine beam ( bi > Bchine ). In such a situation, wetted
0.1 chine length is equal to
0
3 5 7 9 11 13 Lwetchine_i = Lweti − Li for i = 1, 2, 3 (6)
V (m/s)
where Li is Bchine /(2 tan(γi )). Finally, speed coefficient can be
Figure . Wetted surface comparison from mathematical model and experiments. determined by
(This figure is available in colour online.)

V
Cvi = . (7)
gbi
where H j is the height of jth step. It must be noted that the com-
putational procedure is limited to situations where the separa- In the computational procedure, it is possible to acquire
tion continues across the step. dynamic draft and mean wetted length of each planing area as
Consequently, spray angles, γi , can be evaluated as follows: follows:

bi
ti =tan(βi )
2
π tan(τo + τi ) Lweti + Lwetchine_i
γi = tan−1 fori = 1, 2, 3 (4) λi = . (8)
2 tan(βi ) 2bi
S322 A. DASHTIMANESH ET AL.

where Cfi in Equation (9) is well-known ship-model correlation

line of ITTC-1957. Consequently, it is possible to calculate fric-
tional drag as

DF i = 0.5ρSiV 2 (Cfi + Cf ) (10)

where Cf is the roughness factor given as 0.0004. In the next

step, lift coefficient and lift force of each planing surfaces have
been computed:
Figure . Body plan of Lee’s model (Lee et al. ).
2
λi
Cl0i = ((τo + τi )1.1 λi 0.5 ) + 0.012 + 0.0055
Cvi
By using obtained mean wetted length, Reynolds number and
frictional drag coefficient for each individual planing surface can Clβi = Cl0i − 0.0065βi Cl0.6
0i
be written as follows: Ni = 0.5ρV 2 b2i Clβi cos(βi − βlocal_i ). (11)
V λ i bi
Rni = and accordingly, pressure drag is calculated as follows:
ν
0.075
Cfi = . (9) 1
log10 (Rni − 2)2 Rpi = ρb2i Clβi tan (τ 0 + τ i ) (12)
2
Experimental Results of Lee et al. (2014) Proposed Method

0.30 3.50 0.54

Τ (deg)

S (m2)
R/Δ

0.25 3.00
0.52
2.50
0.20
2.00 0.50
0.15
1.50 0.48
0.10
1.00
0.05 0.46
0.50
0.00 0.00 0.44
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(a)
0.25 4.00 0.52
Τ (deg)

S (m2)
R/Δ

3.50
0.20 0.51
3.00
0.50
0.15 2.50
2.00 0.49
0.10 1.50
0.48
1.00
0.05 0.47
0.50
0.00 0.00 0.46
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(b)
0.25 4.50 0.53
Τ (deg)

S (m2)
R/Δ

4.00
0.20 0.52
3.50
3.00 0.51
0.15
2.50
0.50
2.00
0.10
1.50 0.49
0.05 1.00
0.48
0.50
0.00 0.00 0.47
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(c)

Figure . Comparison between predicted results and experimental data, Case#: (a)  = . N; (b)  = . N; (c)  = . N. (This figure is available in colour online.)
 SHIPS AND OFFSHORE STRUCTURES S323

Experimental Results of Lee et al. (2014) Proposed Method

0.25 4.00 0.50

Τ (deg)

S (m2)
R/Δ
3.50
0.20 0.48
3.00
0.15 2.50 0.46
2.00
0.10 1.50 0.44
1.00
0.05 0.42
0.50
0.00 0.00 0.40
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(a)
0.25 4.00 0.52

Τ (deg)

S (m2)
R/Δ

3.50
0.20 0.50
3.00
0.48
0.15 2.50
2.00 0.46
0.10 1.50
0.44
1.00
0.05 0.42
0.50
0.00 0.00 0.40
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(b)
0.25 4.50 0.52
Τ (deg)

S (m2)
R/Δ

4.00
0.20 0.50
3.50
3.00 0.48
0.15
2.50
0.46
2.00
0.10
1.50 0.44
0.05 1.00
0.42
0.50
0.00 0.00 0.40
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(c)

Figure . Comparison between predicted results and experimental data, Case#: (a)  = . N; (b)  = . N; (c)  = . N. (This figure is available in colour online.)

