A STUDY ON VENTILATED HYDROFOIL SECTIONS USING FREE STREAMLINE THEORY AND CFD ANALYSIS Thesis submitted in partial fulfillment of the requirements for the degree in BACHELOR OF TECHNOLOGY in OCEAN ENGINEERING AND NAVAL ARCHITECTURE by Aproop Dheeraj (07 NA 1017) Under the guidance of Prof. Om Prakash Sha DEPARTMENT OF OCEAN ENGINEERING & NAVAL ARCHITECTURE INDIAN INSTITUTE OF TECHNOLOGY, KHARAGPUR May, 2011
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A STUDY ON VENTILATED HYDROFOIL SECTIONS USING FREE
STREAMLINE THEORY AND CFD ANALYSIS
Thesis submitted in partial fulfillment of the requirements for the degree in
BACHELOR OF TECHNOLOGY
in
OCEAN ENGINEERING AND NAVAL ARCHITECTURE
by
Aproop Dheeraj
(07 NA 1017)
Under the guidance of
Prof. Om Prakash Sha
DEPARTMENT OF OCEAN ENGINEERING & NAVAL ARCHITECTURE
INDIAN INSTITUTE OF TECHNOLOGY, KHARAGPUR
May, 2011
Department of Ocean Engineering &
Naval Architecture
Indian Institute of Technology
Kharagpur-721302
CERTIFICATE
This is to certify that the thesis entitled “A Study on Ventilated Hydrofoil Sections using Free
Streamline Theory and CFD Analysis” submitted by Aproop Dheeraj (Roll no. 07NA1017), in the
partial fulfillment for the requirement for the award of degree of Bachelor of technology in
Ocean Engineering & Naval Architecture by the Indian Institute of Technology, Kharagpur during
the academic session 2010-2011 is a bona fide record of the project work carried out by him
under my supervision and guidance. The approval does not necessarily endorse or accept every
statement made. Opinion expressed or conclusion drawn as recorded in the thesis. It only
signifies the acceptance of the thesis for the purpose for which it is submitted.
Date: ____________
Place: IIT Kharagpur Dr. O.P.Sha
Professor
Department of Ocean Engineering & Naval Architecture
Indian Institute of Technology
Kharagpur-721302
ACKNOWLEDGEMENT
I would like to extend my sincere gratitude to Prof. O P Sha, my supervisor for the project, for
his support, patience and constant involvement in the project. I would also like to thank Mr.
Manu Korulla and Mr. Jagadeesh Kadiyam of NSTL, Vizag, for providing me with the requisite
data from experiments and their invaluable inputs at critical stages in the project that led to
significant progress.
Date: ____________
(Aproop Dheeraj)
CONTENTS
1. Introduction
2. Free Streamline Theory Equations
3. Ventilated foil sections – experiment and CFD analysis
4. Extension of the theoretical model to a NACA-66 section
5. Correlation of theory with experimental and CFD results
6. Conclusions
List of Symbols The symbols used in this thesis are generally explained when first introduced and their meaning will
normally be apparent from the context. The following list contains the main symbols used.
