Modelling and CFD Analysis of Traditional Snake Boats of Kerala · 2019. 8. 13. · 482 B. Venkata...

Available online at www.sciencedirect.com

ScienceDirectScienceDirect

Aquatic Procedia 4 ( 2015 ) 481 – 491

2214-241X © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).Peer-review under responsibility of organizing committee of ICWRCOE 2015doi: 10.1016/j.aqpro.2015.02.063

INTERNATIONAL CONFERENCE ON WATER RESOURCES, COASTAL AND OCEAN ENGINEERING (ICWRCOE 2015)

Modelling and CFD Analysis of Traditional Snake Boats of Kerala B.Venkata Subbaiaha*, Santosh.G.Thampia, V.Mustafaa

a Department of Civil Engineering, National Institute of Technology, Calicut, Kerala, India.Pin:673601

Abstract

Kerala has a rich maritime tradition dating back to several centuries. Snake boats, locally called Chundan vallams, are one of the icons of Kerala. Different types of snake boats take part in Vallamkali, the local name for traditional boat races. This paper presents the preliminary results of a study on a typical ChundanVallam. Data were collected at site using a total station and three dimensional drawings of the boat, required for performing the analysis, were developed using MAXSURF-16 and Rhinoceros-6.0. Hydrodynamic analysis was performed using the commercial CFD software package SHIPFLOW 5.1® Analysis was performed by employing coarse, medium and fine meshes. Validation of the model was performed by performing simulations with hull series 60 (S_60). The drag coefficient at different Froude numbers for steady turbulent flow was estimated by performing simulations with the data of the snake boat. Velocity vectors as well as contours of pressure distribution, obtained from the simulations, are also presented. © 2015 The Authors. Published by Elsevier B.V. Peer-review under responsibility of organizing committee of ICWRCOE 2015.

Keywords:Chundan Vallam; Snake Boat; CFD; Hydrodynamic Analysis; Drag Coefficient

1. Introduction

The solution of problems involving determination of resistances using Computational Fluid Dynamics (CFD) analysis is now becoming tractable due to enhanced accessibility to high performance computing. Determination of the resistance characteristics of ships/ vessels is one of the most important topics in Naval Architecture, Offshore

* Corresponding author. Tel.: +91-9440142409;

e-mail address:[email protected]

© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).Peer-review under responsibility of organizing committee of ICWRCOE 2015

http://crossmark.crossref.org/dialog/?doi=10.1016/j.aqpro.2015.02.063&domain=pdf




482 B. Venkata Subbaiah et al. / Aquatic Procedia 4 ( 2015 ) 481 – 491

and Ocean Engineering. Today, several CFD tools play an important role in the design of the ship hull forms. CFD has been used for the analysis of ship resistances, seakeeping, manoeuvering and investigating the variation in resistance encountered due to changes in the ship hull resulting from variation in its parameters. However, in ship hydrodynamics, obtaining accurate results is next to impossible (Aktar et al., 2013). Studying the pattern of waves generated by a ship moving through water is one of the most important objectives in ship hydrodynamics due to its importance in the design process. The waves produced by a ship in motion can radiate at great distances away from the ship. Furthermore, these waves contain energy that must be dissipated in the surrounding fluid. The ship experiences a force that tends to oppose its movement, one of its components being known as the wave making resistance which is the most important component of the resistance encountered by a moving ship (Ahmed, 2011). Nomenclature

B Breadth at mid ship S Wetted surface area T Draft V Velocity P Instantaneous pressure

BC Block coefficient FC Coefficient of frictional resistance TC Coefficient of total resistance WC Coefficient of wave making resistance nF Froude number nR Reynolds number PPL Length between perpendiculars FR Frictional resistance TR Total resistance WR Wave making resistance

g Acceleration due to gravity k Turbulent kinetic energy p Time averaged pressure t Time z Depth of water

General variable ix Cartesian coordinates iU Instantaneous velocity components in cartesian directions

iu Time averaged velocity components in cartesian directions "iu Fluctuating velocity components in cartesian directions ”p Fluctuating pressure iR Volume force ij Total stress tensor Mass density Dynamic viscosity Kinematic viscosity Specific dissipation of turbulent kinetic energy

B Parameter direction crossing the boundary A.P. Aft body perpendicular F.P. Fore body perpendicular

