RINA International Conference THE WILLIAM FROUDE CONFERENCE: ADVANCES IN THEORETICAL AND APPLIED HYDRODYNAMICS – PAST AND FUTURE 24 - 25 NOVEMBER 2010 PORTSMOUTH, UK The Royal Institution of Naval Architects
RINA
International Conference
The William Froude ConFerenCe:advanCes in TheoreTiCal and applied
hydrodynamiCs – pasT and FuTure24 - 25 November 2010
portsmouth, uK
The royal institution of naval architects
CFD MODELLING OF PLANING HULLS WITH PARTIALLY VENTILATED BOTTOM
Stefano Brizzolara
, Alessandro Federici§, Marine CFD Group,
University of Genoa, Italy;
§ University Pole of La Spezia, Italy
SUMMARY
Paper intends to present the main results achieved from a CFD investigation about planing hulls with partially ventilated
bottom, with the aim to characterise the main feature and behaviour of these important class of planing hulls, of which
very few hydrodynamic results and design data are known. After a brief introduction about the level accuracies obtained
in the simulation of conventional planing hulls with a state of the art RANSE VOF solver, with specific adaptation
introduced by the Marine CFD Group, mainly about mesh type generation, the paper will focus on stepped hulls, dealing
with studies aimed to the correct mesh definition, to the correct problem definition and numerical schemes. The
presented application cases regard two quite different hulls types: the first is a deep-V prismatic hull form, with a
cathedral type of spray rail and transverse step whose numerical results are compared with experiments in terms of total
resistance and visual comparison of the two phase flow below the bottom. The second example is a lower deadrise
planing hull with mixed convex/concave ordinates, warped and tapered at stern for which a rather comprehensive set of
towing tank tests results is presented and used for validation purposes of the CFD models.
NOMENCLATURE
BOA Breadth over all (m) BWL Waterline breadth in static condition
CRT Vol Volumetric Total Resistance Coefficient
defined as (2 RT -1 V-2 -2/3)
Fn Volumetric Froude number,
calculated as (V g -1/2 -1/6)
V0 Velocity of undisturbed flow (m s-1)
Displaced volume (m3)
g Gravity acceleration (9.8066 m s-2)
FnL Length Froude number IAV Draft on fwd perp. (mm)
IAD Draft on aft perp. (mm) LPP Length between perpendicular (m)
LWL Waterline length in static condition (m)
LCG Longitudinal Centre of Gravity, zero XG point is at Stern (m)
RO Wave resistance (N)
RA Shear resistance (N) RT Total advance resistance (N)
VS Ship velocity (m s-1)
VM Model velocity (m s-1)
S Trim angle in static condition, positive
means Stern sinking (deg)
D Trim angle in dynamic condition, positive is Stern sinking (deg)
Specific mass of water (kg m-3)
1. INTRODUCTION
The simulation of the hydrodynamics characteristics of
planing hulls is rather complex and involve different
fluid phenomena, such as large free surface deformation,
production of jet flows and sprays, wave breaking,
viscous boundary layer development in pressure gradient
and flow separation from sharp edges. The numerical
solution of the free surface RANSE seems the most
promising and comprehensive method to take into
account of the shape variability of hull forms, as other
faster methods based on potential flow solution [1] can
do, with the capability of considering also the effects of
different types of appendages. In fact, the most recent
examples of numerical studies in the field are using these
kind of flow solvers [2] [3]. The Marine CFD Group of
Genoa also brought its contribution in this field.
This paper, in fact, continues the series of studies
dedicated to RANSE simulations of planing hulls, the
first studies being published in [4] and [5] about an
original 2DOF method developed into a commercial
RANSE solver in order to more quickly find the running
attitude of planing hull at speed; more recent application
on different types of contemporary fast crafts, complete
of details and propulsion system, such as spray rails and
waterjet inlets in [6]; and finally systematic studies about
hydrodynamic forces on stern appendages for trim setting
and control in [7].
Figure 1 – Conventional (top) and stepped planing hull
(bottom) . Wetted surface at rest is evidenced in red.
Present paper, instead, is focusing the attention on the
specific problem of stepped hulls (see figure 1) with
partially ventilated bottom. To the authors knowledge
few studies have been published on this argument,
although planing hulls with transverse steps have been
used since around 1930 (Italy with Baglietto shipyard
being one of the innovators) and are still widely used as
primary mean of drag reduction in very fast planing hulls
from racing to fast attack crafts and motor yachts.
Among the few authors that published studies about the
argument, for sure there is Eugene Clement who after
several model test campaign devised a simplified design
method for planing hulls partially ventilated bottom [8].
The general lack of reference data about the achievable
drag reduction and the portion of ventilated bottom
obtained with a certain design of transverse steps make
the design of these kind of hulls rather arbitrary and in
general always require a set of model tests to confirm (if
not optimize) the intended effect. In this respect, this
paper also intends to introduce a series of reference
experimental tests conducted in the years 1930 in the first
Italian towing tank, located in La Spezia naval base. The
towing tank (146x6x3 meters) built on 1887 and
destroyed during WW2, was at the time the fourth in
Europe, and types of studies made there had many
analogies with those of William Froude and his son
Edmund in Torquay, although aside the tests on warship
models, also smaller fast units, like the planing attack
crafts for Xa MAS, taken as second example of hull for
the validation of CFD results in this paper.
The first type of hull form assumed in the study and
presented in the next chapter, instead, is a rather
conventional deep-V planing hull, conceived for a super
high speed ferry design that was modified by the
designer with a cathedral type of spray rail and a
transverse step with forward rake. The level of
confidence of RANSE simulations for this kind of hull in
its original and stepped version will be presented in the
following two paragraphs. Finally the last paragraph will
be dedicated to a systematic numerical analysis of the
influence of the step height on the resistance
characteristics of a different type of hull.
2. RANSE SOLVER
A state of art of RANSE solver with VoF method to
represent the free surface has been selected to simulate
the Flap and the Interceptor behind a prismatic planing
hull. The software suite [9] has the capability to solve the
turbulent viscous flow around a body in stationary
conditions, with the VoF method to predict the free
surface around it.
