nternational Journal of
D e l f t U n i v e r s i t y o f T e c h n o l o g y
Ship Hydromechanics laboratory
L i b r a r y
iv!el<elweg 2 26282 CD Delft
Phone: +31 (0)15 2786873
Naval Architecture
and Ocean Engineering
V o l . 4 N o . 4
D e c e m b e r 2 0 1 2
Naval Architecture and Ocean Engineering
Volume 4, Number 4
December 2012
Contents
335 Hydrodynamic design of an underwater liull cleaning robot and
its evaluation
Man Hyung Lee, Yu Dark Park, Hyung Gyu Park Won Chul Park
Sinpyo Hong, Kil Soo Lee and Ho Hwan Chun
353 The implementation of the integrated design process in the
hole-plan system
Won-Sun Ruy Dae-Eun Ko and Young-Soon Yang
362 Technological and economie study of ship recycling in Egypt
Yousri fvl. A. Welaya, Maged M. Abdel Naby and Mina Y. Tadros
374 Psycho-acoustic evaluation of the indoor noise in cabins of a
na-val vessel using a back-propagation neural network algorithm
Hyung-Suk Han
386 Suggestion of a design load equation for ice-ship impacts
Yun-Hyuk Choi, Hye-Yeon Choi, Chi-Seung Lee,
Myung-Hyun Kim and Jae-Myung Lee
403 The effect of dynamic operating conditions on nano-particle
emissions from a light-duty diesel engine applicable to prime
and auxiliary machines on marine vessels
Hyung mln Lee and Yeon hwan Jeong
All
On the energy economics of air lubrication drag reduction
Simo A. Makiharju, Marc Perl in and Steven L. Ceccio
Indexed a n d A b s t r a c t e d in Science C i t a t i o n Index E x p a n d e d , J o u r n a l C i t a t i o n R e p o r t s / Science E d i t i o n , a n d C u r r e n t C o n t e n t s ® / E n g i n e e r i n g , C o m p u t i n g , a n d Technology
International Journal of
Naval Architecture and Ocean Engineering
Volume 4, Number 4
December 2012
Contents
423 Ultimate strength of simply supported plate with opening under
uniaxial compression
Chang-Li Yu and Joo-Sung Lee
437 An array effect of wave energy farm buoys
hiyucl<-rvlin Kweon and Jung-Lyul Lee
447 Hydrodynamic characteristics for flow around wavy wings with
different wave lengths
IVli Jeong Kim, if yun Sii< Yoon, Jae Hwan Jung,
Ho Hwan Chun and Dong Woo Pari<
460 Thermo-economic approach for absorption air condition onboard
high-speed crafts
Ibrahim S. Seddiei<, iVIosaad Iviosleh and Adei A. Banawan
All
Endplate effect on aerodynamic characteristics of
three-dimen-sional wings in close free surface proximity
Jae Hwan Jung, Mi Jeong Kim, Hyun Sii< Yoon, Pham Anh Hung,
Ho Hwan Chun and Dong Woo Pari<
488 Experimental investigation on stern-boat deployment system and
operability for Korean coast guard ship
Ho Hwan Chun, Moon Chan Kim, Inwon Lee, Kooichyun Kim,
Jung Kwan Lee and Kwang Hyo Jung
Hydrodynamic design of an underwater hull
cleaning robot and its eyaluation
M a n H y u n g L e e ' , Y u D a r k P a r k ' , H y u n g G y u P a r k ' , W o n C h u l Park^ S i n p y o H o n g ^ , K i l Soo L e e ' a n d H o H w a n C h u n ' '
'Graduate School of Mechanical Engineering, Pusan Nadonal Utnversity, Busan, Korea
'Graduate School of Interdisciplinary Program in Mechatronics, Pusan National Utdversity, Busan, Korea ^Global Core Research Center for Ships and Offshore Plants, Pusan Nadonal University, Busan, Korea
''Department of Naval Architecture and Ocean Engineering, Pusan National University, Busan, Korea
A B S T R A C T : An undenvater hull cleaning robot can be a desirable choice for the cleaning oflaige ships. It can make
the cleaning process safe and ecotwmical. Tins paper presents a hydrodynamic design of an underwater cleaning robot and its evaluadon for an unden vater ship hull cleaning robot. The liydrodynamic design process of the robot body is described in detad. Optimal body design process with compromises among conflicting design requirements is given. Experimental results on the hydrodynamic performance of the robot are given.
KEY WORDS: Under water h u l l cleaning robot; Remotely operated vehicle; Hydrodynamic design.
I N T R O D U C T I O N
Hulls o f large ships i n service need to be periodically cleaned for safety inspections and reduction o f drag force. A ship that operates w i t h a clean under-water surface can potentially save over 5% o f hiel cost. H u l l cleaning also restores effectiveness o f antifouling paint arrd extends the life span o f the paint on the hull. Since expensive non-toxic chemicals are cunently used for the paintmg o f large ship hulls, the uuportance o f hull cleanmg has been growing. However, cleaning o f large ships is usually very expensive and mefficient. Cleaning and inspection o f ship hulls are usually executed by human divers. There are opera-tioiral limitations in the cleanmg thue length and working space for huiuan underwater operations. Thus, an under-water huU cleaning robot (HCR) can be an effective means to reduce economical burden in cleaning and pamting o f ship hulls.
Compared w i t h cleaning ship hulls i n dry docks or on the gr'ouird (lona et al., 2009), underwater hull cleaning can be economical. The preparation and postprocessing for dry dock cleaning or gi-ound cleanmg are quite expensive and tkne-con-suming. Under-water cleaning also mcreases ship availability.
Under-water HCRs are special types o f undei-water robots. Underwater robots can be either connected to the surface w i t h tether cables or isolated without tether. There are several types o f undei-water robots including remotely operated vehicles (ROVs), autonomous underwater vehicles ( A U V s ) , and solar-powered autonomous undewater vehicles (SAUVs). Theh application covers exploration o f undewater resources, fishhig mdusti-ies, port security, and military operations. The tethered undewater robots are usually caUed remotely operated vehicles (ROVs). ROVs can take pictijies undewater and cany out diverse undewater jobs. The fust undewater robot, Pooddle, was developed by a Frenchman D i m i t i i Rebikoff hi 1953 (Marme Technology Society, 2012). In its initial stage o f development, United States Navy stiidied R O V to solve technical problems
C o i T e s p o n d i n g a u t h o r : Man Hyung Lee e - m a i l :
mahlee@pusan.ac.la-336 Inter JNav Archit Oc Engng (2012) 4:335-352
associated w i t l i unmanned undei-water vehicles (Roberts and Sutton, 2006). The design o f an underwater hull cleaning robot was inttoduced by Yuan et a l , 2004. hr ICorea, R O V research has also been done. (Jung et a l , 2009; Jun et a l , 2006; L i et a l , 2004; Jung et a l , 2002; K i m and Shin, 2005).
Even though robots for undei-water operations have been actively developed throughout the world, a commercial A U V for undei-water hull cleanmg has not been introduced. The Office o f Naval Research (ONR) o f the United States Navy is cunently developmg an autonomous undei-water hull cleanmg robot H u l l B U G with SeaRobotics Corporation.
The development o f undei-water hull cleanmg robots requii-es specialized techirologies. The robots are supposed to w o r k near the ship hull without damaging the hull pamt. Theh motions need to have six degrees-of-freedom (DOF) with sub-decmreter accuracy. They demand special propulsion systems to apply wide range o f worldng conditions. They need to cover 2000 nfAi for mspectional operations and 200 ;»-'//? for cleamng jobs. The H C R to be considered in this paper is designed to approach the ship from a mother ship or the gi'ound by autonomous navigation, h i this paper, 3 m/s has been selected as the practical maximum approaclrirrg speed.
While there are many issues m the design o f huU cleamirg R O V , such as the positions o f cleanmg biush and metacenter, auproof, keepmg o f robot's attittide on the inclined huh suiface, and anticonosion (Yuan et a l , 2004), important key issues with the development o f autonomous undei-water HCRs can be related with drag reduction and the design o f t h e positionhig system. The mam focus i n the design o f t h e body is the reduction o f body drag to extend the operation period o f t h e cleaning seivice. Accurate positioning o f H C R requires expensive sensor systems, such as long base line ( L B L ) acoustic poshioning system and high gi-ade inertial sensors.