Chundan vallams (beaked boats) of Kerala, known to the outside world as the snake boats, are one of the icons of

Kerala culture and is used in Vallamkali (boat race). Similar type of boat races are conducted in some other parts of


the world including the Ottawa Dragon Boat Race Festival, Canada, Sydney Dragon Blades, Australia, Dragon Boat Race Festival, Hong Kong and Dragon Boat Race, Germany. The snake boats of Kerala are constructed according to specifications in the Sthapathya Veda, an ancient treatise on the building of wooden boats and are a marvelous example of ancient India's prowess in Naval Architecture. These boats vary from 100 to 138 ft in length. with the rear portion towering to a height of about 20 ft., and a long tapering front portion; it resembles a snake with its hood raised. Its hull is built of planks precisely 83 feet in length and six inches wide. Traditionally each boat belongs to a village and the villagers worship it like a deity. Only men are allowed to touch the boat and to show respect they should be barefooted. To make the boat more slippery while in the water and thereby to reduce the resistance of the hull, it is oiled with a mixture of fish oil, coconut shell carbon and eggs. Repair work of the snake boat is generally done annually by the village carpenter. The major objective of this work is to collect data of a typical snake boat to enable its characterization and thereafter to use this data to perform CFD analysis of its hull form to understand its resistance characteristics. It is extremely important that data on these traditional boats are documented as otherwise it is very likely that this indigenous knowledge base may be lost forever. Also, it is hoped that this study may yield some exciting results which could be useful in the development of new hull forms of vessels which are highly efficient. This is of considerable practical significance when viewed in the background of the ongoing efforts to revive the inland navigation system in Kerala. The application of CFD tools for predicting the flow pattern around the hulls of vessels has made much progress over the last decade.

2. Basic Governing Equations

The equations are formulated based on the reference coordinate system presented in Fig. 1.

Fig.1 Coordinate system

2.1 Reynolds-Averaged Navier-Stokes (RANS) Equations

The Reynolds averaged Navier Stokes equations are time averaged equations of motion for fluid flow. In Reynolds decomposition, an instantaneous quantity is decomposed into its time averaged quantity and fluctuating quantity. These equations are primarily used with approximations based on knowledge of the properties of turbulence in flow to obtain approximate time averaged solutions of the Navier-Stokes equations.

According to the continuity equation, 01

i

i

xU

t (2.1)

For steady incompressible flow, the continuity equation can be written as 0i

i

xU

(2.2)

The Navier-Stokes equations of motion can be written asj

iji

j

iji

xR

xUU

tU )(

(2.3)


ij is the total stress, which for a Newtonian fluid can be written as )31(2 ijkkijijij SSP

(2.4)

ijS is the strain-rate defined as i

j

j

iij x

UxUS

21

(2.5)

For incompressible flow, k kS in the above equation is zero. 021

k

k

k

k

k

kkk x

UxU

xUS (2.6)

The Reynolds-averaged Navier-Stokes equations can be derived from equation (2.3) by splitting the instantaneous velocity components iU into a mean iu and fluctuating ''iu components as

'''' iiiii uuuUU (2.7)

and the instantaneous pressure P into the time mean pressure p , and time fluctuating pressure ''p as

"" pppPP (2.8)

The time mean of a variable is defined as T

TT

dtT21lim

(2.9)

The time averaged continuity equation and Navier-Stokes equations for incompressible flow can be written as

0i

i

xu

(2.10)

i

j

j

i

jii

j

ijiji

xu

xu

xxR

x

uuuu

tu 1

'''' where, (2.11)

2.2. Components of Ship Resistance

In a resistance test, it should be ensured that there is geometric similarity between the full scale ship and the model. Being geometrically similar means having the same shape. In addition the model and the prototype shall also satisfy conditions of kinematic and dynamic similarity. Kinematic similarity implies that the streamlines around the hull are geometrically similar; dynamic similarity means that the force vectors are scaled by the same factor and will have the same direction at model and full scale. If kinematic and dynamic similarity is to be achieved between the model and the prototype, all non-dimensional parameters which characterize the flow and the forces would have to be equal. The main non-dimensional parameters in the flow around the hull are the Froude number )( nF and the

Reynolds number )( nR (Eliasson et al., 2011) expressed as

PPn gL

VF (2.12)

VLRn (2.13)