The solver is applied to the following group of equations
which express the mass and momentum balance with an
Eulerian approach and Reynolds time-Average approach
with the appropriate boundary conditions valid for the
specific type of problem. The RANS equations can be
expressed for an incompressible flow as follows:
MSTUPU
U
Re
0
(1)
where U is the average velocity vector field, P is the
average pressure field, μ is the dynamic viscosity, TRe is
the tensor of Reynolds stresses and SM is the vector of
momentum sources. The component of TRe are computed
in agreement with the k-ε turbulence model selected for
this application:
kC
kCgraddivdiv
t
kgraddivkdivt
k
kkxx
ijijt
k
t
ijij
k
t
ijijtij
i
j
j
itij
2
21
Re
2
2
3
22
3
2
DDU
DDU
DDUU
(2)
Where μt is the turbulent viscosity, k is the turbulent
kinetic energy and ε is the dissipation term of turbulent
kinetic energy. The realizable k-ε turbulence model was
selected to close the hydrodynamic problem together
with a two layer wall function applied for the cell near
the wall.
The wall function is an analytical treatment for the first
cell near the wall where the velocity vector and all other
scalar quantities are extrapolated from the known
quantities on the wall boundary surface. The two layer
wall function model, is a model that impose a first thin
linear layer near the wall, and a second logarithmic layer
over the first, this model assume that the first cell
centroid near the wall is lies within the logarithmic
region of the boundary layer. The wall treatment is
optimized to compute a mesh with a y+<100. All the
available models with relative analytical formulations are
listed in [9].
The RANS solver is based on a Finite-Volume method to
discretise the physical domain. The equation for an
incompressible multi-phase fluid has been used in the
simulation, with one more transport equation for the
VoF, that represent the fraction of water present inside
each cell.
0
UVoFdiv
t
VoF (3)
This new equation guarantees to find the correct shape of
free surface defined by the point of transition between
water and air, this method is powerful for the problem
when occurred the wave breaking, or when the air effects
are important on the free surface shape.
The simulation has been built to reach a stationary
solution but it can reached its only across a time-
marching solution to guarantee to update the solution of
VoF quantity.
To solve the time-marching equations, it’s used an
implicit solver to find the field of all hydrodynamic
unknown quantities for each time step, in junction with a
iterator solver for each time step. The software uses a
SIMPLE method to conjugate pressure field and velocity
field, and a AMG (Algebraic Multi-Grid) solver to
accelerate the convergence of the solution.
3. ON THE SET-UP AND ACCURACY FOR A
RANSE MODEL OF A THE DEEP-V BASIC HULL
We are used to refer to pure planing condition when the
hull is running at an advance volumetric Froude number
of Fn > 2.5-3.0, in that case the weight of the boat is
just almost completely sustained by dynamic lift force
due to the dynamic pressure distribution acting on the
hull bottom: the wave resistance practically becomes
negligible and main drag components are the shear drag,
followed by induced resistance (from lift), both
proportional to the square of the velocity. In this
condition, hydrostatic component of lift, proportional to
the submerged volume is small compared to the total
displaced volume at rest. The dynamic trim is important
since it changes the wetted surface and its length as well
as . So dynamic trim induce an important non linearity
into the resistance component, depending on the center of
gravity position and to the hull form, at pre-planing
speeds, augmenting their value with respect to a pure
quadratic behavior, while in pure planing regime the
dynamic trim generally decreases, the dynamic rise
generally increases, with the result of a reduction of
induced drag and lower frictional resistance. Using this
effect we can highly increase the speed, without having a
prohibitive increasing of drag. Variable deadrise angles
of the ordinates along length (warped hulls), bended
longitudinal sections in the aft part of the hull (rocker or
hook) as well as tapered beam and spray rails can all
have an influence in the hydrodynamic behavior of the
hull in pre-planing and planing regimes, since they can
have a direct influence on the dynamic trim and rise
experienced by the vessel at high speed.
Another way of playing on the running attitude of the
vessel and of reducing its wetted surface at high speeds is
to ventilate part of the bottom surface, by realizing a
transverse step at a proper longitudinal position aft of the
spray root line and by opening a to create and maintain
an air pocket on the hull bottom aft of the step (see
Figure 1). Eventually this the water flow separating from
the lower step edge, forming a free surface, will reattach
to the bottom after a certain longitudinal distance to
create another wetted portion of the hull forward of the
transom, from which will separated again to form the
wave trough in the hull wake. The length and volume of
the air pocket beneath the hull and is of course very
much influenced by longitudinal position, the height and
the sweeping angle of the step.
3.1 CHARACTERISTICS OF FIRST HULL
The first deep-V planing hull (about 25 degree of
deadrise) was given to the authors as a reference hull for
the CFD validation study and was already tested with a
systematic campaign of model tests in towing tank by the
designer, with and without steps. The hull devised for a
super-fast ferry of about 70m in length, has been created
without deck and superstructure, and it is cut with a
simple flat deck also in the CFD model and the
aerodynamic forces developing on it are neglected in the
calculation of the total drag.
The deep-V planing hull has a monohedric shape aft of
midship, ending in a vertical transom and has a rather
long slender entrance body to ensure good seakeeping. It
has a single hard chine with a flat spray rail to better
diverge the separated spray flow from the side at high
speeds, while augmenting the lift force. Figure 9 presents
the body plan of the stepped version of this hull, whose
simulations are described later.
As already discussed in previous studies in case of the
some hulls of series 62 [5], the correct prediction of the
jet spray separation from chines is the most critical issue
to obtain valid results from the RANSE solvers. This is
critically discussed in the next paragraph with regards to
the mesh type to be used.
3.2 MESH SPECIFICATIONS
The key for a correct solution of the flow in the jet spray
region forward of spray root line and the separated flow
from chine aft of the spray root line is a correct mesh
topology and resolution. In parallel to the mesh density,
to obtain a valid courant number also at high speeds, the
time-step must be reduced. These two requirements
together imply a rather high computing time needed to
reach the steady state solution.
As regards the mesh, a very high refinement of cells is
preferable around the free surface in the spray region to
minimize a non physical inclusion of air under the keel.
Figure 2 present the VoF distribution obtained with two
different mesh resolution: colour scale represent the
percentage of fraction of water (volume of fluid)
calculated in any cell. So the cells with 100% of water
are coloured in red, while in blue those with 100% of air.
For the less dense mesh, a marked numerical diffusion of
air under the hull bottom is noted; this error artificially
decreases the frictional resistance, because, as from the
volume of fluid theory, cells with only a fraction of
water, will experience a proportionally less dense and
viscous fluid. On the other hand the wetted surface tends
to increase, with this numerical air diffusion, because the
spray root line becomes less sharply defined, and the
pressure area increases its extension and so the pressure
resistance. The result of the above described numerical
diffusion has not much influence on the predicted
resistance. In fact, the decrease of shear force is balanced
by a virtual increase of pressure resistance, so that the
total drag does not vary much between different mesh
resolutions. So the mere looking at the resistance will not
give a good indication about the proper mesh to use. One
should better analyse the VOF distribution under the hull
bottom to judge if the mesh type and resolution are valid.