In this paper, the design o f the robot body starts from a smrple basic model. A comparative sttidy on the fluid dynaiuic behavior is performed with the basic model and several optional models. From this stiidy, an optimal shape o f the body that satisfies the design requhements is detenumed. Analysis o f the flow field around the body is made with computational fiuid dynamic (CFD) techniques using tiirbulence models, standard and R N G ic-e models. The CFD analysis is perfoimed with the standard ic-s model. To mvestigate the influence o f hirbulence model on the analysis, an additional C F D analysis w i t h the R N G ic-s model on the optuual shape o f the body is executed. Analysis results o f the two tirrbulence models are compared to evaluate the effect o f the tiarbulence model.
B O D Y D E S I G N O F A H U L L C L E A N I N G R O B O T
The design o f t h e H C R body involves appropriate comproiuises between coirflicting design goals: mobility and stability, mamtenance convenience and stabihty, and perfonuance and budget. Autonomous navigation o f a long path requhes stabihty to overcome large attittide changes induced by external forces such as waves. Thus it is desirable to have a long metacenter, the distance between the centers o f gravity and buoyancy. On the other hand, for the robot to clean the sides o f t h e ship hulls, h may be necessary to roll up to 90 degi'ees and keep its attittide for an extended time, h i this case, a short metacenter is advantageous for the large roll motion. I f cleanmg equipments are installed on the top side o f the robot body, h is able to clean the bottom o f the huh w i t h oirly a small change o f attittide. The maintenance o f cleanmg equipments can also be gi'eatly simplified. However, the center o f gi'avity o f the body moves downward and stability decreases.
The sttategy taken m the design o f H C R body is to luhrmrize as much drag reduction as possible. To meet this sttategy, a basic model o f H C R is designed at f u s f Then drag forces o f the body are computed w i t h various fiow speeds. The components that have relatively di'ags are to be found and then shapes are changed so that the drags on the components are decreased.
General structure of R O V
ROVs can be largely classified as open frame stiuctiire or closed frame stiuctiire. A n R O V w i t h an open frame stiuctiire has stable three DOFs motions based on a large metacenter. This type o f stmctiire has various additional advantages. This stiuctiire is well k n o w n and generally adopted by most ROVs. h is convenient f o r large payloads. hispection and cleanmg equipments can be easily attached to or removed from the body. However, ROVs with this stiucmre have difficuhies w i t h motions that require more than thr'ee DOFs. Due to these advantages and disadvantages, open frame stractiare is mainly applied to ROVs for general worlcs.
Even though an R O V with a closed fi'ame stiuctirre has potential for greater mobility, it has disadvantages. Due to a short metacenter, it has relatively unstable motions, tt also has a small payload. It is veiy inconvenient in handling large payloads for
Fig. 1 R O V s w i t l i open frame stmcture.
Fig. 2 ROVs w i t h closed frame structure.
cleaning and mspecting. Thus ROVs o f closed frame shucUire are applied to special works that need fast motions.
There are several requhements f o r HCR. Four DOFs for surge, yaw, heave, and roh are necessary for stable navigation and cleaning works. Attached cleaning bmshes can be easily exchanged. Body stability is neghgibly affected by the height o f t h e cleanuig devices. Fmally, mamtenance o f the body and payload is corrvenient. Thus, an open frame stiuctiire is adopted for the H C R in this stiidy.
In the design o f t h e general arrangement o f a R O V , compromise between operational convenience and design complexity is required. The cleanmg equipments can be attached to the front side, upper side, left or right sides o f t h e body o f HCR. I f the equipments are placed on the upper side, H C R is able to clean the bottom o f huU w i t h almost the same attitude as when h navigates, ft only needs to roll w i t h a considerable angle to clean the sides o f the hull. A n R O V o f this stiructiire is shown m Fig. 3. W i t h this type o f stmcttire, tire center o f gravity tends to move upward making the R O V more unstable. To mamtam the cen-ter o f buoyancy above the cencen-ter o f gravity, an additional payload may be placed at the bottom. The cleanmg equipment is placed on the upper side o f H C R in this study. Even though the size and weight o f R O V can be large, the operational conve-nience is considered much more important.
Horizontal thmst is used to obtain force to gam horizontal motion. The H C R is requhed to have horizontal speed o f 3 knots. Prelhuuraiy estuuation o f t h e horizontal speed is made with a simplified body model. The estimated m m i m u m horizontal speed is 3.01 hwts w i t h the thiust force o f 50 ligf and drag coeflicient o f 1.1.
Body
Camera and Light
5/
Front watch sonar M o o n Pooi t h m s t Cleaning brush U5BL
*-:
^ P ' C e n t e r of buoyancy < | | t F - # ' T ' f = M Housing for c o n t r o l and C o m m u n i c a t i o n gravity Cover and - Buoyancy f o a m Horizontal propeller ^Body frame — Power housing Weight Manipulator \ M a n i p u l a t o r housing Manipulator Manipulator controller housing338 Inter JNav Archit Oc Engng (2012) 4:335-352
Veitical thrust is necessaiy for heave motion. The force is also used to adhere to the suiface o f the ship hull to be cleaned or to change the roh attihide o f HCR. The m m i m u m requhements on the H C R for this shidy relating to veitical thiust are veitical speed o f 1.7 Imots and roll angle o f 90 degrees. The H C R has tluee vertical thrusts. Each thmst has the maxhnum output o f 25
kgf. The three thmsts f o n n a triangular shape and the cleaning equipment is placed at the center o f t h e hiangle. The maximum
roll moment o f 18.5 kgf-m can be produced with the thmsts. W i t h this anangement, the vertical thiusts can produce a moment to overcome restoiing moment b y gravity force when the roll attihade o f t h e body is 90 degrees.
The prelhninary estimation relatmg to the vertical thmst verifies that the design requhements can be satisfied with the selected thiusts. The vertical drag force is estimated to be less than 75 kgf which is the maxhnum available vertical thiust force. When the roU attitiide o f t h e body is 90 degrees, the restormg moment produced by the gi'avity is 12.09 kgf-m which is less than the maxmium roll moment o f the thrusts.
Design of thrusters
Horizontal thiusters are used for surge and yaw motions o f HCR. I n order to design the horizontal tlrmsters to satisfy the requhement o f t h e 3 hmts o f t h e maximum surge speed, a rough estimation o f drag force is calculated usmg the f o l l o w m g relation:
F=-p-C,-AA^' (1)
2 '
where F is the drag force, P is the density o f the fluid, is the drag coefficient, A is the frontal area, and V is the velocity o f moving object. Assuming that the body shape is a box o f rectangular parallelepiped fi-ontal area o f 0.28 nr, the velocity o f 3.4 hmts is obtamed i n the seawater with the horizontal tluust force o f 50 kgf and drag force coefficient o f 1.1. I f a cable drag force o f 9.91 kgf is added to the body drag force, the speed o f 3.01 hwts is obtained. Thus, the requhement on the maximum surge speed can be satisfied with 50 /cgf thmst force.
Vertical thiusters are used when H C R moves vertically or changes hs roh attitiade for cleanmg operations. To clean the sides o f t h e ship hulls, 90 degi-ees o f t h e maximum roll angle is necessaiy. W i t h thi'ee thiusters, each has 25 kgf thmst force, the H C R can have the maximum vertical speed o f 1.7 hwts. When H C R adheres to the bottom o f the hull for cleanmg, it can produce the maximum adhesion force o f 75 kgf. T w o o f t h e tluee vertical thi'usters are located as close as possible to the sides o f t h e H C R so that it can produce a large roU moment. It is designed to produce the maxhnum roll moment o f 18.5 kgf-m that exceeds the maximum expected restoruig moment 12.09 kgf-m when the roll angle is 90 degrees.
Analysis method
The body o f the H C R is designed to minimize the drag force. The computation o f the drag force is assisted w i t h com-putational fiuid dynamics (CFD) techniques. Turbulent fiow field around the H C R is analyzed with the general CFD analysis code F L U E N T / U N S . The govemmg equations employed in the F L U E N T / U N S are the three-dimensional equations for mass and momentiim conseivations w i t h the standard k-B ttirbulence model. The equation o f mass conservation or contmuity equation in tensor forms is
dp , ^{P"i) dt av,
. 0 (2)
where P is the specific mass, " y is tire velocity component along the coordinate axis -v ,. , and t is the time. Assumuig m-compressible fiow, the equation for the momentiam conseivation is
d{pu,) d[pu,u^)_ dp ^ a r ,
where P is the pressure, g , is the gravity component along the coordinate axis x. , F. is the extemal force component along the axis x. , and T^IJ is the sti'ess tensor defined as
^,7 = P dn. du. —' - + —- ~ ^dx. dx, ^ 2 du (4)
where P is the viscosity and S,., the Ki-onecker delta.