It is worthwhile to note that the effect of local variation of minimising the constrained nonlinear multivariable Equation
transverse flow has been included in Equation (11) (Figure 5). (1), trim and wetted length of two-stepped hull can be obtained.
Moreover, although some authors have presented the wave resis- By using fmincon command, both heave and pitch equations
tance in their computations (Sahoo et al. 2008), it is neglected in would be solved, simultaneously.
the current calculation. At this step, total drag of stepped planing Moreover, optimisation algorithm behind the fmincon
hull can be easily computed as is a subspace trust region method and is based on the
interior-reflective Newton method described in Coleman

3 and Li (1996). Each iteration involves the approximate solution
Rt = (DF i + Rpi ). (13) of a large linear system using the method of preconditioned
i=1 conjugate gradients (Coleman and Li 1994).

At this point, both expressions of Equation (1) are iteratively

solved until equilibrium condition has been achieved. To solve 4. Results and validation
the equilibrium equations, a nonlinear optimisation algorithm In this section, an attempt has been made to validate the
has been utilised, as shown in Figure 6. The solution procedure mathematical model presented in earlier sections. The authors’
for optimisation is based on constrained minimisation of Equa- anguish has been noted in the dearth of experimental data avail-
tion (1) as an objective function. For this purpose, MATLAB able for validation purpose. However, two sets of experimental
provides the command fmincon in which limits of guessed val- work have been used for validating the accuracy of the mathe-
ues (i.e. trim and wetted length) are considered as inputs. By matical model.
S324 A. DASHTIMANESH ET AL.

Experimental Results of Lee et al. (2014) Proposed Method

0.25 4.00 0.52

Τ (deg)

S (m2)
R/Δ
3.50 0.48
0.20
3.00 0.44
0.15 2.50 0.40
2.00 0.36
0.10 1.50 0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(a)
0.20 4.50 0.52

Τ (deg)

S (m2)
R/Δ

0.18 4.00 0.48

0.16 3.50 0.44
0.14 3.00
0.12 0.40
2.50
0.10 0.36
2.00
0.08 0.32
0.06 1.50
1.00 0.28
0.04
0.02 0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(b)
0.20 5.00 0.52
Τ (deg)

S (m2)
R/Δ

0.18 4.50 0.48

0.16 4.00
0.44
0.14 3.50
0.12 3.00 0.40
0.10 2.50 0.36
0.08 2.00 0.32
0.06 1.50
0.28
0.04 1.00
0.02 0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(c)

Figure . Comparison between predicted results and experimental data, Case#: (a)  = . N; (b)  = . N; (c)  = . N. (This figure is available in colour online.)

Table . Details of model C (Taunton et al. ). wetted surface thus evaluated have been presented in Figures 8
Characteristics Values and 9, and have been compared against experimental data of
Taunton et al. (2010). It is abundantly clear that there is a
L . m generally good agreement between mathematical model and
B . m
T . experimental data over the entire range except at velocity
Displacement,  . kg of 4.05 m/s. The average relative error evaluated for trim is
β . degrees around 17% and has been derived as per Equation (14) as
lcg from transom . L
shown:

|τEXP − τPredicted |/τEXP
n
In the first instance, model C2 (Figure 7) introduced by ετ = (14)
n
Taunton et al. (2010) has been used. Model C2 is a two-stepped
hull whose characteristics have been presented in Table 1. Local
features of model C2 can easily be extracted from hull lines pre- However, if the velocity 4.05 is ignored, the error will reduce
sented in Figure 7. to 9%. Hence, it can be argued that LWP is a reasonable and suit-
Model C2 has been tested over a wide range of velocities rang- able assumption for high-speed range.
ing from 4.05 to 12.05 m/s. Accuracy of developed mathemati- Furthermore, obtained wetted surface area has been com-
cal model have been evaluated over the entire range. Trim and pared against experimental data. The average relative error is
 SHIPS AND OFFSHORE STRUCTURES S325