In fact, in the high refined mesh case, it is noted that the
spray root line results very well defined and it has a more
realistic swept back angle as well.
This is because the VoF interface capturing method, fill
those mesh cells cut by the free surface, with a fraction
of water (between 0% and 100%) and the complement of
air, so that those cells have a hybrid fluid obtained by a
relative mixture of the two fluids, air and water, with
material proprieties, density and viscosity, that
correspond to the weighted average of filling ratio of
each fluid in the cell. The transport equation, then,
diffuse the mixture from partially filled cells to the
contiguous ones, bringing the flow mixture also below
the hull bottom. Naturally by refining the mesh (compare
figure 2), the same effect is still experienced, but on a set
of smaller cells, occupying a smaller overall volume. In
this way, the effect on the results is less pronounced,
although, not completely eliminated.
Graph 1 –Shear and pressure resistance ratio on total predicted
resistance for different mesh resolutions, obtained by CFD on
the first hull model without step
Graph 1 presents the variation of relative weight of
shear/pressure resistance on total resistance predicted
with CFD models by systematically varying the mesh
resolution from 250k cells to 5.7M cells, confirming the
above mentioned trend.
For practical purposes, if a good trade off between
computational costs and solution accuracy, a number of
about 1M cells with a nested refinement as that presented
in figure 3 is sufficient to obtain accuracies in the range
of 2% as verified in many different cases[11]. The
comparison of the predicted resistance by the present
method and model test total resistance volumetric
coefficient curve is given in Grahp 2 at the end of the
paper.
4. STEPPED HULL FORMS
A stepped hull is generally obtained from a conventional
planing hull by cutting, or inversely extruding, one or
more transverse steps on the bottom surface (figure 1).
As already discussed earlier in the paper, the
hydrodynamic function of the step is to create a negative
pressure zone immediately behind itself, in order to
naturally (sometimes also forcedly) create a separated
free surface flow downstream of its lower edge, with an
air pocket of a certain extent attached to the hull. Its
action must be complemented, in the simplest case, by a
way for the air to be connected with and flow into the
depressurized air pocket. Otherwise the step would only
increase the resistance, directly because the negative
pressure pulls behind the step back face, indirectly
because the reduced pressure below the hull bottom tends
to sink the hull and increasing the necessary trim angle to
obtain the same lift force. It is clear that the position and
sweep angle of the step line and the step height as well as
the way of air ventilation, are all important parameters to
be considered to optimize the resistance reduction of a
stepped hull. In general the beneficial effect is
experienced only above a certain speed, around the hump
speed of the hull. Below that the resistance
characteristics of the stepped version will be poorer than
the original hull form.
Incidentally it is worth to mention that beside the
advance resistance, another important effect to be
investigated to design stepped planing hulls is the
dynamic stability of the hull, i.e. porpoising and chine
walking phenomena, which can be effectively studied
with the same kind of RANSE solvers presented in this
paper.
Two rather different hulls forms are presented in this
paper: a deep-V, straight section, prismatic hull and
mixed convex-concave sections with much lower
deadrise angle. They will be referenced in the paper as
first and second model and their body plans are given in
figure 9 and figure 10, respectively.
4.1 GENERAL DESIGN
The Step divide the hull in two portions, one forward of
it, namely the forward-body, the other aft of it: the after-
body. The reduction of shear resistance obtained with the
stepped hulls starts in general to be significant for around
Fn=3.0 and rather important for Fn>5.0. So in general,
it is a good solution for very high speed crafts. For the
specific case of model-1, the predicted resistance
reduction obtained by ventilation of the bottom with the
given step design at different speed is presented in Graph
3, in which a rather limited reduction of about 5% is
noted at the top speed (Fn4.7).
For a proper design of the step it should be considered to
use only the Fwd-body to carry out the major part of the
lift (90%) to sustain the craft, while using the Aft-body
to give a small portion of lift force (10%) only to adjust
the trim angle to an optimum value; so attention is to be
paid for the proper definition of the geometry of the
After-body that must be consistent with the geometry of
the step.
4.2 DESIGN OF 1
st STEPPED HULL MODEL
The model, as originally designed, had a quite small step
(in height) with a negative sweep-back angle. The flat
spray rail at the chine, still present in the forward body
has been substituted by a very sharp-V type chine, as
visible on the vertical of the body plan of figure 9. The
0 20 40 60 80
100 120
5.700.000 Cells 2.100.000 Cells 1.000.000 Cells 465.000 Cells 250.000 Cells
Shear Pressione
original idea of the designer, was probably that to
contain the trapped air pocket below the hull bottom, in
similitude to what it is done in SES crafts. Model was
also equipped, in fact, with an air feeding system
composed by rectangular conduits connecting the deck
with the bottom. Tests and simulations were all
performed for the naked hull version, without
appendages.
4.3 DESIGN OF 2nd
STEPPED HULL MODEL
The Vessel model was made in scale 1:10 according to
Italian Navy's towing tank tests made in La Spezia on
1951. The model is named b112, as tested in series III
and series IV reported in [10].
The parent hull was a MAS hull, of the type still widely
used on hull form for small and fast planing crafts, also
for pleasure purposes. The step is rectilinear, without any
sweep-back angle, and it is obtained with a reduction on
the vertical plane of the step, namely "Bassofondo", to
bring the air from the side to the keel just downstream of
the step, as visible in the body plan of Figure 11. All The
sections are not straight and typical for this class of
boats; there is no chine plate, and an upright side with a
small negative angle. The width is reduced at stern with a
flat transom inclined of a small backward angle. The bow
is round with a deep-V sections.
The deck in the CFD model, has been realized with a
transversally flat surface that connect both sides, but, as
in the first model, the aerodynamic forces arising on it
are not considered in the calculation of total resistance.
Among all the other tests reported in [10], this model was
chosen because it was one of the few tested without
appendages and propellers.
4.4 MESH SPECIFICATION
If the problem of refining mesh around the free surface is
important for conventional planing hulls, this is even
more true for stepped hulls. In fact, in the latter case,
there are not only the problems of virtual air inclusion at
the bow (spray root line) and the flow separation from
the chine (water on side), but there is also that of the
correct prediction of the separated flow from step that
reattaches on the bottom of the after-body keel with the
same above problems. In fact, if the flow arriving at the
step already has a numerical inclusion of air coming from
the first spray root line, this gets worse when it reattach
on after-body after having mixed, also in the reality with
air in the air-cushion downstream the step.