Since the f l o w field around f f C R can have turbulence under severe flow conditions, the selection o f an appropriate turbulence model may be necessaiy. f n order to analysis the turbulence around a large scale model such as f f C R , the standard
lc-£ two equation Uirbulence model that is generally used m industries is considered in this paper. The li-s turbulence
model is derived fi-om the Reynolds Averaged Navier Stokes (RANS) equations.
Ö ( / ? « , ) d{pu,u.) dt dx, _d_ dx, ( ( p
du, du. \ 2 du.
^dxj^ dx, , 3 ^ dr.. •' -P
—
S,i dx, dx,(5)
where u,u. is the Reynolds sti'ess due to the random turbulent fiuctuations i n the fiuid momenUim. Using the Boussmesq eddy viscosity assumption, the Reynolds stress can be written as
-pu,Uj = p, du, dUj
2'
Pk + P,— du,dx,
(6)
where p, is the turbulent viscosity, p, can be computed fl-om the turbulent kineric energy lc and turbulent dissipation rate f a s follows:
P,=PC, (7)
where C^, is the model constant. The turbulent Idnetic energy and turbulent dissipation rate can be obtained fl'om the equations:
d{pk) ^ d{pu,k) ^ d [[^^^ p, \ dk ^ dt dx, dx, VV ^ t ) dx, J + +G,-p£ (8) d[p£) d(^pu,£) dt _d_ dx P + VV de dx: (9) + C , - ( G , + ( l - C 3 j G , ) - C , , . / , ^
where Gf. and G^ are terms due to the turbulent stress and buoyancy, respectively. They can be obtamed fl-om the equations as follow:
340 Inter J Nav Archit Oc Engng (2012) 4:335-352
— du.
Pr, dx,
(10)
(11)
where T is the teinperature, Pr,, the turbulent Prandtl number, and , the coefficient o f thermal expansion. Model constants i n equations (8), (9), and (11) are given as follows:
C,. = 1.44, C , _ = 1.92, C^,= 0.09, ^^^^ cr, = 1 . 0, 0 - ^ = 1 . 3 , Pr, = 0 . 8 5
These model constants are obtamed from experiments on the basic tijrbulent shear flows o f ah and water. To take into consideration o f t h e f l o w near walls in the standard / f- f model, tiirbulence intensity and hydrauhc diameter are used i n the tiirbulent specification method. Numerical computations are performed w i t h standard waU fiinction. Analysis algorithm used m F L U E N T / U N S is the Semi-Imphch Method for Pressure-Lmked Equation (SDVIPLE) algorithm o f tiie finite volume method.
Computafion of drag force
Since tiie actiial shape o f t h e H C D body is complicated to calculate the drag force, fiow field analysis can be easily done b y using a simple model o f t h e body. Simple model o f the components that have relatively large drag force such as camera, hori-zontal and vertical thrusts, and roUer on the upper side are chosen for easy C F D analysis. Turbulent fiow field around the sim-plified model o f the selected components are analyzed with F L U E N T .
h i order to detennine the best shape design o f body, the drag force computation is made m two steps. Fhst, a simple basic model is selected and drag force on it is computed. Based on the drag force computation, several options f o r the shape o f t h e body that may have smah drag force are proposed. A f t e r the evaluation o f the drag force for each option is made, the smallest drag force among the options is chosen as the best option.
Drag force computafion on the basic model
The shape o f a general R O V is shown m Fig. 4. As shown in the figure, the stiuctiires o f the inside and lower part o f t h e body are quite complicated. The thin frames that connect the upper and lower stiuctiires may have relatively small drag forces.
However, they requhe large time for the generahon o f mesh around them f o r CFD analysis and may cause uncertainties on the CFD analysis results due to their undesii-able effects on the overall mesh quality. Therefore, simple fomis o f t h e upper and low-er shuctures that have major effect on the drag force are dlow-erived for the basic model o f HCR. Fig. 5 shows the basic model.
Fig. 5 The shape o f basic model.
The mesh for the f l o w analysis with the basic model is generated as follow. A n unsUuchjred hiangle mesh is generated by mnning Gambh v.2.3.16. The mesh size is 0.5 near the H C R body. Tlie size increases with the distance from the body. To deci-de the proper number o f mesh elements, the tests o f the deci-dependeci-dency on the element number are conducted. 150 miUion mesh elements are chosen after testmg 80 million, 150 million, and 200 million mesh elements. The details o f t h e test results was omitted in consideration o f paper's length.
The computation region and boundary conditions are roughly shown in Fig. 6. Let the length o f H C R i n the flow dhection be L . Then the length fi-om the inlet to the fi-ont o f t h e body is 4 Z , the length from the rear o f t h e body to the outiet is 8Z., and both the height and width o f the computation region are AL. The inlet condition is determined b y the fiow speed. Outlet conditions are given on the convection and vertical and horizontal planes o f symmetiy. Viscosity conditions are given on the surface o f the HCR.
The drag force is computed m the steps o f 0.5 laiots fiow speeds hi the range o f 0-5 laiots. Drag force can be written as
Fr=F+F„ (13)
where i v is the total drag force, , the pressure or f o n n resistance, and F,, , the viscous resistance. Thus, consideiiirg equ-ation (1), the drag coefficients can be written as
2F^ pU^A 2F^ pU^A 2F^ pU'A (14)
342 Inter J Nav Archit Oc Engng (2012) 4:335-352
where U is the inlet f l o w velocity, P is the f l u i d density, and A is the frontal area. As shown hr Table 1, the drag force mputation results show that pressure resistance is dominant and the effect o f viscous resistence is quite weak. Total drag force 9.3 Af w i t h the f l o w speed o f 0.5 Knots. The drag mcreases i n a parabolic shape as the f l o w speed increases. The maximum di-ag force is 911 A at the maxmium speed o f 5 Knots. The drag coefficient shows similar values for the speed range w i t h the variation less than 1 % as shown i n Fig. 7. I n order to derive optimum body shape, the pressure resistance o f each component o f H C R body is computed to conclude which component has the major influence on the drag.
co is
Table 1 Drag forces and coefficient w i t h fiow speeds. V e l o c i t y (Knot) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 FAN) ^.24956 33.13114 74.70490 132.99836 208.02904 299.81985 408.35072 533.65909 675.74219 834.60425 FAN) .07466 3.82893 .09883 13.81142 20.91993 29.39068 39.19628 50.31608 62.73170 76.42770 FAN) 9.32422 36.96008 82.80373 146.80978 228.94897 329.21052 447.54699 583.97516 738.47388 911.03194 0.59164 0.59403 0.59520 0.59615 0.59678 0.59729 0.59768 0.59802 0.59831 0.59856 CDV 0.07707 0.06865 0.06545 0.06191 0.06001 0.05855 0.05737 0.05639 0.05554 0.05481 Cn 0.66872 0.66268 0.65984 0.65806 0.65679 0.65584 0.65505 0.65440 0.65385 0.65337 0.8 O.G 0.4 0.2 1 2 3 4 S|)ecd ( K j i o t ) Fig. 7 Total drag coefficient.
Surface pressure distribufion and flow pattern for the basic model
In order to find the shape o f H C R body that has low pressure resistance, pressure dish-ibution on the suiface o f t h e basic body model is analyzed. The components that have relatively large positive or negative pressure are found. Fig. 8 shows the pressure dishibution on the body surface w i t h the maximum fiow speed 3 Knots fi'om various view angles, h shows the stagnation pomts at the fi-ont o f parts 1, 3, and 4 have relatively high pressures, h i the case o f housing, part 4, the shape o f t h e
front parts are hemispheres and tire area with high pressure is relatively small. This shape has a desirable pressure dishibution such that the pressure has the luaxiruum value at the center and decreases as the distance from the center mcreases.