Experimental Results of Lee et al. (2014) Proposed Method

0.25 3.50 0.52

Τ (deg)

S (m2)
R/Δ
3.00 0.48
0.20
2.50 0.44
0.15 0.40
2.00
0.36
0.10 1.50
0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(a)
0.25 4.00 0.56

Τ (deg)

S (m2)
R/Δ

3.50 0.52
0.20 0.48
3.00
0.44
0.15 2.50
0.40
2.00
0.36
0.10 1.50
0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(b)
0.25 5.00 0.56
Τ (deg)

S (m2)
R/Δ

0.52
0.20 4.00 0.48
0.44
0.15 3.00
0.40
0.36
0.10 2.00
0.32
0.05 1.00 0.28
0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(c)

Figure . Comparison between predicted results and experimental data, Case#: (a)  = . N; (b)  = . N; (c)  = . N. (This figure is available in colour online.)

around 27%. It appears that trend of both mathematical and be a reasonable assumption in the speed regime investigated.
experimental data deviates from each other to a considerable Therefore, it is extremely important to perform extensive set of
extent. In fact, there is an unusual oscillatory behaviour in exper- experiments to develop a general empirical formula for flow sep-
imental wetted surface areas. With increased velocity, flow sep- aration from each step. It may be a good suggestion to use Savit-
aration from steps is intensified and wetted surface will reduce sky and Morabito’s (2010) wake relations in low speed range and
(Lee et al. 2014). However, obtained wetted surface areas have linear wake theory in high-speed regime.
a reasonable reduction with increased velocity. Overall, it is Lee et al. (2014) have conducted a systematic set of model
expected that by increasing velocity, ventilation length increases tests to understand the effects of displacement and step loca-
and wetted surface reduces. The LWP of the mathematical tion on performance of a two-stepped hull. Their experimental
model reproduces this behaviour quite convincingly. In the sec- work involving seven different step arrangements and at three
ond part of validation, comparative analysis of wetted surface displacements over a range of four velocities in calm water have
against velocity will be revisited. been shown in Table 2. However, the Case#1 had no step and
In Figure 10, predicted resistance has been compared against will not be considered for our validation exercise. The body
experimental data. As can be observed from the figure that there plan of Lee’s model known as NSWC15E has been shown in
is a relatively good correlation between experimental and math- Figure 11. The principal characteristics of NSWC15E have also
ematical model solution especially for speed exceeding 7 m/s. been presented in Table 3. It is suggested that the interested
Average relative error for resistance has been computed to be reader may refer to article of Lee et al. (2014) for further details
only 9%. It is quite apparent that linear wake theory appears to about NSWC15E.
S326 A. DASHTIMANESH ET AL.

Experimental Results of Lee et al. (2014) Proposed Method

0.25 4.00 0.52

Τ (deg)

S (m2)
R/Δ
3.50 0.48
0.20
3.00 0.44
0.15 2.50 0.40
2.00 0.36
0.10 1.50 0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(a)
0.25 4.00 0.52

Τ (deg)

S (m2)
R/Δ

3.50 0.48
0.20
3.00 0.44
0.15 2.50 0.40
2.00 0.36
0.10 1.50 0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(b)
0.20 4.50 0.52
Τ (deg)

S (m2)
R/Δ

4.00 0.48
0.16 3.50 0.44
3.00
0.12 0.40
2.50
0.36
2.00
0.08 0.32
1.50
1.00 0.28
0.04
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)
Resistance Trim angle Weed Surface
(c)

Figure . Comparison between predicted results and experimental data, Case#: (a)  = . N; (b)  = . N; (c)  = . N. (This figure is available in colour online.)