So to accurately predict this complicated flow mixture
the mesh has to be highly refined in the afterbody,
between step and transom, as shown in the example of
Figure 3, that refers to the CFD model-1. In the same
figure, the ventilation pipes (in white) and the hole on the
chine necessary to bring that air behind the step.
Figure 4, instead, presents the comparison between
After-body and Forward-body spray root line solution. In
the Fwd-body the inclusion of air is limited, because the
free surface of the incoming is quite sharp and interests
only one cell; in the Aft-body, instead, the inflow is more
complex and many more cells are interested by a mixed
flow (air/water) that is caused by the wrongly predicted
separation behind the step, as visible in the Figure 5. A
method to predict a sharp separation also aft of the step is
needed to improve the solution, as will be presented in
the next paragraph.
4.5 INTERFACE SOLUTION
The proposed solution is to build a virtual interface
surface behind the step aligned with the local flow, in
order to refine the mesh in the proximity of the interface
with the mesher routine. The interface has only the role
to topologically divide the mesh along the guessed free
surface to align the cells sides with the free surface,
reducing the number of cells with a mixture of fluids.
The effect of this different mesh can be seen in Figure 5,
showing the flow capture without interface (top picture),
with interface and double prism layer around (bottom
picture) the surface. In the second case the flow is very
well captured. The combination of this virtual interface
with nested mesh refinement, previously discussed, can
ensure good results as presented in the comparison of top
and bottom pictures in figure 6.
The problem associated with this virtual interface
method, though, is of course that the free surface is not
known a priori, so a guess must be done and refined after
the first obtained solution. In our cases we notice that the
free surface is initially aligned with the Fwd-body keel
and after about 10 heights it turns in the direction of the
undisturbed flow. This behaviour can be approximated
with two planes at a certain angle between them.
Secondly, at present, the solver [9] does not accept the
interface model with the 6 DoF solver, so the trim and
rise must be first found with a more approximated model
and then imposed by trial and error in this more accurate
model. This procedure usually works with few
adjustments, but it is very time consuming. In fact, the
use of interface modifies the pressure distribution in the
after-body, so the trim changes from one interface shape
and the other and needs an iterative search of the right
trim (and sinkage). The following table summarises the
differences found between the simulation without
interface and the simulation with it. The drag values of
the two simulations have been corrected to correspond to
exactly the same lift force taken from the experimental
model.
Table 1 – Model1: calculated CFD resistance components
and correlation error with experimental results, Fn= 4.1
Drag Exp. Drag CFD Error % Drag Shear Drag Prex N N N N
Without 187.6 169.6 -10 90.0 79.6 With Interf. 187.6 176.1 -6 106.1 70.0
As expected the model with specially refined mesh get
closer to the experimental results, and as commented
before shear drag is increases and pressure drag reduces
with respect to the coarse mesh model. This is consistent,
since in the first case (w/o interface) the wetted surface is
higher, but is in contact with a fluid with less density and
viscosity than in the second case.
The underestimation error on shear force noted on the
model with a coarse mesh implies an corresponding
underestimation of the resistance reduction obtained with
a certain step design with respect to the original hull
form. In fact this underestimation error on shear is
expected also for full scale CFD simulations, though,
obviously it will be smaller in relative sense, being lower
at full scale the relative weight of shear to pressure
resistance.
4.6 COMPUTATIONAL ACCURACY
From a qualitative point of view, the predicted air/water
mixed flow pattern obtained with the above mentioned
CFD models is very similar to the observations done in
towing tank; we can see, for instance figure 7 and 8,
which present a photograph from below of the model
towed at Fn= 4.1 and the same view obtained from the
CFD model, in the same conditions. The two cases are
very close, in terms of compared the stagnation line,
spray area and wetted and ventilated portion of the
bottom: the spray extent in the far field, instead, is not so
accurately captured due to the increasing mesh cells size
toward the boundaries.
From a quantitative point of view, the level of accuracy
that can be obtained on the CFD simulation of stepped
hull can be better discussed on a more consistent and
systematic series of experimental tests made in case of
model 2. The tests on this hull model (figure 10) has
been repeated with two different longitudinal center of
gravity values. The complete result of the tests are
reported in Table 3 and Table 5, respectively, referred to
as Series III and Series IV. For each series of tests, not
only the resistance components, calculated with
ATTC’47 method used at that time, but also dynamic
trim and sinkage are reported for each run speed.
4.6 (a) Series III
Series III are a 27 runs testes with Fn going from 0.627
to 6.269. The model is a 1.55 LPP m long and 0.39 BOA m
large; with a weight of 15.107 kg and a LCG of 0.8 m.
The craft has a static angle of trim of S= -0.81°, positive
bow down, as reported in table 2.
The model geometry scale is 1:10.
Table 3 presents the measured resistance values in model
scale, and a tentative of interpretation of the resistance
components in model scale, performed on the basis of the
mean wetted length and surface taken from visual
observation during each run and ATTC’47 frictional
correlation line, following the practice used at the time of
tests.
In pure planning condition, the error on total resistance
between towing tank tests and CFD simulations, reported
in table 4 is very small; never exceeding 3.4 %. For the
case at 20 kn (Fn=2), the error is higher, but the flow
regime there is not purely planning and another mesh
typology and domain size should be used. For improving
the results, this case was studied with a more refined
mesh obtained with the virtual interface technique and
the respective result is reported at bottom of table 4; as
visible the error in the predicted resistance reduces, but it
is still high (10%) due to the limited domain size, valid
for pure planing conditions and not for semi-planing
regimes.
Trim predictions, differently to drag ones, present some
noticeable uncertainty, varying around ±10%. The
precision of the measurement of trim values at those
time, though is also uncertain, rending the discussion a
little difficult. Moreover, due to the ventilation of a large
portion of the aft body (at some speed the whole aft
body) the dynamic trim angle is not stationary but it
oscillates around a mean value, as noted from CFD
simulations. The same it is expected to happen also in the
model tests, though a single (averaged?) value was
reported. Again also in this case, the virtual interface
solution approach can minimize this error (from 12% to
5%) being the pressure distribution on after-body part,
that has a large influence on the trim angle, more
correctly predicted .
Graph 4 represents the trends of the predicted total drag
volumetric coefficient against volumetric Froude number
obtained by CFD simulations with the towing tank
results. As clear the accuracy is very good for Fn>2.5.