Fig. 9 shows f l o w pattems around the basic model. The cross sections in tlie figure are defined m Fig. 8. Cross section 1 is the ttansverse section that cuts tluough the center o f the model. Cross sections 2 and 3 are the ttansverse sections that cut thr-ough the center o f housing and a vertical tliruster, respectively. Cross section 4 is the longitiidinal section to watch the wake distiibution at the front part o f the horizontal thrusters.
The flow pattems around upper and lower stiuctiires are very shnilar to those around blunt objects. A stagnation point is generated at the front part o f each shuchire. The flow accelerates after the stagnation pohit until the point o f mflection. There is no large vortex around the fi-ont part o f the model. However, it can be seen that there are relatively large vortices around the rear part o f t h e model. There ai'e also small vortices m the wakes o f cleaning bmshes on the upper side and vertical tluusters on the lower side o f t h e model. The areas which generate vortices have relatively large magiutiide o f negative pressures.
Fig. 8 Pressure distribution on the basic Fig. 9 Flow patteru around the basic H C R body model surface. H C R body model.
Design ofthe fuial model with small resistance
W i t h small modifications on the basic model, a new model that has small drag w i l l be designed. The arrangement o f equip-ments is unchanged and the shape o f each component that has large resistance w i l l be changed. The previous analysis results on the pressure disttibution and flow pattern show that parts 1, 3, and 4 have relatively large resistance. To reduce resistance, each shape o f t h e parts is changed to have streamlined form.
346 Inter J Nav Archil Oc Engng (2012) 4:335-352
Based on the shape design and anangement o f body detaü, the design is made. I f the proposed shape or anangement cannot be satisfied during the detail design, tire shape or anangement is modified and tire detail design is processed agam. Material sel-ection, shuchiral stability analysis, machuring methods, detail dimension such as thiclmess and length, and thennal analysis are made for the detail design. Fig. 11 shows the fmal H C R taken from various view angles.
E X P E R I M E N T S O N T H E H Y D R O D Y N A M I C P E R F O R M A N C E O F H C R
More tests may be necessaiy to verity the peiformance o f HCR. However, tests that are dhectiy related to the hydrodynamic performance o f H C R such as drag force and speed are presented m this paper.
Experhnents on the body drag force
In order to examine hie drag force o f the H C R body, tests are conducted i n a towmg tanlc. H C R is towed by a towmg carriage at speeds from 0 to 4 Knols and drag force is measured by a resistance dynamometer. The shape o f t h e shut to connect the H C R body with resistance dynamometer is sh-eainlined to reduce drag force on the submerged part. During the tests, floats and weights are added inside the H C R body to change the centers o f gi'avity and buoyancy so that the ttansverse and longitti-dinal ttuns are minimized. The required magnittide o f t h e liorizontal tlmist force to m n w h h the target maximum speed can be detennined from the drag force measurements.
The tests for drag force measurement are conducted with seven discrete speeds from 0 to 2.1 m/s at a regular mterval, 0.3
m/s. Three measurements are obtamed for each speed. The drag coefficients are obtained w i t h the drag measurements using tire
equation (1) where the frontal area o f t h e H C R is 0.670 nr. The resistance dynamometer used i n the drag tests can measure up to 3,000 N. Its accuracy is 0.2 % o f t h e frill scale. The dynamometer is calibrated with a standard dynamometer whose maxi-m u maxi-m force range is 5,000 N and accuracy 0.03 % o f tlie frill scale. Fig. 12 and 13 show the resistance dynamaxi-momaxi-meter and side views o f H C R mounted on the carriage, respectively.
F i g . 12 Resistance dynamometer.
Table 4 Drag measurement results.
1st Trial
V e l o c i t y (tn/s) Total resistance Fx (N) Strut resistance Fs (N) H C R resistance Fd (AO Drag coefficient Cd
0.0 0.0000 0.0000 0.0000 0.0000 0.3 26.0935 1.1121 24.9815 0.8302 0.6 99.4686 2.6913 96.7773 0.8041 0.9 225.9598 5.2725 220.6872 0.8149 1.2 400.1130 5.1324 394.9806 0.8204 1.5 509.4221 6.6479 502.7742 0.6684 1.8 742.3665 9.7201 732.6464 0.6764 2.1 889.0749 14.0249 875.0500 0.5935 2nd T r i a l
V e l o c i t y (tn/s) Total resistance Fj (N) Strut resistance Fs (N) H C R resistance Fq (AO Drag coefficient Cd
0.0 0.0000 0.0000 0.0000 0.0000 0.3 27.5458 1.1121 26.4338 0.8785 0.6 111.0022 2.6913 108.3109 0.8999 0.9 234.8187 5.2725 229.5461 0.8476 1.2 368.8709 5.1324 363.7385 0.7555 1.5 501.4406 6.6479 494.7927 0.6578 1.8 651.3177 9.7201 641.5976 0.5923 2.1 906.0459 14.0249 892.0210 0.6050 3rd T r i a l
V e l o c i t y (m/s) Total resistance Fx (iV) Strut resistance Fs (N) H C R resistance Fd (AO Drag coeffrcient Cd
0.0 0.0000 0.0000 0.0000 0.0000 0.3 28.9755 1.1121 27.8634 0.9260 0.6 113.8630 2.6913 111.1716 0.9237 0.9 233.0049 5.2725 227.7323 0.8409 1.2 336.7526 5.1324 331.6202 0.6888 1.5 450.7082 6.6479 444.0603 0.5903 1.8 822.0404 9.7201 812.3203 0.7499 2.1 951.5357 14.0249 937.5107 0.6359
348 Mer J Nav Archit Oc Engng (2012) 4:335-352
Table 4 and Figs, 14 and 15 show H C R drag measurement resuhs. The relationship between the drag force and velocity can be described with the second order polynomial
156.LY- + 1 I 2 . 0 . V - 9 . 5 0 5 (15)
where X is the velocity m m/s and ) ' is the drag force in N. The measured average drag coefficients are dishibuted between 0.60 and 0.83. For each velocity, the measured drag forces show small differences. These differences can be due to the h i m changes i n the tests. The average values o f drag force and drag coefficient measurements are larger than those obtained fi-om CFD. The reason for these discrepancies can be considered to be the result o f t h e sehips for the experiments and CFD are quite different from each other.
1000 800 5-600 Q 400 200 • Trial : A 2"' Trial * 3"* Trial 0 Average < CFD y = 156.1x^+112.0x+9.505/^ c , 4 . . . '/ A \ : X '• '/ A \ " y X A 1 1 0.3 0.6 0.9 1.2 1.5 I.S Speed (m/s) F i g . 14 Drag measurements. 2.1 2.4 0.8 O.S 0.4 0,2 0 T Trial Trial 3'^ Trial Average CFD 0 0,3 0.6 0.9 12 1.5 1.8 2.1 2.4 Speed (m/s)
Fig. 15 Drag coefficient.
Experhnents on the fonvard velocity
The tests on the horizontal speed o f H C R are conducted with five levels o f speed. The f o m a r d speed o f H C R is measured w i t h two methods. The speed is determined by an electromagnefic fiow meter or by the time o f tiavel for a predetermined distance. The fiow meter is attached to the bottom side o f the H C R body.
During the tests, H C R navigates fi-eely near the surface o f the water with neuttal buoyancy. The accuracy o f the electtomagnetic fiow meter used m the experiments is 2 % o f fiill scale or i^'-OOS m/s. Fig, 16 shows the electtomagnetic flow meter used in the tests. The flow meter attached on the bottom side o f H C R is shown m Fig, 17,
Fig. 17 Electromagnetic f l o w meter installed on H C R .
The speed commands for the velocity test have discrete values fi-om 0 to f u l l range at the regular mterval, one fifth o f the f u l l range. The test consists o f three sets o f expeiiments. During each set o f experiments, H C R velochies are measured for the velocity commands mcreasing fi-om 0 to the f u h range. Average velocity for a given velocity command is derived from the elapsed thne to ttavel 10 m distance. The mstantaneous maxhnum and minimum velochies obtamed fi-om the fiow meter is used to detennine measurement deviations. The measurement deviation is defmed as the half o f the difference between the maximum and minmiunr values.
Table 5 and Fig. 18 show the velocity test results. It shows the maximum velocity is 1.016 ni/s for the maxhnum velocity command. The relation between velocity measurement sand cormnands can be described with the linear equation
j = 1 . 0 9 x - 0 . 0 9 4 (16)
where X is the velocity command and ) ' is the velocity measurement in m/s. Duiing velocity tests, H C R moves along a relatively sti-aight luie m the f o w a r d dhection. Thus, it can be considered that the dhection o f the f o w a r d thmst force is appropriately adjusted and the weight distribution o f HCR body is well balanced.