In this part of the study, the presented mathematical model Table . Main characteristics of Lee’s model (Lee et al. ).
has been used to predict trim, resistance and wetted surface Length overall  inches . m
for each case against the experimental data of Lee et al. (2014). Beam overall  inches . m
Comparisons have been performed against the three displace- Maximum chine beam . inches . m
Design displacement  lbf . kg
ments and four velocities which are given in Figures 12–17 Deadrise angle  degrees  degrees
depicting Cases#2–7.

Table . Step arrangement for Lee’s model (Lee et al. ). It is observed that there is a relatively good agreement
between results of the mathematical model and experimental
Height of forward Height of aft step from
Cases step (%B) forward step (%B) data. In the case of resistance to weight ratio (Rt/) for Case#2,
average error is 13%, 17% and 19%, respectively (Figure 12).
  
 . .
Therefore, it can be concluded that accuracy of mathemati-
 . . cal model decreases with increase in displacement. However,
 . . accuracy of trim calculation appears to be unaffected by change
 . .
 . .
in displacement (32% for all displacements). Mathematical
 . . model for prediction of wetted surface has been shown to be
excellent and further improves with increasing displacement
 SHIPS AND OFFSHORE STRUCTURES S327

Experimental Results of Lee et al. (2014) Proposed Method

0.25 3.50 0.52

Τ (deg)

S (m2)
R/Δ
3.00 0.48
0.20
2.50 0.44
0.15 0.40
2.00
0.36
0.10 1.50
0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)

(a)
0.25 4.00 0.52

Τ (deg)

S (m2)
R/Δ

3.50 0.48
0.20
3.00 0.44
0.15 2.50 0.40
2.00 0.36
0.10 1.50 0.32
1.00 0.28
0.05
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)

(b)
0.20 4.00 0.56
Τ (deg)

S (m2)
R/Δ

3.50 0.52
0.16 0.48
3.00
0.44
0.12 2.50
0.40
2.00
0.36
0.08 1.50
0.32
1.00 0.28
0.04
0.50 0.24
0.00 0.00 0.20
7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00 7.00 8.00 9.00 10.00
V (m/s) V (m/s) V (m/s)

(c)

Figure . Comparison between predicted results and experimental data, Case#: (a)  = . N; (b)  = . N; (c)  = . N. (This figure is available in colour online.)

(from 5% to 1% average error). In contrast, resistance calcu- However, accuracy of wetted surface calculation decreases with
lation shows degradation in accuracy while it is well known increasing displacement (from 7% to 12% average error) which
that resistance depends to a large extent on wetted surface area. can be due to correctness of linear wake theory by heightening
It is the understanding of the authors that the degradation of the aft step.
resistance with increasing displacement may have resulted from The accuracy of wetted surface prediction appears to degrade
spray resistance. Previous studies have indicated that both trim with increasing height of aft step as depicted in Figure 14. It
and wetted surface tend to increase with increasing displace- may be due to variation in fluid flow separation from aft step
ment (Ghadimi et al. 2015, 2016). With increasing spray angle, which cannot be appropriately modelled by linear wake the-
spray resistance tends to be amplified (Ghadimi et al. 2014). ory. Precision of trim calculation appears to be unreliable (36%
Consequently, it can be concluded that to improve the math- average error). However, resistance is reasonably well predicted
ematical model in resistance prediction, other components of (8% average error). It can be justified that by increased height
resistance such as spray resistance should be included in math- of the aft step, wetted surface of aft hull decreases, significantly
ematical model in near future. and consequently, this area has small effect on total resistance.
In Figure 13, the results thus determined have been com- Generally, it can be concluded that accuracy of linear wake the-
pared against experimental data for Case#3 which show simi- ory is limited by the ship’s displacement and step height. How-
lar trends as exhibited in Case#2. Again, average errors in resis- ever, lack of sufficient data precludes the quantification of these
tance prediction increase with displacement (from 12% to 16%). limitations.
S328 A. DASHTIMANESH ET AL.