The resistance of a ship at a given speed is the force required to tow the ship at that speed in smooth water, assuming no interference from the towing ship. Neglecting eddy making, appendage and air resistances, the equation for total resistance can be written as

FWT RRR (2.14)


where, WW CSVR 25.0 (2.15) FF CSVR 25.0 (2.16) FWT CCC (2.17) Typically, the friction drag coefficient is predicted using the ITTC’57 ship model correlation line

SVC

C fF 25.0

(2.18)

210 )2(log

075.0

nf R

C (2.19)

3. Computational method

3.1. Computational method for wave making drag and frictional drag

For computing the wave making drag, the computational domain was divided into two numerical grids - one on the ship hull surface and the other on the water surface using the XPAN module in SHIPFLOW 5.1®. Flow lines, wave profile along the length of ship and wave making resistance were to be calculated. Analysis was performed for coarse, medium and fine meshes and different combinations of draft and Froude numbers. To solve the governing equations in the computation of viscous drag, the domain was subdivided into a finite number of cells. Analysis was carried out using the XCHAP module in SHIPFLOW 5.1®. It uses several turbulence models. The solver can be used in a zonal or a global approach and can handle overlapping grids. Flow was computed with a double model or with a prescribed free-surface for both coarse mesh and very coarse mesh. The total drag, and pressure and velocity contours on the body surface were determined.

3.2. Collection of data and modelling

Data required to develop CAD drawings of the Chundan Vallam proposed for analysis were collected using a total station (Fig. 2). The feasibility of employing 3D scanners for this purpose was examined and this plan was abandoned because of difficulty in positioning the scanners. The Chundan Vallam was at Aranmula in Pathanamthitta District of Kerala, India. Coordinates i.e. North, East, and Elevation (X,Y,Z) of points were collected at each offset with respect to a reference point (benchmark). Using MS-EXCEL, these data points were converted about to F.P. reference point and a wire frame model with edge surf was developed using AUTOCAD as shown in Fig.3; later the *.dwg files were converted to *.dxf files and imported into MAXSURF-Pro (Fig.4). The NURUB surfaces were generated using the data and *.dxf files with markers (Fig. 5) and these were exported as *.3dm files. These files were thereafter imported through RHINOCEROS and converted to IGES files. Fig. 6 presents the offset file generated in SHIPFLOW and this was used in the analysis.

Fig. 2 Data collection using with total station


Fig.3 AUTOCAD wireframe model with edge surfing Fig.4 Development of NURUB surface in MAXSURF-Pro

Fig.5 NURUB surface model Fig.6 Offset file in SHIPFLOW5.1

4. Computational domain and boundary conditions

The models used in this study are a typical snake boat and Hull Series 60 (S_60), a standard ship for ship- hydrodynamics research used by the ITTC. The model of a typical snake boat was prepared from the data collected from the field and this was analysed using XPAN, a flow solver for the potential flow around three dimensional bodies. The flow lines, wave profile along the length of the ship and the coefficient of wave making resistance

wC were computed. Using the XCHAP module, the grid for performing viscous flow computations was generated. This is a finite volume code that solves the RANS equations and is used for turbulence modelling. The velocity and pressure fields on the surface of the hull and the coefficient of frictional resistance ( FC ) were computed. The dimensions of the prototype employed in this study and its hydrostatic particulars are tabulated in Table 1. Fig.7 presents the coarse mesh around the snake boat with 1050504 cells and 939330 nodal points, the size of the fluid domain being 1 PPL from sF.P., 1.5 PPL from A.P. and of radius of 4 PPL . Analyses were performed for the boundary conditions presented in Table 2. Fig. 8 presents a 3D view of the Kelvin wave pattern at nF = 0.24 and 0.5m draft. Analyses was performed for drafts 0.5m and 0.6m, the Froude number ranging from 0.24 to 0.34.