Table 5 Forward speed w i t h main thrusters.
V e l o c i t y Commands V e l o c i t y (m/s) Averaged V e l o c i t y V e l o c i t y Commands F ' T r i a l 2"" T r i a l 3''' T r i a l Averaged V e l o c i t y V e l o c i t y Commands
Mean Deviation Mean Deviation Mean Deviation Mean Deviation
0/5 0.000 0,000 0,000 0.000 0,000 0,000 0.000 0,000 1/5 0,013 0.003 0,013 0,003 0.013 0.003 0,013 0,003 2/5 0.310 0,060 0,385 0,115 0,375 0.105 0.357 0.093 3/5 0,450 0.050 0.551 0,065 0.512 0.075 0.505 0.063 4/5 0,725 0,175 0,846 0.200 0.871 0,250 0.814 0,208 5/5 0.990 0.040 1.002 0,250 1,055 0,165 1.016 0.152
350 Inter J Nav Archit Oc Engng (2012) 4:335-352 1,2 1 •O 0.6 0) a. CO 0,4 0,2 I t Trial • A 2"'Trial • 3"' Trial : 0 Awrage . y = 1.090x-0.094 „ _ ^ : . . . . , . , 1 , , . , 1 ( 1 1 0.2 0.4 0.6 0.8
Input signal (part cf 5)
Fig. 18 Forward speed w i t h main thmsters.
Experiments on the vertical thrust
The test for the vertical thmst force measurement is conducted m the towing tank. A commercial weighing scale w i t h 300
leg capacity and 0.1 leg accuracy is used to measure the tluust fDrce. The scale hangs down from a ceiling hoist and the H C R
hangs down from the scale. Discrete input commands from 0 to the fiiU range at the resolution, one eighth o f t h e fiiU range, are sent to the vertical thmsters. Due to the hysteresis symptom m the scale, only one measurement is taken fr)r each command mput. For a given command input, the mean value o f vertical thmst force is defmed as the half o f the sum o f t h e maxhuum and m m i m u m values o f the weighting scale measui-ements. The deviation o f thiust force is defined as the half o f the difference between the maxhuum and m m i m u m values o f the measurements. Fig, 19 shows the weighting scale for the vertical tluust force measurement.
Fig, 19 The w e i g h t i n g scale f o r vertical thrust force measurement.
Table 6 and Fig. 20 show the experhuental results o f the vertical thmst force measureruent. During the test, only two vertical thrusters near both sides o f H C R are used. W i t h the tivo thmsters, sufficient roll moment can be generated to maintain 90
degree roll attihide. The maximum tiuust force 41.5 kgf'xs obtained w i t h the maxhnum thmst command mput. tt can be seen
Table 6 Experimental results o f vertical thmst test,
Command Vertical thmst force (kg/)
input Mean Deviation 0 0.0 0,1 1/8 4,0 2 3/16 17.5 5 4/16 39,5 3 8/16 40.5 5 12/16 39.0 2 1 41.5 7 0,25 0.5 0,75 Input signal (part of 16)
Fig, 20 Vertical thrust force measurements.
C O N C L U S I O N S
The design o f H C R body requires compromises among conflicting requhements. Tradeoffs may be necessary between stability and mobility, A compromise between operational convenience and design complexity is also necessary. From these considerations, an open fi-ame R O V that has a cleaning bmsh on the top side is selected as the stiuctirre o f HCR.
The robot body design is optimized to reduce drag force, C D F analysis results reveal that drag due to pressure difference is dominant and the influence o f viscosity on it is negligible. A n extensive analysis on the f l o w field around the robot body with C F D suggest that a significant reduction in the drag force can be obtained by adoptmg hemisphere form o f lower rear part and streamlined form o f the bow part o f upper stiuctiire.
The mfiuence o f tijrbulence is included with k - s model in the CFD analysis. Most o f the C F D analysis is conducted w i t h the standard k - s model. To mvestigate the hifluence o f hirbulence model on the drag analysis, additional analysis w i t h the R N G k - s model on the f m a l H C R body model is conducted, h is shown that both the standard and R N G k - s tiir-bulence models produce shnilar ti-ends ui the drag analysis. The comparison o f the drag fore reduction ratio o f the two tiarbul-ence models may not be justified to use k - s model for fiow field analysis. For m-depth analysis, comparisons o f t h e analysis results o f drag force reduction ratio distribution over speed and characteristics o f vortices may be necessaiy as a further research. Experiments are conducted to inspect on the hydrodynamic peiformance o f the robot. The results show that the measure-ments o f t h e drag force are larger than that obtahied fi-om the C F D analysis. This discrepancy is considered to be the different semps f o r the CFD analysis and acmal test. Thus, the measured maximum forward velocity is below the target velocity.
352 Inter J Nav Archit Oc Engng (2012) 4:335-352
A C K N O W L E D G E M E N T
This w o r k is the result o f regional mdustrial technology development project "Development o f ship huU cleaning under-water r o b o f ' supported by the Korea Commerce, Indushy, and the Energy Department. This work was also supported b y the National Research Foundation o f Korea ( N R F ) grant fimded by the Korea govemment ( M E S T ) through GCRC-SOP (No. 2011-0030658).
R E F E R E N C E S
lorra. A . , Caceres, D., Oritz, E., Franco, J., Palma, P. and Alvarez, B . , 2009. Design o f service robots. IEEE Robotics &
Automation Magazine, 16(1), pp.24-33.
Jung, K . Y . , K i m , I.S., Yang, S.Y. and Lee, M . H . , 2002. Autopilot design o f an autonomous underwater vehicle using ro-bust control. Transaction on Control Automation, and Systems Engineering, 4(4), pp.264-269.
Jun, B . H . , Lee, J.H. and Lee, P . M . , 2006. Repethive periodic motion planning and dhectional drag optimization o f under-water articulated robotic arms. InternationalJournal of Condol, Automation, atid Systems, 4(1), pp.42-52.
Jung, Y.S., Lee, K . W . , Lee, S.Y., Choi, M . H . and Lee, B . H . , 2009. A n efficient underwater coverage method f o r m u l t i -A U V w i t h sea current distirrbances. InternationalJournal of Control, -Automation, and Systems, 7(4), pp.615-629. K i m , H.S. and Shin, Y . K . , 2005. Design o f adaptive fuzzy sliding mode conti-oUer based on fiizzy basis function expansion
f o r U F V depth control. InternationalJournal of Condol, Automation, and Systems, 3(2), pp.217-224.
L i , J.H., Lee, P . M . and Jun, B . H . , 2004. A neural network adaptive controller f o r autonomous driving corrtrol o f an auto-nomous underwater vehicle. International Joumal of Control, Automation, and Systems, 2(3), pp.374-383.
Marme Technology Society, 2012. ROVs-A BRIEF HISTORY. [Onlme] Available at: <http://www.rov.org/rov_history.cfm> [Accessed 19 November 2012J.
Roberts, G. and Sutton, R., 2006. Advances in unmanned marine vehicles. l E E , London.
Yuan, F . - C , Guo, L . - B . , M e n g , Q.-X. and L i u , F.-Q., 2004. The design o f underwater hull-cleaning robot. Journal of
The implementation of the integrated design process
in the hole-plan system
W o n - S u n Ruy\ D a e - E u n K o ^ a n d Y o u n g - S o o n Yang-^
'Department of Ocean System Engineering, Jeju National University, Korea Department of Naval Architecture and Ocean Engineering, Dong-Eui Universit)', Korea ^Department of Naval Architecture and Ocean Engineering, RIMSE, Seoul National Utdversit)', Korea
A B S T R A C T : Ad current shipyards are using the customized CAD/CAM programs in order to itnprove the design
qu-alit)' and increase the design efficiency. Even though the data structures for ship design and construction are ahnost completed, the implementation related to the ship design processes are sdll in progress so that it has been the main causes of the bottleneck and delay during the middle of design process. In this study we thought that the hole-plan system would be a good example which is remained to be improved. The people of outfitting division who don't have direct authorit}' to edit the structural panels, should request the hull design division to install the holes for the outfitting equipment. For acceptance, they should calculate the hole position, determine the hole type, and find the intersected contour of panel. After consideradon of the hud people, the requested holes are manually instaUed on the hull structure. As the above, many processes are needed such as communicadon and discussion between the divisions, drawings for hole-plan, and the consideration for the structural or production compadbdit}'. However this iterative process takes a lot of worldng time and requires mental pressure to the related people and cross-division conflict. This paper wid han-dle the hole-plan system in detail to automate the series of process and minimize the human efforts and time-consumption.