For Case#5 and Case#7, height of forward step has been r As observed in obtained results, the accuracy of linear
increased while the height of aft step is fixed, as shown in wake theory is limited by the ship’s displacement and step
Figures 15 and 17. It is evident that increased height of the for- height. Moreover, linear wake theory is only usable for a
ward step has also some effects on validity of linear wake the- particular range of step height. At present, there is no suf-
ory for wetted surface predictions. However, general agreement ficient experimental information to be able to recognise
between mathematical results and experimental data has been relations of step height, displacement and LWP. Therefore,
shown to be acceptable and observed variations in results can be these relations must be quantified in future experiments.
described by both increasing in ship displacement and increased r It is evident from results that increased height of the for-
height of forward step. Results presented so far indicate that the ward step has an insignificant effect on linear wake theory
increased height of the forward step has insignificant effect on validity rather than increased height of the aft step.
linear wake theory validity rather than increased height of the r Although the presented mathematical model has achieved
aft step. Nonetheless, there exists an urgent need to conduct a relatively good agreement in almost all configurations,
extensive set of experiments for various stepped hulls and there exists an urgent need to conduct extensive set of
extract the flow pattern behind each step. experiments for various stepped hulls and extract the flow
Finally, the height of forward and aft steps was increased, pattern behind each step.
simultaneously. The obtained results have been compared
against experimental data shown in Figure 16. Effects of Disclosure statement
increased step height on accuracy of wetted surface predictions No potential conflict of interest was reported by the authors.
show that again, linear wake theory appears to be only usable for
a certain range of step height which must be quantified in future
References
experiments.
Clement EP. 2003. A configuration for a stepped planing boat having min-
It is apparent that presented mathematical model exhibits rel- imum drag. Bethesda (MD): David Taylor Model Basin, US Naval sur-
atively good agreement in almost all configurations and can cer- face Warfare Center.
tainly be utilised for engineering applications especially in the Clement EP, Desty DH. 1980. The BP Dynaplane high-speed research boat.
absence of any other validated mathematical model for two- Paper presented at: High-Speed Surface Craft. 3rd International Hov-
stepped hulls. ering Craft and Hydrofoil Exhibition; Brighton, UK.
Clement EP, Koelbel JO. 1991. Effects of step design on the performance of
planing motorboats. Paper presented at: 4th Biennial Power Boat Sym-
posium. The Society of Naval Architects and Marine Engineers; Miami
5. Conclusion (FL), USA.
In this paper, a new mathematical model has been presented Clement EP, Koelbel JO. 1992a. Optimized designs for stepped planing
for performance prediction of two-stepped hulls. A detailed monohulls and catamarans. Paper presented at: HPMV-92, Intersociety
High Performance Marine Vehicles Conference and Exhibit; Washing-
description of two-stepped hulls has been described by divid- ton (DC), USA.
ing the hull into three components and planing characteristics Clement EP, Koelbel JO. 1992b. Progress during the past century toward the
of each individual planing surface have been illustrated. In the development of efficient, load-carrying, stepped planing boats. Paper
absence of any theory, LWP has been adopted between the plan- presented at: 5th Biennial Power Boat Symposium. The Society of Naval
ing surfaces. A computational procedure has been discussed and Architects and Marine Engineers; Miami (FL), USA.
Clement EP, Pope JD. 1961. Stepless and stepped planing hulls graphs
computational algorithm has been developed for the determi- for performance prediction and design. Bethesda (MD): David Taylor
nation of trim and resistance of two-stepped hulls. In order to Model Basin, US Naval surface Warfare Center.
validate the mathematical model, experimental data from the Coleman TF, Li Y. 1994. On the convergence of reflective Newton methods
literature survey have been utilised. A comparative analysis per- for large-scale nonlinear minimization subject to bounds. Math Prog.
formed between the mathematical model and experimental data 67:189–224.
Coleman TF, Li Y. 1996. An interior, trust region approach for nonlinear
illustrates reasonably good agreement for both cases. In light minimization subject to bounds. SIAM J Optim. 6:418–445.
of these observations, the following conclusions and limitations Danielsson J, Strømquist J. 2012. Conceptual design of a high speed supery-
related to presented mathematical model could be made: acht tender hull form analysis and structural optimization [thesis].
Stockholm: Marina System Centre for Naval Architecture, KTH Uni-
r Based on the obtained results, it was observed that the pre- versity.
Froude W. 1872. Experiments to determine the resistances at various speeds
sented model has insufficient accuracy in low speed range. of a ship of 2500 tons: Designed by reverend C. Meade Ramus. Torquay
Authors believe that mentioned inaccuracy is related to (UK): Institution of Naval Architects, Admiralty.
LWP which is improper for low velocities. Therefore, Garland W. 2011. Stepped planing hull investigation. Trans Soc Nav Arch
implementation of Savitsky and Morabito (2010) wake for- Mar Eng. 119:448–458.
mulas in low speed range and linear wake theory in high Ghadimi P, Tavakoli S, Dashtimanesh A, Pirooz A. 2014. Developing a com-
puter program for detailed study of planing hull’s spray based on Mora-
speed regime are suggested.
r From previous experimental studies, it has been under- bito’s approach. J Mar Sci Appl. 13:402–4015.
Ghadimi P, Tavakoli S, Dashtimanesh A, Zamanian R. 2016. Steady per-
stood that fluid flow spreads out from step and chine formance prediction of a heeled planing boat in calm water using
in some situation which can lead to an added resistance asymmetric 2d+t model. Proc Inst Mech Eng M J Eng Marit Environ.
known as spray resistance. In order to improve the math- doi:10.1177/1475090216638680.
Ghadimi P, Tavakoli S, FeiziChekab MA, Dashtimanesh A. 2015. Introduc-
ematical model accuracy, other components of resistance ing a particular mathematical model for predicting the resistance and
such as spray resistance should be included in mathemati- performance of prismatic planing hulls in calm water by means of total
cal model in near future. pressure distribution. J Nav Arch Mar Eng. 12:73–94.
 SHIPS AND OFFSHORE STRUCTURES S329