4.6 (b) Series IV
Series IV present similar trends with respect to Series III,
the mainly difference being the LCG set at 7.5m from the
transom. The weight and the wetted length are the same;
the static trim angle is S= -0.07°, as reported in table 2.
In table 5 we have all the experimental results from
towing tank about this condition.
This series of tests was investigated in more details with
a larger number of CFD runs, to derive a better defined
resistance curve. For this type of test, a very good
correlation is again found for the runs at high speeds, for
which a maximum error of about 2% is noted. At slow
speed, in pre-planning regime, the error rapidly increases
due to low Froude number, but in any case remains lower
than 6%.
In Graph 5, it can be noted also graphically how the total
drag coefficient trend over speed is close to that obtained
from towing tank measurements, at high speed. A final
notice can be done on dynamic trim angle D that
presents an average error of about 8%. Also in this case
the error on trim is increasing with speed, highlighting a
possible problem casued by the unsteady and fluctuating
nature of this signal over time. In fact particularly in this
series, at very highest speeds, the particular shape of the
free surface calculated aft of the step edge, keeps the
after-body floating in air although at a very close
proximity to the water. This interesting and rather
unusual condition is also confirmed by mode test
sketches, reporting wetted part of the after-body.
4.7 INFLUENCE OF GEOMETRIC
PARAMETERS
The influence of geometric parameters were studied with
regard to the last case, Series IV: the original step height
of the model tested in towing tank was taken as reference
height, while two additional ones have been added,
namely +30% and -30%. The CFD simulations covered
the complete planing regime range, with three calculation
speeds: 20, 30 and 40 kn in ship scale.
The geometry modification was restricted to the after-
body of the hull. In this respect, to increase the step
height the after-body bottom surface was parallelly
shifted up of a normal distance corresponding to the
height reduction; inversely to decrease the step height the
bottom surface is shifted downward.
4.7 (a) Step Height
Table 6 presents the CFD results obtained for the model
with the original step height, while Table 7 reports the
same Lift and Drag force components subdivided
between after-body and forward-body.
Complementally, Table 8 to Table 11, report the same
kind of results for two modifications of the step height: a
variation of ±30%. Graph 6 presents an extract of the
obtained CFD results in terms of friction and pressure
resistance components for the three different step heights
as a function of speed. As visible, the pressure resistance
of the forward body decreases when the step height
reduces, ad it is partly compensated by a contemporary
increase of the frictional resistance: this changes are
primarily due to the difference in dynamic attitude
assumed by the hull with different steps height (trim
increases and sinkage decreases with step height).
The shear resistance of the after-body reaches almost null
values at high speeds (Fn>3.0) for 0% and +30% step
heights since the wetted area of the bottom is nearly zero,
while for the lowest step the shear resistance maintains
on a considerable value. As regards the pressure
resistance, for high speeds, the best step variant seems to
be that of the original design: the lowest height being
better at the lowest speed as expected.
These conflicting trends noted for the shear and pressure
resistance make the final results in terms of total
resistance not very much dependent from the step height,
at least in the investigated range of variation. Total
resistance seems anyhow to be slightly decreased in the
highest speed range (Fn>3.0) for the smaller step.
Some more definite trends could be found with a more
significant step height variation. For this reason new
CFD calculations are currently planned and will give
definitive guidelines about step influence on drag
reduction. Naturally, if the step height is reduced too
much, there is the possibility to largely reduce or even to
loose the ventilated portion of the after-body, with the
result of having a significant increase of drag, due to
separated recirculating flow induced aft of the step
irregularity in the fully wetted flow under the hull
bottom.
5. CONCLUSIONS
Paper outlined the current techniques devised and applied
by the Marine CFD Group to increase the accuracy in the
numerical prediction of fast planing hull resistance in
calm water, in both cases for conventional hull forms as
well as for hulls with transverse steps. The techniques
regards mainly the generation of a proper topology for
the mesh and a refined mesh cells, aligned locally with
the predicted free surface shape in the stagnation line
region as well as in the separation region aft of the step.
In particular the presented CFD models are able to
accurately define the main physical aspects of the flow
under a planing hull with steps at high speeds: the
position and shape of the spray root line and the spray
area are correctly predicted; so is correctly predicted the
free surface separation from the step lower edge and the
ventilated portion of the bottom downstream of it, also
for complicated hull, step and chine shapes like those of
the presented model-1.
The level of accuracy on the numerical results achievable
with the described techniques are generally satisfactory
for engineering purposes, in the pure planing regime,
being in the average error within a 3% with respect to
reliable experimental results.
The hydrodynamic effect of a transverse step has been
discussed on the basis of the detailed CFD results, on two
very different hull shapes, and analysing the results
obtained for model-2 from a systematic variation of the
step height and test speed. In general, for the examined
cases, the effect of the step results beneficial, due to an
important reduction of the shear drag component with
respect to the original hull without step.
When the step height is reduced but still sufficient to
induce ventilation on the aft-body bottom, the CFD
simulations seem to indicate a reduction in total
resistance, primarily justified by a lowest value of the
running trim angle which in general means higher
induced drag (pressure) resistance on the forward body.
CFD models are useful in this respect not much to
evaluate the drag reduction difference between one step
height and another, but rather to find the lower step
height to ensure a proper after-body ventilation, the final
key for drag reduction for stepped hulls.
Eventually, once the minimum step height to induce
ventilation is set, a final CFD study on the most
appropriate shape of the after-body to ensure the lowest
trim angle and pressure resistance on the forward and
after-body is suggested and can be effectively performed
by the proposed CFD models.
6. AKNOWLEDGEMENTS
Authors wish to acknowledge the support of this work by
Promostudi University Pole of La Spezia within a
broader research program dedicated to the numerical
hydrodynamic simulations for the design of motor/sailing
yacht. A special thanks to Prof. Marco Ferrando for his
expert hint in using model tests of the towing tank of La
Spezia as reference data for CFD validations.
7. REFERENCES
1. Savander B.R., Scorpio S. M., Taylor R.K., “Steady
hydrodynamic analysis of planing surfaces”, Journal of
Ship Research, vol. 46, no.4, pp. 248-279, 2002
2. Caponnetto, M. “Practical CFD Simulations for
Planing Hulls”, 2nd
High Performance Marine Vehicles
Conference (HIPER), pp. 128-138. Hamburg, 2001.
3. Thornill E., Bose N., Veitch B., Liu P. “ Planing Hull
Performance Evaluation Using a General Purpose CFD
Code”, Proceedings of Twenty-Fourth Symposium on
Naval Hydrodynamics, NAP, 2003.
4. Brizzolara S., Serra F. “Accuracy of CFD Codes in the
Prediction of Planing Surfaces Hydrodynamic
Characteristics”. 2nd
Int. Conference on marine Research
and Transportation, ICMRT”07. ISCHIA. 28-30 June,
2007. (vol. 1, pp. A-1-A-12). ISBN: 88-901174-3-5
5. Villa D., Vatteroni G., Brizzolara S.“CFD Calculations
of Planing Hulls Hydrodynamics”, Star European
Conference, London, March 2009
6. Brizzolara S., Villa D., “CFD Simulation of Planing
Hulls”, Seventh International Conference On High-
Performance Marine Vehicles Melbourne, Florida, USA
13-15 Oct. 2010.
7. Brizzolara S., Villa D. “A systematic CFD Analysis of
Flaps / Interceptors Hydrodynamic Performance”, Fast
2009 Int. Conference on High Speed Fast Ship Design,
Athens, 2009.
8. Clement E.P., Pope J.D. “Stepless and Stepped
Planing Hulls – Graphs for Performance Predictions and
Design” DTMB Report 1490, Jan 1961.
9. CD-ADAPCO “Star-CCM+ User and Theory
Manual”, version 4.04.011, 2009
10. Permanent Commission for War Material and
Experiments, “Annale n°124, Fascicolo IV” series of
model tests results in the towing tank of La Spezia, 1951.
11. Federici A., “Prediction by CFD Methods about
Hydrodynamic Behavior of Planing Hulls with
Cambered Step and Stern Vee Hydrofoil”, Master
Degree Thesis with Press Dignity in Yacht and Power
Craft Engineering at University of La Spezia, December
2009.
8. AUTHORS BIOGRAPHY
Stefano Brizzolara, Naval Architect and Marine
Engineer, MRINA, MSNAME, PhD in numerical
hydrodynamics. is researcher at the University of Genoa,
where he holds the two courses of “computational
hydrodynamics for ship design” for the MSc degree in
naval architecture and yacht design.
Head of the Marine CFD Group, besides specialist
consultancy work for industries, he is currently guiding
different research projects for ONR, Italian Ministry of
Defence and EU all dealing with non conventional high
efficiency hull form and propulsors design, devised and
optimised by CFD methods. His previous experience
includes navy ships and propellers design in the
hydrodynamic design office of Fincantieri Naval
Business Unit and experimental research at the cavitation
tunnel of the Italian Navy in Rome.
Alessandro Federici, MSc in yacht engineering, holds a
research grant in La Spezia University Pole for
specializing, inside the Marine CFD Group, on the
application of RANSE solvers to typical yacht design
problems; in particular fast and non conventional hulls
simulations for resistance in calm water and motions in
waves.
Figure 2 – Model 1: VOF distribution on the hull surface. Mesh with 1M (top) and 5.7M cells (bottom)
Figure 3 – Model-1: generated mesh typology for the stepped version
Figure 4 – Model 1: VoF distribution in the spray root line region in the fore-body (top) and aft-body (bottom)
Figure 5 – Predicted Air/Water (VoF) distribution behind the Step with standard (top) and refined mesh typology (bottom)
Figure 6 – Comparison of VOF distribution between the case without interface (top) and with it (bottom)
Figure 7 - Underwater Photo from model tests– Flow pattern and bottom ventilation at Fn= 4.1
Figure 8 - RANSE Simulation results in the same case of figure 7
TABLES OF MODEL TESTS-CFD RESULTS for Model-2
LPP 15.500 dm
LWL 14.959 dm
BOA 3.900 dm
BWL 3.890 dm
IAV 72.7 mm
IAD 50.7 mm
XG 0.800 m
S -0.81 deg
LPP 15.500 dm
LWL 14.912 dm
BOA 3.900 dm
BWL 3.890 dm
IAV 62.0 mm
IAD 60.0 mm
XG 0.750 m
S -0.07 deg Table 2 - Main Geometrical and Load Characteristics of Model-2, for Test Series III (left) and IV (right)
Table 3 - Series III - Towing Tank Results
Table 4 - Series III - CFD Results
V S V S FnV FnL V M R O R A (15°C) R O R A (15°C) R T C RT Vol I AV I AD I AV I AD [Kn] [m/s] [m/s] [Kg] [Kg] [N] [N] [N] *1000 [dm] [dm] [mm] [mm] [deg]
6 3.087 0.627 0.255 0.976 0.104 0.106 1.020 1.040 2.060 70.78 0.03 -0.02 75.7 48.7 -1.00 8 4.116 0.836 0.340 1.301 0.302 0.184 2.963 1.805 4.768 92.19 0.09 -0.04 81.7 46.7 -1.29 10 5.144 1.045 0.425 1.627 0.812 0.285 7.966 2.796 10.762 133.05 0.13 -0.02 85.7 48.7 -1.37 12 6.173 1.254 0.510 1.952 1.258 0.390 12.341 3.826 16.167 138.86 -0.10 0.20 62.7 70.7 0.30 14 7.202 1.463 0.595 2.278 1.407 0.494 13.803 4.846 18.649 117.61 -0.26 0.31 46.7 81.7 1.29 16 8.231 1.672 0.680 2.603 1.574 0.628 15.441 6.161 21.602 104.34 -0.34 0.32 38.7 82.7 1.63 18 9.260 1.881 0.765 2.928 1.959 0.778 19.218 7.632 26.850 102.50 -0.37 0.31 35.7 81.7 1.70 20 10.289 2.090 0.849 3.254 1.953 0.872 19.159 8.554 27.713 85.66 -0.40 0.30 32.7 80.7 1.77 22 11.318 2.299 0.934 3.579 1.806 0.923 17.717 9.055 26.771 68.40 -0.52 0.31 20.7 81.7 2.25 24 12.347 2.507 1.019 3.904 1.689 0.958 16.569 9.398 25.967 55.76 -0.67 0.32 5.7 82.7 2.84 26 13.376 2.717 1.104 4.230 1.558 1.005 15.284 9.859 25.143 45.99 -0.79 0.34 -6.3 84.7 3.36 28 14.404 2.925 1.189 4.555 1.466 1.041 14.381 10.212 24.594 38.79 -0.89 0.33 -16.3 83.7 3.69 30 15.433 3.134 1.274 4.880 1.389 1.022 13.626 10.026 23.652 32.50 -0.96 0.31 -23.3 81.7 3.88 32 16.462 3.344 1.359 5.206 1.323 1.003 12.979 9.839 22.818 27.55 -1.00 0.29 -27.3 79.7 3.95 34 17.491 3.552 1.444 5.531 1.266 0.989 12.419 9.702 22.122 23.67 -1.04 0.27 -31.3 77.7 4.02 36 18.520 3.761 1.529 5.856 1.219 0.985 11.958 9.663 21.621 20.63 -1.08 0.26 -35.3 76.7 4.13 38 19.549 3.970 1.614 6.182 1.188 0.987 11.654 9.682 21.337 18.27 -1.12 0.24 -39.3 74.7 4.21 40 20.578 4.179 1.699 6.507 1.111 1.003 10.899 9.839 20.738 16.03 -1.14 0.21 -41.3 71.7 4.17 42 21.607 4.389 1.784 6.833 1.040 1.028 10.202 10.085 20.287 14.22 -1.15 0.19 -42.3 69.7 4.13 44 22.636 4.597 1.869 7.158 0.981 1.064 9.624 10.438 20.061 12.81 -1.16 0.17 -43.3 67.7 4.10 46 23.664 4.806 1.954 7.483 0.924 1.111 9.064 10.899 19.963 11.67 -1.16 0.14 -43.3 64.7 3.99 48 24.693 5.015 2.039 7.809 0.869 1.171 8.525 11.488 20.012 10.74 -1.14 0.11 -41.3 61.7 3.80 50 25.722 5.224 2.124 8.134 0.817 1.243 8.015 12.194 20.209 10.00 -1.12 0.08 -39.3 58.7 3.62 52 26.751 5.433 2.209 8.459 0.766 1.330 7.514 13.047 20.562 9.40 -1.11 0.06 -38.3 56.7 3.51 54 27.780 5.642 2.294 8.785 0.727 1.428 7.132 14.009 21.141 8.96 -1.09 0.04 -36.3 54.7 3.36 56 28.809 5.851 2.379 9.110 0.624 1.537 6.121 15.078 21.199 8.36 -1.08 0.03 -35.3 53.7 3.29 58 29.838 6.060 2.464 9.435 0.533 1.655 5.229 16.236 21.464 7.89 -1.08 0.02 -35.3 52.7 3.25 60 30.867 6.269 2.548 9.761 0.462 1.790 4.532 17.560 22.092 7.59 -1.09 0.01 -36.3 51.7 3.25
Trim Ship Model
V S V S FnV FnL V M R PREX R SHEAR R T Error C RT Vol I AD D Errore [Kn] [m/s] [m/s] [N] [N] [N] [%] *1000 [mm] [deg] [%]
20 10.289 2.090 0.849 3.254 16.953 6.745 23.698 -14.49 73.25 78.76 1.99 12.03 30 15.433 3.134 1.274 4.880 15.603 7.250 22.853 -3.38 31.41 82.03 4.26 9.92 40 20.578 4.179 1.699 6.507 15.100 5.922 21.022 1.37 16.25 72.13 4.65 11.47 60 30.867 6.269 2.548 9.761 11.419 11.042 22.461 1.67 7.72 47.00 2.91 -10.37
20 10.289 2.090 0.849 3.254 17.766 6.957 24.723 -10.79 76.42 82.10 1.87 5.43
Ship Trim Model
Table 5 - Series IV - Towing Tank Results
Table 6 - Series IV - CFD Results
Table 7- Series IV - Aft & Fwd body's resistance component
Table 8 - Series IV - Step Height -30% - CFD Results
Table 9 - Series IV - Step Height -30% - Aft & Fwd body's resistance component
V S FnV Lift Overall R PREX R SHEAR R T Overall Lift Overall R PREX R SHEAR R T Overall [Kn] [N] [%] [N] [N] [N] [%] [N] [%] [N] [N] [N] [%]
20 2.090 62.221 41.98 2.495 1.857 4.351 19.72 86.005 58.02 12.945 4.769 17.714 80.28 30 3.134 32.639 22.02 2.487 1.409 3.897 17.81 115.617 77.98 12.121 5.859 17.980 82.19 40 4.179 11.781 8.01 1.448 0.299 1.747 8.72 135.374 91.99 12.191 6.109 18.300 91.28
After-Body Forward-Body Ship
V S V S FnV FnL V M R O R A (15°C) R O R A (15°C) R T C RT Vol I AV I AD I AV I AD D [Kn] [m/s] [m/s] [Kg] [Kg] [N] [N] [N] *1000 [dm] [dm] [mm] [mm] [deg]
6 3.087 0.627 0.255 0.976 0.108 0.102 1.059 1.001 2.060 70.78 0.04 0.03 66 63 -0.11 8 4.116 0.836 0.340 1.301 0.315 0.175 3.090 1.717 4.807 92.94 0.11 0.03 73 63 -0.37 10 5.144 1.045 0.425 1.627 0.784 0.266 7.691 2.609 10.301 127.35 0.04 0.10 66 70 0.15 12 6.173 1.254 0.510 1.952 1.171 0.369 11.488 3.620 15.107 129.76 -0.18 0.26 44 86 1.55 14 7.202 1.463 0.596 2.278 1.414 0.481 13.871 4.719 18.590 117.24 -0.32 0.33 30 93 2.33 16 8.231 1.672 0.681 2.603 1.609 0.606 15.784 5.945 21.729 104.96 -0.39 0.32 23 92 2.55 18 9.260 1.881 0.766 2.928 1.775 0.725 17.413 7.112 24.525 93.62 -0.46 0.29 16 89 2.70 20 10.289 2.090 0.851 3.254 1.666 0.784 16.343 7.691 24.035 74.29 -0.55 0.28 7 88 2.99 22 11.318 2.299 0.936 3.579 1.640 0.800 16.088 7.848 23.936 61.16 -0.67 0.29 -5 89 3.47 24 12.347 2.507 1.021 3.904 1.575 0.830 15.451 8.142 23.593 50.66 -0.77 0.28 -15 88 3.80 26 13.376 2.717 1.106 4.230 1.507 0.893 14.784 8.760 23.544 43.06 -0.85 0.27 -23 87 4.06 28 14.404 2.925 1.191 4.555 1.418 0.912 13.911 8.947 22.857 36.05 -0.91 0.25 -29 85 4.21 30 15.433 3.134 1.276 4.880 1.340 0.928 13.145 9.104 22.249 30.58 -0.94 0.23 -32 83 4.24 32 16.462 3.344 1.361 5.206 1.277 0.939 12.527 9.212 21.739 26.25 -0.97 0.21 -34.5 81 4.26 34 17.491 3.552 1.446 5.531 1.224 0.946 12.007 9.280 21.288 22.77 -0.99 0.19 -37 79 4.28 36 18.520 3.761 1.531 5.856 1.186 0.951 11.635 9.329 20.964 20.01 -1.02 0.16 -39.5 76 4.26 38 19.549 3.970 1.617 6.182 1.163 0.957 11.409 9.388 20.797 17.81 -1.04 0.14 -41.5 74 4.26 40 20.578 4.179 1.702 6.507 1.122 0.970 11.007 9.516 20.523 15.86 -1.06 0.11 -43.5 71 4.22 42 21.607 4.389 1.787 6.833 1.091 0.993 10.703 9.741 20.444 14.33 -1.06 0.08 -44 68 4.13 44 22.636 4.597 1.872 7.158 1.055 1.025 10.350 10.055 20.405 13.03 -1.07 0.06 -44.5 65.5 4.06 46 23.664 4.806 1.957 7.483 1.013 1.067 9.938 10.467 20.405 11.93 -1.07 0.03 -45 63 3.99 48 24.693 5.015 2.042 7.809 0.961 1.121 9.427 10.997 20.424 10.96 -1.07 0.01 -45 61 3.91 50 25.722 5.224 2.127 8.134 0.896 1.191 8.790 11.684 20.473 10.13 -1.07 0.00 -45 60 3.88 52 26.751 5.433 2.212 8.459 0.814 1.278 7.985 12.537 20.523 9.39 -1.08 0.00 -46 60 3.91 54 27.780 5.642 2.297 8.785 0.728 1.372 7.142 13.459 20.601 8.74 -1.11 0.01 -49 61 4.06
Trim Ship Model
V S V S FnV FnL V M R PREX R SHEAR R T Errore C RT Vol I AD D Errore [Kn] [m/s] [m/s] [N] [N] [N] [%] *1000 [mm] [deg] [%]
20 10.289 2.090 0.851 3.254 16.323 6.303 22.626 -5.86 69.93 90.66 3.19 6.69 24 12.347 2.507 1.021 3.904 16.133 6.844 22.977 -2.61 49.34 91.23 4.13 8.65 30 15.433 3.134 1.276 4.880 15.436 5.753 21.190 -4.76 29.12 71.74 4.66 9.82 34 17.491 3.552 1.446 5.531 15.432 6.094 21.526 1.12 23.03 76.65 4.61 7.83 40 20.578 4.179 1.702 6.507 15.075 6.031 21.106 2.84 16.31 70.20 4.56 7.90 44 22.636 4.597 1.872 7.158 15.229 5.077 20.305 -0.49 12.97 69.26 4.74 16.80 50 25.722 5.224 2.127 8.134 14.762 5.674 20.435 -0.19 10.11 64.50 4.46 15.15
Ship Model Trim
V S FnV Lift Overall R PREX R SHEAR R T Overall Lift Overall R PREX R SHEAR R T Overall [Kn] [N] [%] [N] [N] [N] [%] [N] [%] [N] [N] [N] [%]
20 2.090 59.671 40.24 2.990 1.720 4.710 20.82 88.599 59.76 13.333 4.583 17.916 79.18 30 3.134 9.623 6.37 1.422 0.339 1.761 8.31 141.386 93.63 14.015 5.414 19.429 91.69 40 4.179 9.591 6.47 1.565 0.590 2.155 10.21 138.646 93.53 13.509 5.442 18.951 89.79
Forward-Body After-Body Ship
V S V S FnV FnL V M Lift Error R PREX R SHEAR R T Var on 0% C RT Vol Rise D Var on 0% [Kn] [m/s] [m/s] [N] [%] [N] [N] [N] [%] *1000 [mm] [deg] [%]
20 10.289 2.090 0.851 3.254 148.226 0.02 15.440 6.626 22.065 -2.48 68.20 0.39 2.77 -13.17 30 15.433 3.134 1.276 4.880 148.256 0.04 14.608 7.269 21.877 3.24 30.06 3.78 4.11 -11.85 40 20.578 4.179 1.702 6.507 147.155 -0.71 13.639 6.408 20.047 -5.02 15.50 4.15 4.07 -10.81
Trim Ship Model
Table 10 - Series IV - Step Height +30% - CFD Results
Table 11 - Series IV - Step Height -30% - Aft & Fwd body's resistance component
Graph 2 – Model-1 (without steps): comparison of the predicted total resistance by CFD results and towing tank measurements
V S FnV Lift Overall R PREX R SHEAR R T Overall Lift Overall R PREX R SHEAR R T Overall [Kn] [N] [%] [N] [N] [N] [%] [N] [%] [N] [N] [N] [%]
20 2.090 57.351 38.71 3.252 1.399 4.650 20.31 90.807 61.29 13.706 4.540 18.246 79.69 30 3.134 21.959 14.81 2.460 0.518 2.979 13.25 126.320 85.19 14.906 4.598 19.504 86.75 40 4.179 3.876 2.62 0.928 0.069 0.997 4.72 144.105 97.38 16.159 3.963 20.122 95.28
After-Body Forward-Body Ship
V S V S FnV FnL V M Lift Error R PREX R SHEAR R T Var su 0% C RT Vol Rise D Var su 0% [Kn] [m/s] [m/s] [N] [%] [N] [N] [N] [%] *1000 [mm] [deg] [%]
20 10.289 2.090 0.851 3.254 148.159 -0.03 16.958 5.938 22.896 1.19 70.77 -0.50 3.49 9.20 30 15.433 3.134 1.276 4.880 148.278 0.05 17.366 5.116 22.482 6.10 30.90 7.70 5.28 13.21 40 20.578 4.179 1.702 6.507 147.981 -0.15 17.087 4.032 21.119 0.06 16.32 8.70 5.50 20.64
Ship Trim Model
15
20
25
30
35
40
2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8
103
CRT
Fn
Without Step
With Step
Graph 3 – Model-1: comparison of the hull resistance in the conventional and modified with step versions (numerically predicted)
0
10
20
30
40
50
60
70
80
90
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5
103CRT
Fn
Towing Tank
CFD
Graph 4 - Model2, Series III: Total Resistance Coefficient Vs Volumetric Froude Number
0
10
20
30
40
50
60
70
80
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
103CRT
Fn
Towing Tank
CFD
Graph 5 – Model2, Series IV: Total Resistance Coefficient Vs Volumetric Froude Number
Graph 6 - Model2, Serie IV: Shear and pressure resistance components dependence on step height at different speeds, evaluated on
the after-body and forward body parts of the hull