KEY WORDS: Hole-plan system; H u l l and o u t f i t t i n g design; Process automation.
I N T R O D U C T I O N
I n the ti-adihonal design envhonment, the designers had been satisfied that the C A D / C A M system could provide the dra-wmgs for products. However, the expectation o f the shipbuilding C A D / C A M has been updated around these days (Misti-ee et al. 1990). M a n y major shipyards are deploying 3D product model to support the variety o f requhements, such as factoiy auto-mation data, reproduction o f the data and the detail product infoiauto-mation.
Aiuong the current issues for the development o f mtegi-ated design support system, this paper focuses on the process automation (Andritsos and Perez-Prat, 2000) which, m general, are occuned betiveen the related several divisions. For this, the design information management system which can reflect the ship design process and solve the process conflict is certainly nee-ded. The generation process o f outfitting holes is the typical example which needs the co-work between the outfitting and hull design divisions. The difference o f design-tune, work scope o f each division, and the numerous revisions are the mam reason w h y the hole-plan system should be automated.
To solve this complicated process on the outiïtting design, a B i l l - o f Material ( B O M ) approach has been inttoduced and suggested by Lee et al. (2010) and Lee (2010). However there are a f e w stirdies dhectiy related to the hole plan system. The
Corresponding author: Dae-Eun Ko e-mail: deko@deu.ac.kr
354 Mer J Nav Archit Oc Engng (2012) 4:353-361
related studies had been miplemented by Y e and K u n (1992), Lee and K i m (1992) and Suh and Lee (2006). A t that tune, the outfitting division did not have the same C A D system w i t h the huh sfructure division. Tiiese researches, therefore, had fried to solve the interface problem between the two different systems. For reference, the major shipyards are recently usmg the unified C A D system. So we are gomg to focus on the process automation which includes the generation o f vufiial holes, the manage-ment o f holes lustoiy, the insert o f automated hole on the panels, and the hole visualization connectmg with C A D system, hi-stead o f t h e interface problem. Based on the concept o f this article, we have developed the hole-plan system and the detail pro-cesses are explahied on the remaming chapters.
T H E C O N C E P T A N D D E F I N I T I O N O F H O L E - P L A N
Holes on the hull structure
There are many lands o f holes on the hull smictiire. They are used f o r the various purposes which hiclude the passage usage (Access hole, M a n hole), the discharge usage (Drain hole, A h hole), the weldmg or assembly usage (Scallop, Slot), and lighte-nmg usage (Lightening hole) or the special usage (Lashmg hole, etc.). Here, the lashing hole is used for tightelighte-nmg the cars on the Pure car and Truck Canier (PCTC) ship, and the tightening holes are also made for the load dispersion, passage usage, besides the weight reduction. Additionally, the outfitting holes which are the main theme o f this aificle consist o f t h e pipe holes, the ventilation holes, and cable holes. Fig. 1 shows the outfitthig pipe network on an engme block.
Fig. 1 A p i p i n g network on the engine block.
Generafion process of outfitting holes
Fig. 2 describes the oveiview o f t h e developed hole-plan system which compares "As-Is" and "To-Be" beuig developed, h i case o f "As-Is", the people o f outtittmg division should make the complicated hole drawmgs (refer to Fig. 3, Ye and ICim, 1992) which could indicate the exact position, type o f holes and the target hull panel. For reference, the hole drawmg had been used as
As-Is To-Be i.Aj^iiUjriij Mo ll.l ü IU>li f.k.delmg f^"^ » \.,lHoh!-;plafl ' b f f w i n g Holt.' LifaMug
main medium between outfitting and hull division, and the designated hole-plan system generates this drawmg only for the evidence material. This drawmg j o b is executed i n repeatmg way eveiy thue w i t h revision works as well as the inhial work. O n the other hand, the people o f hull division must manually add the actual outfitting holes after review and consideration about the requested holes.
Fig. 3 A n example o f hole drawing.
However, all requested holes cannot be approved with the several reasons, which w i l l be explamed later, hr that case, the relevant holes w i l l be sent back to the outfittmg division. It could, therefore, be expected that a lot o f confliction would be ge-nerated between the relevant divisions. Naturally, discussion and meetings are fi'equently needed to settle down such problems, hr conclusion, rt seems that the current mechanism has a lot o f problems.
tvleanwhile, "To-Be" process does not require drawing works. The hole-plan system covers for its role, and the designers o f two divisions are not necessaiy to leave their seats and go to the meeting room.
In detail, the designers o f outfittmg division could make the viitual holes which are the temporaiy ones calculated by the system, not the acttral hole. (Refer to Fig. 4) The hole-plan system generates ah vhtual holes which are located at the inter-section pomt between the hull panel and the outfitting equipment. In addition, their shape is determined by considermg the ins-taUation. For example, a separated pipe has the flanges or sleeves at the both ends for connection with the other pipe. The size o f corresponding virtual holes, therefore, should contain the larger one. For reference, all vutual holes have aheady fixed depen-ding on the outfitting components at the stage o f instaUation. Wlien the vhtual holes are generated, the system checks theh coru-patibility conditions on the shuctural or production view. IVlain checking rtems are the distance between the designated hole and
356 Inter JNav Archit Oc Engng (2012) 4:353-361
the peripheiy parts which are sensitive to the strength, h w i l l be explamed in detail at the next chapter. Anyway, the allowable holes and the mtended holes are fransferred to the hull division for the approval. The huh person hr charge decides whether the hole should be approved or rejected. The approval hole would be mseited on the hull panel. On the other hand, the rejected hole would be retumed to the outfitting with the proper opinion.
The series o f this process on the hole-plan system proceed without any drawmgs or the meeting. Without the sweat o f out-fitting people, they can achieve the vhttial holes and get the approval through the inttanet or web system and fmally all related resuhs can be shown on the C A D / C A M system. Meanwhile, the hull people have only to check the request vhttial holes and determine whether to give their approval or not.
Definition and necessity of hole-plan system
I f we h y to defme "ffole-plan system", it could be said that the requested holes by outfitting division would be automa-tically inserted on the hull panel w i t h the approval o f the hull division with the foundation o f the hull shuchire and the outfitting model on the C A D / C A M D B . In addifion, it ought to be comparatively easy to m o d i f y the holes' poshion and hs type and to insert and delete.
Most o f the holes on a hull structirre are treated by the h u l l division and they are planned at the relatively early stage o f design before node o f the product design on Fig. 5. However, the outfitting holes are a little different to other holes. There are so many o f variability during the ship constmction. The periods o f installation are spread over the constmction sche-dule. For reference, the outfitting installation is possible to be divisible i n the precedent (Shin et a l , 2009), the dock, and the quay installation w i t h the time when outfitting equipment is established. I n case o f t h e precedent installation, it can be subdivided b y " H u l l sttucmre outfitting," " U m t outfitting," "On-block outfitting," and "Pre-Erection outfitting". There is therefore a lot o f probability o f revision on account o f the change o f equipment and facilities, incompatibility w i t h the pre-designed structure and some inevhable situation. For these reason, all design information o f outfitting holes cannot partici-pate i n the preliminary stmctural drawings. Hole-plan system has to handle these unexpected holes.
Estimates Preliminary Desiqri Detail 1 Product 1 Estimates Contract Preliminary Desiqri
Design 1 Design |
Part Assembly
j Sub Block
j Assembly Assembly Painting P.E, F Erection
Launching \ - \ Sea Trial p H Delivery
F i g . 5 Various stage o f outfitting erection during the ship constmction ( W i j n o l s t and Wergeland, 2008).