CL Lift coefficient
CD Drag coefficient
CM Moment coefficient
Cp Coefficient of pressure
Fr Froude number
g Acceleration due to gravity
pv Vapour Pressure
p Ambient Pressure
α Angle of attack
ν Kinematic viscosity
ρ Density
σ Cavitation number
1. Introduction
Ever since sea-faring began, man has made continuous efforts to conquer the seas and establish his supremacy
over the mighty oceans. In this, one of the prime areas of focus has been to make waterborne craft faster. Ships
and boats have evolved from wind powered and paddle-oar systems to steam engines and turbines, nuclear
powered vessels and reciprocating diesel engines. But all of these focus on the propulsion of the vessel, to
overcome the resistance that is encountered while moving through water. Figure 1.1 depicts a typical drag (skin
friction)-velocity curve :
Frictional drag as a percentage of total resistance vs speed
The contributory factors to the total drag are the skin friction and wave resistance, amongst other losses (eddies,
vortex shedding etc). Depending on the type of vessel and the operating speed, different resistance components
have different degrees of contributions. The definition of “High Speed” in the case of waterborne vessels is
ambiguous, as it depends not on the absolute velocity, but the velocity in comparison to the size of the vessel. This
leads to the definition of the dimensionless “Froude number” (Fn):
Where, ‘V’ is the velocity of the craft, ‘L’ its length and ‘g’ the acceleration due to gravity. In cases like SWATH hulls
where the definition of length is not clear, an alternative definition of the Froude number is used:
Where, is the displacement of the vessel. For high speed craft, the operation range is typically at Efforts to
reduce drag have resulted in modification of the hull form, and vary from planing craft and multihulls to hydrofoil
ships and hovercraft. Again, the idea behind most of these concepts to either lift the hull out of the water, thereby
reducing wetted surface area and thus the skin friction component of resistance, or the wave making component.
Favorable wave resistance characteristics are also observed simultaneously in some cases, like hydrofoil ships.
High speed craft typically operate at , and they fully or partially employ hydrodynamic lift. Apart from
reducing frictional losses, they also contribute to stabilizing the ship’s motion in waves. Hydrofoils may also be
used in conjunction with other designs, like multihulls and air cushion vehicles (hovercraft). Generation of lift is
basically a consequence of conservation of Energy from Bernoulli’s theorem, where the change in local velocities
creates a pressure difference, and thus a lifting force. It is however most effective at higher speeds and the lift
varies in proportion to the velocity squared. So two phases are observed in the operation of hydrofoil craft, where
at lower speeds there is not much lift, and a “take off” speed at which the hull lifts out of the water and
subsequently is a foil – operation mode. Prior to take off, as the speed increases, the trim also increases and the
center of gravity gradually rises. The maximum “hump drag” occurs at just around the take off speed, and on
further increase a speed it reduces due to reduction in wetted surface area, along with reduction in the trim angle.
However, at higher speeds, the skin friction increases again at higher speeds relatively and causes a rise in
resistance. The effective power however, continuously increases with velocity.
Foils may be used in various configurations, depending on the distribution of lift over the forward and aft foil
systems, the submergence of the foil system, the type of foil section used and the way the craft is stabilized.
Hydrofoils may be surface piercing or fully submerged. Surface piercing hydrofoils have the advantage of lift
control by means of lifting area variation so that these types of foils are inherently stable and do not necessarily
require additional lift control mechanisms for safe operations. Fully, but shallowly submerged hydrofoils obtain
some lift control through the free surface effect which reduces lift as the hydrofoil approaches the water surface,
but this is not a very reliable design approach for lift control as this effect is observed in a small range of immersion
depths. These types of hydrofoils are used in inland waterways only, as the foils are vulnerable to broaching in
waves. Fully, relatively deeply submerged hydrofoils do need a lift control system since their lifting area does not
vary with submergence. Trailing edge flaps linked to a digital ride control system are nowadays commonly used for
lift control, not only for operation in waves but also to assist take-off and to enable maneuvering. Due to the lower
wave excitation and high effectiveness of digital ride control systems coupled to trailing edge flaps, the sea keeping
behavior of deeply submerged hydrofoils is superior to that of surface piercing hydrofoils.
Virtually all craft today employ sub cavitating sections, which have an upper speed limit of 40 to 45 knots. Higher
spends require vented and super cavitation sections and have been employed in experimental and prototype craft.