Table1. Prototype dimensions

Dimensions Snake Boat Hull Series 60

Length between perpendiculars PPL (m) 19.695 121.92

Breadth ( B ) (m) 2.32 16.26

Draft (T ) (m) 0.7871 6.502

Block Coefficient ( BC ) 0.38 0.6

Surface area ( S ) (m2) 30.528 2526.4


Table 2. Boundary conditions

Fig.7 Computational fluid domain Fig.8 Kelvin wave pattern

5. Results and discussion

Fig. 9 presents the Kelvin wave pattern (wave contours) and wave profiles at Froude numbers 0.24, 0.280, and 0.30. The variation in the wave pattern as the wave propagates away from the hull can be observed. The resolution of the free surface reduces significantly and the wave cuts along the length of the boat increase with speed. The wave height increases significantly from bow to stern; the profile of the snake boat also increases from mid-point to stern, and hence there is no problem with this. The wave profile near the bow starts with a trough at Froude number 0.30 and it starts with crests as the Froude number increases. The computed values of resistance coefficients of snake boat are tabulated (Table 3).

No slip Slip Inflow Outflow

u 0iu 0,0B

iii

unu iu constant 0

B

iu

p

0B

p 0

B

p 0

B

p

0p

k

0k

0B

k

k constant 0

B

k

,...)(uf 0 constant B


Fig. 9. Kelvin wave pattern and wave profile at different Froude numbers for 0.5m draft

Fig. 10 is a plot of freeboard versus Froude number for two different drafts. As evident from Fig. 10(b), for a draft of 0.6m, the freeboard values attain negative values beyond a certain Froude number. It means that the boat is under water which is unacceptable. Hence the design draft for the boat needs to be limited to 0.5m and this condition shall be strictly adhered to for Froude numbers in the range of 0.24 to 0.34.

Table 3. Computed values of resistance coefficients for a typical snake boat

Fig.10 Froude number vs freeboard at (a) draft = 0.5m and (b) draft = 0.6m

nF wC FC TC

0.24 0.00215 0.00230 0.00444

0.26 0.00213 0.00223 0.00435

0.28 0.00210 0.00217 0.00428

0.3 0.00209 0.00213 0.00421

0.32 0.00207 0.00209 0.00416

0.34 0.00205 0.00206 0.00411

-0.075

-0.025

0.025

0.075

0.125

0.175

0.225

0.22 0.27 0.32

Fre

eboa

rd le

vel i

n m

eter

s

Froude Number

FreeboardNet freeboard

0.02

0.07

0.12

0.17

0.22

0.27

0.32

0.22 0.27 0.32

Fre

eboa

rd le

vel i

n m

eter

s

Froude Number

Freeboard

Net freeboard


Figs. 11 and 12 show the plots of pressure and velocity coefficients respectively along the length of the boat. The non-dimensionalized pressure is extracted through the expression, p=1/2Cp (measured in Pascals) and the non-dimensionalized velocity is extracted by multiplying with ship speed, offers useful information regarding pressure and velocity; these are useful for improving the ship lines. A Cp value of zero indicates that the pressure is the same as the free stream pressure, a Cp value of 1 indicates the pressure is stagnation pressure and the point is a stagnation point and a Cp value of minus one (-1) indicates a perfect location for a "Total energy". By plotting the isowakes (contours of constant axial velocity) or the iso-vorticity contours, the effect of the stern geometry on the propeller can be studied. The pressure contour diagram may be used for hull optimization Hull series 60 model was used for validation of the analysis using SHIPFLOW and the corresponding offset, wave profile and wave pattern are presented in Figs. 13(a), (b), and (c). and the color scale referring zero is still water level, to extract wave height these values are multiplied by LPP. Fig.14 presents the wave making, frictional and total drag on the snake boat considered in this study at a design draft 0.5m for different Froude numbers. It is observed that the maximum wave making drag is 178.2kN, the maximum frictional drag is 358.3kN and the total drag is 536.6kN. Fig.15 is a plot showing the variation of the coefficient of wave making resistance for hull series 60; the values of the coefficient of wave making resistance were compared with published results (Salina Aktar,2010) and are in good argument with each other.

Fig. 11 Pressure coefficient contours on the surface of the body of the snake boat

Fig. 12 Velocity coefficient contours on the surface of the body


Fig.13(a) Offset file of hull series 60 (S_60)

Fig.13 (b) Wave profile along the length of hull series 60 (S_60) & Fig.13(c) Kelvin wave pattern of S_60

Fig.14 Froude number vs drag force @ 0.5m draft Fig. 15. Validation of SHIPFLOW results

6. Conclusions

In this work, hydrodynamic analysis of a typical Chundan Vallam was performed using the commercial CFD software package SHIPFLOW 5.1® Analysis was performed by employing coarse, medium and fine meshes. Validation of the model was performed by performing simulations with hull series 60 (S_60). The drag coefficient at different Froude numbers for steady turbulent flow was estimated by performing simulations with the data of the snake boat. Velocity vectors as well as contours of pressure distribution are presented.