Expected effect of hole-plan system
It is possible to expect the f o l l o w i n g effect by the application o f Hole-plan system. First o f all, this system can reduce the human e i T o r . Considering the installation constraints and the characteristics o f outfitting equipment, all outfitting com-ponents are assigned and have the proper hole's size and type before this system is implanted on the C A D / C A M software. Secondaiy, the generated holes get to have their histoiy w h i c h contains the generation date, time, designer, and inspector infoimation. So this management system can clarify where the responsibility lies. T h h d l y , h can reduce the w o r k i n g hour. A U related jobs are connected to the C A D / C A M D B . So, the users can find the exact position o f each hole using the h u l l and outfitting model, v e r i f y the hole shape and poshion w i t h the C A D system. I n addition, people o f hull team can auto-matically make the real and approved hole on the h u l l stiaictiare. To reinforce the panel having the several holes, h is usual
that the hull team uisert the additional stmctures such as "Coaming" or "Cariing" around the hole. This system also su-pports to insert autoirratically. Finally, this system is very helpful to increase the rate o f design standardization. There are a lot o f people who execute the outfitting design. Depending on the designers, the installed holes can have various types and sizes without this system.
A U T O M A T E D P R O C E S S O F O U T F I T T I N G D I V I S I O N
Vh-tual hole
Table 1 Various hole types. D H R M r H O H T H O R H E H R FT Contour Cmtcf-r
The vhtual holes are aheady mttoduced at the previous chapter and Fig, 4. The reason that these holes are referred as the vutual hole not the regular hole, is that the hole-plan system temporarily made for the puipose o f the approval o f the hull division. Fig. 6 shows an enlarged vhtual hole which contams the pipe component. A h generated holes have the D-type shape for the efficiency o f automation. Later, the user can change the holes' type with easy usmg this system. Each vhtual hole dis-played on the C A D view has its own annotation w h i c h consists o f a hole name (ship project name + division name / block name / serial no. consti'aint stahis), hole position (XIYIZ), and hole type ( D / H O / H O R / H R / H R M / H T / H E / Contour-refer to Table 1),
Fig, 6 Detail o f a virtual hole.
Constraints of v h t u a l holes
A l l vutual holes must get thi-ough the several compatibility consti-ahits which generally consist o f "interference with the other sbucttires" and "the distance with the mahr components o f panels". These consttaints have relevance to the local stiucttiral sttength and the consttuction reasons which are usually determined with division or company mles.
358 Inter J Nav Archit Oc Engng (2012) 4:353-361
Table 2 Compatibility checlc items o f the virtual holes.
Constraint type Reason
Stiffener Approach ( S A ) Stmctural strength
Seam Approach ( S M ) StrucUiral strength & Production
Overlap ( O L ) Production
None Spec. N o data
The hole-plan system classifies the verification resuh by the specific name which was explained as the "consh-aint stahjs" at the previous sub-chapter. I f a vhtual hole pass all consfraints, it has the "Stand-by (SB)" mark on theh note dhnension. Table 2 shows the checkmg marks when the vhtual holes violate the compatibility condhions.
Note that the verification is hnplemented at the stage o f outfihmg process not the huh stage. The system uitends to check hr advance and the people o f hull to check only the vutual hole which has passed all compatibility condition. The oufiitting people also must identify whether the relevant vhtual hole has passed, and then request for approval.
Table 3 shows the representative case o f violating the constramts. The fust shows a hole approaches the stiffener too closely, the second is located on the seam line, and the last get close to other holes veiy much, h i the last case, the user must combine near holes so as to make them one hole.
Automated process
This system is mauily implemented usmg "Python" (ver. 2.6, littp://www.python.org/) which is the only language for cal-Img the A P I functions mherent m 'I ribon M 3 system (http://www.aveva.coin/). The designer o f oulfittmg division starts with selectmg the outtittmg components ( @ on Fig. 7) and the related hull bloclcs ( ® ) , and then makes the vutual holes usmg the button (3). That resuh would be hsted up at the ® which mcludes aU generated vutual holes, theh name, and the compatibility statiis. Tlie view button((9)) displays the 4 views which include the hull stiuctiire, the oufiitting network, and the vutual holes. Fig. 8 displays only a plan and ISO view among 4 views about an enghie block. A f t e r solvmg the violation o f compatibility condition, the button © enables to request the selected viitual holes to the huh division by e-mail or mttanet. hr addition, the users can verify the hole histoiy using the button (7) and make the hole drawmg usmg the button ® . I f the hull team accept the requested holes, this hst can be checked on the © hst box. I f not, the rejected holes are listed on the ® list box.
1 Miilll'ar.,-1 1( il J O E Ï C - B ( » - P 1 1 O E ö : . e i ( e - p i i O E ï s : - e n ) : O E £ t - 6 t o : ra <1 L ' : !.M0 O E K - e x c - p s O E ï c - e i i w ,g Pip* Vttlly I Show SeUctlofi 1 riiii,/:;tniciu(B 1 0 I SK-Mf, O E K - « f ) O E £ i ; - « : « O E K - * : » O E r C- ^ 1 3 3 O E r e - « l i l O E3X-MI3Ï I Vliluil llolo 1 C r. ,> l . V , , 1,1 ..I H o l . ' O f'.v {.IA O HP J K J . V t K C / W i . ï S O KP?)^Mrt£CAW6-SI O KPW«.VE5.V/W7.SB O K P ! l » * y E 3 C t / t M - S a . j i ^ . . \ T ^ / ^ ' . : ' P .1 1-:: : v ' . t : c ' g HP»W.I/EÏ.V/C6Ï.SB O H P ï y ? ï . i / t ï > : / W . O i . O HPSrJW/EïtAW.C'L O H P / Ï ^ W / E S ; t A « - S B ^^iuMA •'•'•"'•I-a f . M f'lrl fl,-.*-',';»-?)»! _ |:cL.»;.';.'i: s i - ' m i ;
Table 3 Violation cases o f virtual holes.
< F R 3 7 - ' 1 3 . 7 1 8 . 5 , 8 :
Fig. 8 The involved h u l l , outfitting, and generated virtual holes.
A U T O M A T E D P R O C E S S O F I f U L L D I V I S I O N
The huU team needs only to determme whether the requested hole is acceptable or worth refusing. He can comment his opmion about the rejected ones on its histoiy record. He w i l l start with the mput o f his own I D and check the requested hole at the each b l o c k ( ® on the Fig. 9). When click the designated block, the list o f correspondhig viitual holes listed on the box((2)). Anythne he want, he can check the selected hole mformation t f o m the view button((7)). The button © enables the selected vutual holes to be the confmued holes and automatically reflect on the real shucttiiul D B and they are also listed on the @. Fig.
10 depicts the confmrred holes on the shuchiral panel. Note that all resultant holes on the Fig. 10 have not passed the given consttamts and generated for just example. As soon as the user presses the button © , the reject resuh wiU be sent back to the outfittmg person m charge. I n addition, he can check the hole histoiy from the @ box.