Cavitation
Cavitation is a phase change observed in high speed flows wherein the local absolute pressure reaches the vicinity
of the vapor pressure at the ambient temperature. It is signified by what is known as the cavitation number,
defined as
Physical cavities usually have finite length and positive cavitation number. Some mathematical formulations have
been suggested previously to investigate the possibility for obtaining the solution of steady cavity flows of ideal
fluid satisfying the following conditions:
(i) The cavitation number is greater than zero
(ii) The cavity pressure is uniform throughout the cavity
(iii) The pressure of the fluid is nowhere less than the cavity pressure
(iv) The pressure is continuous across the cavity boundary (no pressure jump)
Cavitation occurs in three stages, according to the degree of development of the cavity:
(i) Initial cavitation, which is characterized by the noise of collapsing bubbles. It is harmful as bubble
implosion can cause cracking and shock in moving parts like propellers and rudders and damage them,
hampering performance.
(ii) Partial cavitation, in which the cavity sustains over a part of the cavitating body and has a pulsating nature
(iii) Fully developed cavitation, in which the cavity dimensions exceed that of the body. It is observed for high
velocities (≥ 70 m/s). It can also be observed for considerably small velocities at penetration of bodies
through the free water surface or at horizontal motion in the conditions of crossing the free boundary. In
this case the supercavity is filled by atmospheric air and is related to artificial or ventilated supercavities
forming at the lower speeds.
2. Free Streamline Theory Equations
Many models have been developed on potential flow theory, where the cavity is treated as a single bubble with
constant internal pressure. We bring our attention to three main models, all based around the classical Helmoholtz
boundary theory, which says that the cavity is of infinite length.
(i) Riabouchinksy model: An ‘image’ plate is introduced downstream of the plate to model the cavity. The
coefficient of drag is given by
CD(σ)={1+σ+[8(π+4) ]-1σ 2}CD(0) ≈(1+σ)CD(0)
Where CD(0)=0.88 is the classical Helmholtz solution for σ=0
The free streamlines run from A to A’ and the flow is assumed to be irrotational everywhere.
(ii) Re-entrant jet model: The free streamlines reverse direction at the rear of the cavity to form a jet which
flows upstream in the cavity and it supposed to not impinge on the real body but is removed
mathematical by allowing it to flow on a second Riemann sheet in the physical plane to infinity. The
point infinity acts as a doublet plus source, while the jet represents a sink.
(iii) Roshko’s dissipation model: In this model, the flow past a bluff body is divided into two parts. Near the
body it is described by the free streamline theory to allow a possible adjusment of the under
pressure. The flow farther downstream is described by an equivalent potential flow so that its
pressure increases constinuously to the free stream value as it approaches infinity. The actual
dissipative wake flow is thus represented.
Roshko’s model is simpler in many respects and gives more accurate results, although it involves more complex
manipulations. All theories agree and give the same value for the coefficient of drag in case of pure drag problems
(given by the equation previously).
The main assumptions in employing the Roshko’s model are:
(i) Sharp leading and trailing edges of the foil
(ii) Separation of the streamlines occurs at the leading and trailing edges
(iii) The foil thickness is assumed to be zero (since it does not affect the flow once the cavity is fully
developed)
(iv) Totally dry (ventilated) upper surface of the foil
Using some conformal mappings and complex potential flow equations, we obtain for general profiles (as
illustrated in the diagram) , the coeffiecients of drag, lift and moment (for small cavitation numbers, typically <
1.5) as:
Radius of curvature definitions
1) Ai
2) Bii
2 3/ 2
1
2
2
(1 ) 1)
A
A
A
iii Radius of curvature at A RKd y
dx
2 3/ 2
2
2
2
(1 ) 1)
B
B
B
iv Radius of curvature at B RKd y
dx
22
21 2 12(1 )[sin ]
6 6 2D
AC
J
22
1 22(1 )
6 6LC
J
2
1{ sin cos cos2
AA
231 1 1
2 34
1
cos 5(sin ) sin (sin )( )
4 (sin / 2) 2 4 2 4}
A A AA A
A
0
2
2
4(1 ) ( cos sin )
3M L DC K C C
J
1 2 3
8 84 sin ( sin ) cos sin
3 2 15J A A A
For small angles: 24 sin( )J .