Kaidy R. 2013. Advanced topics in stepped hull design. Paper Presented Svahn D. 2009. Performance prediction of hulls with transverse steps [the-
at: IBEX 2013. International Boat Builders Exhibition and Conference; sis]. Stockholm: Marina System Centre for Naval Architecture, KTH
Louisville, USA. University.
Lee E, Pavkov M, Mccue-Weil W. 2014. The systematic variation of step Taunton DJ, Hudson DA, Shenoi RA. 2010. Characteristics of a series of
configuration and displacement for a double-step planing craft. J Ship high-speed hard chine planing hulls part I: performance in calm water.
Prod Des. 30:89–97. Int J Small Craft Technol. 152:B55–B75.
Moore WL. 1967. Cambered planing surfaces for stepped hulls – some the- Taunton DJ, Hudson DA, Shenoi RA. 2011. Characteristics of a series of
oretical and experimental results. Bethesda (MD): David Taylor Model high-speed hard chine planing hulls part 2: performance in waves. Int
Basin, US Naval surface Warfare Center. J Small Craft Technol. 153:B1–B22.
Sahoo PK, Mason S, Tuite A. 2008. Practical evaluation of resistance of Vitiello V, Miranda S, Balsamo F, Bove A, Caldarella S. 2012. Stepped
high-speed catamaran hull forms-part II. Ships Offshore Struct. 3: hulls: model experimental tests and sea trial data. Paper Presented at:
239–245. 17th International Conference on Ships and Shipping Research; Naples,
Savitsky D. 1964. Hydrodynamic design of planing hull. Mar Technol. 1:71– Italy.
95. White G, Beaver W, Vann D. 2012. An experimental analysis of the effects of
Savitsky D, Morabito M. 2010. Surface wave contours associated with the steps on high speed planing boats. Paper Presented at: The 3rd Chesa-
forebody wake of stepped planing hulls. Mar Technol. 47:1–16. peake Power Boat Symposium; Annapolis (MD), USA.

View publication stats