Acknowledgment

We are thankful to the Indian Maritime University (IMU), Visakapatnam, Andhra Pradesh, India, for the permission granted to use the MAXSURF-16, Rhinoceros-6.0 and SHIPFLOW-5.1® software for performing the analysis. The help rendered by Mr. V. F. Saji and Dr. Manu Korulla, Scientists at NSTL, Visakapatnam is also gratefully acknowledged.

0100200300400500600

0.23 0.28 0.33 0.38

Dra

g F

orce

(kN

)

Froude Number

Wave making dragFrictional dragTotal drag

0

0.2

0.4

0.6

0.8

1

0.15 0.25 0.35 0.45 0.55

Coe

ffic

ient

of w

ave

mak

ing

resi

stan

ce (

Cw

)

Froude Number

SHIPFLOW

Published


References

Aiguo Shia., Ming Wu., Bo Yang., Xiao Wang., Zuochao Wang., 2012. Resistance Calculation and Motions Simulation for Free Surface Ship Based on CFD. Journal of Procedia Engineering 30, 68-74.

Bhattacharyya. R., 1978 Dynamics of Marine Vehicles, Wiley, New York, 28-39. Edward. V. L. Ed., 1988. Principles of Naval Architecture: Vol II - Resistance, Propulsion., The Society of Naval Architects & Marine Engineers,

25-32 Feng Zhao., Song-Ping Zhu., Zhi-RongZhang., 2005.Numerical Experiments of a Benchmark Hull Based on a Turbulent Free-surface Flow

Model. Tech science Press.CMES 9. 273-285. George Tzabiras., Konstantions Kontogiannis., 2010. An Integrated Method for Predicting the Hydrodynamic Resistance of Low-CB Ships.

Journal of Computer-Aided Design 42, 985-1000. I. Senocak., G Iaccarino., 2005.Progress towards RANS Simulation of Free-Surface Flow around Modern Ships. Center for Turbulence Research

Annual Research Briefs. 151-156. J. M. A. Fonfach., C.GuedesSoares., 2010. Improving the Resistance of a Series 60 Vessel with a CFD Code. Proceedings of the V European

Conference on Computational Fluid Dynamics, June-2010. 1-14. Md ShahjadaTarafder.,GaziMd Khalil., 2004. Numerical Analysis of Free Surface Flow around a Ship in Deep Water. Indian Journal of

Engineering & Material Sciences 11, 385-390. Salina Aktar., Goutam Kumar Saha., Md. Abdul Alim., 2010. Drag Analysis of Different Ship Models using Computational Fluid Dynamics

Tools. Proceeding of International Conference on Marine Technology, MARTEC, Dec-2010. 123-129. Salina Aktar., Goutam Kumar Saha., Md.Abdul Alim., 2013. Numerical Computation of Wave Resistance Around Wigely Hull using

Computational Fluid Dynamics Tools. Journal of Advanced Shipping and Ocean Engineering 2, 45-50. SHIPFLOW Theoretical Manual XCHAP., 2007. Flowtech International AB. 1-20. Sofia Eliasson., Danil Olssion., 2011. Barge Stern Optimization ,Analysis on a Straight Shaped Stern using CFD. A Thesis for the Degree of

Master of Science. Chalmers University of Technology, Gothenburg, Sweden. Report No. X-11/266, 21-22. Volker.B., 2000.Practical Ship Hydrodynamics. Butterworth-Heinemann. Y.M. Ahmed., 2011. Numerical Simulation for the Free Surface Flow around a Complex Ship Hull Form at Different Froude Numbers.

Alexandria Engineering Journal 50, 229-235. Zhang Zhi-rong., Liu Hui., Shu Song-ping., ZHAO Feng., 2006. Application of CFD in Ship Engineering Design Practice and Ship

Hydrodynamics. Conference of Global Chinese Scholars on Hydrodynamics. 315-322.

Modelling and CFD Analysis of Traditional Snake Boats of Kerala · 2019. 8. 13. · 482 B. Venkata...

Documents

Transcript of Modelling and CFD Analysis of Traditional Snake Boats of Kerala · 2019. 8. 13. · 482 B. Venkata...