22 21 12 1 cos sin
2 2 4 2 2 2B AK K
22 2
1
9 1 12 1 cos sin8 2 2 4 4 2 2
B AA K K
2 2
2
1cos sin
2 2 2B AA K K
22 2
3
1 1 12 1 cos sin8 2 2 4 4 2 2
B AA K K
The above result shows that ( - ), 1A , 2A and 3A are all of 2( , )O for all ; and in particular, ( -
), 2A and 3A reduce to 2 4( , )O ) for close to /2. Moreover, the fact that 3A is much smaller
.than 1A indicates a good accuracy of the expansion given by Eq. (3.13). If the above quantities are used to check
the curvature and slope of the solid boundary at some other points, say at / 2ie one can easily find that the
agreement is within a factor at most of 2( , )O . It should also be remarked here that if more terms were taken
in the expansion then the first three coefficients 1A , 2A , 3A would differ slightly from the above value. However,
it can be verified that the “improvement” of the solution by taking terms more than three is actually
unappreciable.
For segmental sections, the above equations will simplify to:
2 2212
(1 )[sin ]6 6 2
D
AC
J
2 2
21
2(1 ){[sin cos cos ]
6 6 2L
AC A
J
231 1 1
2 34
1
cos 5[(sin ) sin (sin )( )]}
4 (sin / 2) 2 4 2 4
A A AA A
A
0
2
2
4(1 ) ( cos sin )
3M L DC K C C
J
Where,2
2 2
2 2 1/ 2
1{ cos sin }
2 4( ) 4 sin [( ) ] 2 2
2 2 1/ 2( ) ] ( )
2
1 2 2 1/ 2
sin 9 (1 cos ) 9{1 } {1 }
8(4 sin ) 16 9[( ) ] 32( )A
2 2
2 2 2 1/ 2
1
{ cos sin }4 ( sin ) [( ) ] 2 2
AA
3 1 2 2 1/ 2
1 2 (1 cos )( ) { sin }
9 9(4 sin ) [( ) ]A A
Definition of curvature for circular arc
3. Ventilated Foil Sections
As discussed earlier, the value of the lift generated varies as the square of the velocity. Post take-off, with
increasing velocity, the hull rises higher and higher out of the water, and the foils move closer to the surface, and
consequently some of the pressure gradient is lost, causing a reduction in the lift. The craft then goes to a lower
draft where the hull’s position relative to the water surface can be maintained, and the process is repeated in a
‘bobbing’ kind of motion. It is thus desirable to control the lift beyond a certain velocity so that there is better ride
control. Usually, this is done either by varying the angle of attack of the foil by a control mechanism, or to provide
a trailing edge flap that effectively produces a camber in the foil section. Such methods are usually very
complicated in terms of instrumentation and expensive. For the purpose of lift dumping, and reduction in drag, an
alternative method has been proposed. Air is injected over the surface of the foil to create a cavity on the suction
side. This does away with moving parts designed for flap systems and changing the angle of attack, at the same
time reducing drag and providing ride control. The mechanism has been illustrated below
Lift dumping through ventilation of foils
Towing tank experiments have been performed at the Naval Science and Technological Laboratory (NSTL),
Vizag for NACA 66 airfoil sections at velocities between 3-6.5 ms-1 and small angles of attack (up to 8
degrees). The position of the slit for outlet of air was tried on the top surface, bottom surface and the
nose of the foil. However, it must be noted that the results so obtained are 2 D data, i.e. the foil is
assumed to have an infinite span (and consequently there is no bleeding of pressure or end effects).