360 Inter J Nav Archit Oc Engng (2012) 4:353-361
( V WB BI Iteta I r E ! E E P t ^ O t » . I Actotl ttoto 1 IW EA r ö M t S
r T i ï a ^ ^ H r I H r :'••;; 1 r £!BH I H P M i V I . . . . , , - ,'l : inniTT; V P — » EllC-iCf F.'J S ^ W r U - j : 1) 0 E I M ^ C F : ^ F.0: 0 W . D NO; 0 Q E y ? {CF:IW fiO: 0.W: 0 fW: 3 Q E31P-(CF;I"8 P.0: 0 Vi; 0 f » : D) '
Q EXT-tCf HJ fill; O.VZ; 0.fM: 11 Ü EliPHCF in P-0: 2 Vi. 0, ffö: Ï) Ü EtP-(Cr il7,fïO; O.VÏ: Ö fJö: P) 0 EO:-<Cf: er, F.0: 0 V? 3, KO; Q) 0 f J - t C F ; a HO: 0 V?; a KO: Cn 0 E%>:-(Cf: 0 (Vl: O-VÏ OJiO: 0)
e
_ J / . • , • • -, 0 E I I C - ^ X I P . 0 Eitc-ET^'.'.-n H E S . 0 EMC-ff'.;7A . 0 EIIC-FRjCA-PI . Ó ÉltC-fHÏS . 0 ElIC-fBÏ» * 0 EllC-ffi«0 . 0 EllC-fB*>a-Pi . 0 E i i c - f a w . 0 EllC-fft"M . 0 EllC-FTiJO . Q EIIC-QBCF > 0 Eiic-'S'tte . 0 EHC-t'-fltB+S -.- 0 EHC-^i''.^ . 0 E I ! C - 6 = : B - P - M ^ . 0 €IIC-G»WI zT r ö M t S r T i ï a ^ ^ H r I H r :'••;; 1 r £!BH I H P M i V I . . . . , , - ,'l : inniTT; V P — » EllC-iCf F.'J S ^ W r U - j : 1) 0 E I M ^ C F : ^ F.0: 0 W . D NO; 0 Q E y ? {CF:IW fiO: 0.W: 0 fW: 3 Q E31P-(CF;I"8 P.0: 0 Vi; 0 f » : D) 'Q EXT-tCf HJ fill; O.VZ; 0.fM: 11 Ü EliPHCF in P-0: 2 Vi. 0, ffö: Ï) Ü EtP-(Cr il7,fïO; O.VÏ: Ö fJö: P) 0 EO:-<Cf: er, F.0: 0 V? 3, KO; Q) 0 f J - t C F ; a HO: 0 V?; a KO: Cn 0 E%>:-(Cf: 0 (Vl: O-VÏ OJiO: 0)
inniTT; V P —
» EllC-iCf F.'J S ^ W r U - j : 1)
0 E I M ^ C F : ^ F.0: 0 W . D NO; 0
Q E y ? {CF:IW fiO: 0.W: 0 fW: 3
Q E31P-(CF;I"8 P.0: 0 Vi; 0 f » : D) '
Q EXT-tCf HJ fill; O.VZ; 0.fM: 11 Ü EliPHCF in P-0: 2 Vi. 0, ffö: Ï) Ü EtP-(Cr il7,fïO; O.VÏ: Ö fJö: P) 0 EO:-<Cf: er, F.0: 0 V? 3, KO; Q) 0 f J - t C F ; a HO: 0 V?; a KO: Cn 0 E%>:-(Cf: 0 (Vl: O-VÏ OJiO: 0)
° 3 E 1
TOiï'Sa si;ïUBj EMasj sBsa» m a m BoaeB» .w5i<»si. CF:BS«
Fig. 9 The procedure o f h u i l division.
U
0 ° ——• J 0 06o°0
° 0
ó°
Fig. 10 The c o n f m n e d o u t f i t t i n g holes.
C O N C L U S I O N S
A f t e r the preliminary ship design, the detail design is generally carried out by the two major divisions. One is the huh d i v i -sion and the other is the outfittmg design, ft is fact that the efficiency o f these divi-sions has been mcreased and theh own pro-cesses get to be mcreasmgly smooth b y help o f t h e C A D / C A M system. However, it is an exception to generate the oufiitting holes on the structtiral panel, ft is easy to image that there could be the meetmg point where two division must cooperate to make the oufiitting hole which satisfy the space o f oufiitting equipment and also the compatibility condition o f stiuctiiral behavior.
This article tiles to automate the all design process o f t h e oufiittmg holes by tivo division pomts o f view. The automatic generation o f virtual holes, the hnkage with the C A D system, the communication usmg the mttanet or E-mail, the holes' history management and the automation o f compatibilify checkmg are the developed catalog at the oufiittmg division side. O n the other hand, the automatic scheme, the hole verification usmg the C A D system, and the information management system are the advantages at the hull division side.
A C K N O W L E D G E M E N T S
This work was supported by the hrdustiial Sti'ategic Technology Development Program (10035331, Simulation-based Manufacttirmg Technology f o r Ships and Offshore Plants) fiinded b y the M m i s t i y o f Knowledge Economy ( M K E , Korea)" and also a part o f t h e project titled "Optunized logistics ttansportation system on ports" fiinded by Mmistry o f Land, Transport and Maritime A f f a h s , Korea.
R E F E R E N C E S
Andritsos, F. and Perez-Prat, J., 2000. The automation and integration of production processes in shipbiulding. European Commission Joint Researcli Centre Institute f o r Systems, Informatics & Safety.
Lee, J.K. and ICim, J.H., 1992. The constmction o f ship outfitting design system based on the C A D system. Bulletin ofthe
Society of Naval Architects of Korea, 29(3), pp.28-35.
Lee, J.J., 2010, Data association between design stages f o r the B O M management o f piping i n ship design. Master disserta-tion, fnha University
Lee, J.H., K i m , S.H., Lee, H . B . and Jang, G.S., 2010. A n implementation enterprise B O M f o r marine vessel's pipe equip-ment. Korea CAD/CAM Conference, pp.632-643.
Mistree, F., Smith, W . F . , Bras, B . , A l l e n , J.K. and Muster, D . , 1990. Decision-based design: A contemporaiy paradigm for ship design. Transactions, Society of Ncn'al Architects and Marine Engineers, 98, Jersey City, N J , 1990, pp.565-597. Suh, H . W . and Lee, S.G., 2006. Integrated C A D system f o r ship and offshore projects. InternationalJournal of CAD/CAM,
6(1), pp.41-48.
Shin, S.C, Cho, J.B., Shin, K . Y . and K i m , S.Y., 2009. Process improvements for elevating pre-outfitting rate o f FPSO.
Journal ofthe Societ}' of Naval Architects of Korea, 46(3), pp.325-344.
Wijnolst, N . and Wergeland, T., 2008. Ship innovation. IOS Press B V under the imprhrt D e l f t University Press.
Ye, K . H . and K i m , D.J., 1992. The impleiuentation o f the Hole Plan/Coaiuing System on the ship customized C A D . DSME
Inter J Nav Archit Oc Engng (2012) 4:362-373 Iittp://clx. doi. org/10.3 744/JNAOE.2012.4.4.362
T e c h n o l o g i c a l a n d e c o n o m i c s t u d y o f s h i p r e c y c l i n g i n E g y p t
Y o u s r i M. A . W e l a y a ' , M a g e d M . A b d e l N a b y ' a n d M i n a Y . T a d r o s '
'Department of Naval Architectm-e & Marine Engineering, Alexandria Universit}', Eg}'pt
A B S T R A C T : TIte ship recycling industiy is growing rapidly It is estimated that the Internadonal Maritime
Organiz-ation's (IMO) decision to phase-out single hud tankers by 2015 will i-esult in hundreds of ships requiring disposal. At present, the ship recycling industry is predominantly based in South Asia. Due to tlie bad praedce of current scrapping procedure, the paper wdl highlight the harm occurring to health, safet}' and environment. The efforts of the Marine Environment Protection Committee (MEPC) wliich led to the signing ofthe Hong Kong Internadonal Convention are also reviewed. The criteria and standards required to reduce the risk and damage to the environment are discussed and a proposed plan for the safe scrapping of ships is then presented. A technological and economic study for the ship recycling in Eg}>pt is carried out as a case study This includes the ship recycling facdit}' size and layout. The equipment and staff required to operate the facility are also evaluated A cost analysis is then carried out. This includes site development, human resources, machineries and equipment. A fuzzy logic approach is used to assess the benefits ofi the ship breaking yard. The use of Ihe fiuzzy logic approach is found suitable to make decisions fior the ship brealdng in-dustiy. Based on given constraints, the proposed model has proved capable ofi assessing the profiit and the internal i-ate ofi return.
KEY WORDS: Ship r e c y c l i n g ; Ship scrapping; M a r i n e environment protection; Ship breaking y a r d ; F u z z y l o g i c approach.
N O M E N C L A R U R E and A B B R E V I A T I O N S
A Annual Aiuounts MEPC Marine Enviromnent Protection Committee
i hitemal rate o f renini per year N Number o f years
ICS Indushy Code o f Practice NPV Net Present Value
ILO hiteiuational Labor Organization OSH Occupational Safety and Health
IMO International Maritime Organization P Present worth
IRR Intemal Rate o f Retum SPW Single Present W o i t t i
I N T R O D U C T I O N
Due to the fast growth m ship recycUng indushy worldwide, especially m Asian countties such as hidia, Bangladesh, Chma and Paldstan, it has become a major source o f income for those countties. However, there is a growing concem i n the mtema-tional maritime community over the coirditions m which ships are scraped at present. The need to develop strmgent laws to guarantee the safety o f personnel working m this mdushy and the sunounding envhomuent is exfremely hnportant.
There are gi'eat prospects for this indushy in Egypt due to the excellent location o f Egypt i n the intemational ti'ade route.
Corresponding author: Yousri M. A. Welaya e-mail: y_wela}'a@hotmad. com