Additionally, strut effects have to be considered, for which data was taken over 75% and 100% of the foil
and multiplied by factors to obtain drag and lift coefficients. Another set of data was obtained from wind
tunnel testing of the foil, which takes into consideration only the effects of geometry of the foil (Abott
curve). Experimental results were compared with those obtained by CFD analysis in Fluent, with modeling
of the slit at various positions on the foil. Consistency was found for 3.33 ms-1, but difficulties arose in
both experiment and CFD results of higher velocities. The corresponding results have been plotted in the
below:
Coefficient of lift (experiment) vs. angle of attack for different slit positions
Coefficient of drag (experiment) vs. angle of attack for different slit positions
Coefficient of lift (CFD) vs. angle of attack for different slit positions
Coefficient of drag (CFD) vs. angle of attack for different slit positions
4. Extension of the theoretical model to a NACA-66 section
In order to extend the free streamline theory equations developed by Wu to a general airfoil section, such as the
NACA-66 (used widely in Naval industries), it is necessary to approximate it either to a symmetric (circular arc) or
asymmetric general section, to calculate the values of A1,A2,A3 and Beta, and thus the lift and drag coefficients.
The first approach that was followed was to use the co-ordinates of the foil underside to calculate the first and
second derivatives and substitute and find the values of phi1 and phi2.
Lower surface of a NACA-66 section
First derivatives obtained from coordinates
Second derivatives obtained from coordinates
It can be observed that the nature of variation of the slopes and curvatures by this method is rather
erratic and thus seems to be an unreliable approach. Next, a trend line was fitted through the curve to
obtain a polynomial equation of the surface:
Curve fitting through coordinates
By this method, the slopes and curvatures were calculated at the leading and trailing edges and were
obtained as:
Φ1=.346
Φ2=.064
Ka= -0.00337597
Kb= 0.000676
Substituting these values into the general case equations, the lift and drag curves were obtained, but
again it was observed for non zero cavitation numbers that for small angles in the range of 0-5 degrees
there were sudden spikes in the values of lift (as illustrated), which were unacceptable. It was observed
that the large variation to the order of thousands came due to changes in beta to the order of 0.1, and
beta was directly dependent on phi1. This method was also thus discarded.
Lift coefficient trend obtained by using co-ordinates/curve fitting
Since it was apparent that it was desirable to have small leading and trailing edge angles, a segmental
approximation for the foil surface seemed to be a feasible option. Arcs were fitted through the mid chord
and maximum chord ordinates in AutoCAD. The value of the included angle was 27 degrees and 30
degrees respectively for each case.
Segmental approximation using AutoCAD
The value of the curvature was then substituted into the special case equations and the lift and drag
curves were obtained. There wasn’t any major difference between the values of the maximum chord and
the mid chord approximations. The curves showed a good trend at small as well as large angles of attack.
The cavitation number was increased up to 1.5, around which the lift curves started to show a decreasing
trend which is logically incorrect, validating Wu’s assumption of small cavitating numbers, corresponding
to larger sustained cavities and opposed to partial cavitation.
Coefficient of lift vs. angle of attack for maximum chord segmental approximation
Coefficient of drag vs. angle of attack for maximum chord segmental approximation
Comparison of the lift coefficient values for different approximations
Comparison of the drag coefficient values for different approximations
Lift coefficient -segmental approximation for large angles of attack
Drag coefficient – segmental approximation for large angles of attack
It needs to be pointed out that the values obtained for the lift and drag coefficients using the segmental
approximation lie in the same range as those obtained for the actual section taken in the general case,
barring the problematic region lying between angles of attack of 0 and 5 degrees. Even in this region,
however, the curve obtained matched closely with the extrapolated values for the actual section, i.e. it
followed the trend that should have been expected, which validates the segmental approximation thus
made and any errors that it could have entailed.
5. Correlation of theory with experimental and CFD results
In order to compare the experimental and theoretical results, first a relationship needs to be established between
the velocity and the cavitation number, or, in the case of ventilated hydrofoils, the ventilation number.
We have,
If we substitute the value of Pv, the vapor pressure with the cavity air pressure, the following values are obtained
(for 3.33 ms-1) under different ventilating conditions: