nd Regional Conference On Enhancing Transpor Technology For Regional Compeiiveness SESSION C
TBLE OF CONTENTS PREFCE... 4 ORGNISING COITTEE... 5 KEYNOTE DDRESS... 6 Session : UTOOTIE... 7 Session B : ERONUTICS... 37 Session C : RINE TECHNOLOGY... 7 Session : UTOOTIE... 9 Session B : ERONUTICS... 55 Session C : RINE TECHNOLOGY... 77 Session 3 : UTOOTIE... 5 Session 3B : ERONUTICS... 55 Session 3C : UTOOTIE... 37 Session 4 : UTOOTIE... 353 Session 4B : UTOOTIE... 385 Session 4C : UTOOTIE... 45 Session 5 : UTOOTIE... 455 LIST OF REIEW... 486
STUDY ON SHIP CPSIZING DUE TO COLLISION Zobair Ibn wal a & Dr.. Rafiqul Islam b a cciden Research Insiue (RI), Bangladesh Universiy of Engineering & Technology (BUET) b Faculy of echanical Engineering, Universii Teknologi alaysia (UT) E-mail: rafiqulis@fkm.um.my bsrac This paper invesigaes he ship capsizing due o collision wih anoher ship in calm waer. mahemaical model on collision dynamics has been developed and validaed. The dynamic characerisic (roll in paricular) has been sudied numerically. Searches have been made o find he survivabiliy boundaries in erms of sriking velociy, coefficien of resiuion, collision angle, collision ime and verical posiion of hiing poin. aximum ampliude of rolling moion has been deermined agains he said parameers and have been presened as 3D surface chars which enable o idenify safe boundaries of operaion for wo given variables a a ime. This paricular approach of collision simulaion has he poenial of analysing operaional risks during a ship s design sage and hereby incorporaes necessary modificaions if required. Recommendaions have been pu forward for fuure sudies. Keywords: Ship capsize, simulaion, collision dynamics, rolling.. Inroducion any mariime counries around he world frequenly encouner `problems of ship collision quie ofen boh in he inland waerways and in he seas as well. Ship collisions are of paricular imporance due o he following reasons: The environmenal impac, especially in he case where large anker ships are involved. However, even minor spills from any kind of merchan ship can creae a hrea o he environmen. The loss of human life is invaluable. Financial consequences o local communiies close o he acciden sie and consequences o ship-owners, due o ship loss or penalies. The increase in business aciviy in many developing counries has resuled in denser sea roues and a he same ime faser ships for quick ransporaion. Therefore, he possibiliies ha a ship may experience a major acciden during her lifeime are higher. Denser sea roues increase he probabiliy of an acciden in paricular collisions involving ships or offshore srucures. Due o exremely large masses and relaively high velociies he energy involved in such an acciden is asonishing. n even like his may confrons a ship o susain severe srucural damage and capsizing as well. Over he years here have been numerous cases of ship capsizing and collision accidens. sudy by wal and e. al [] on inland shipping accidens reveals ha around 4 percen of oal accidens in Bangladesh occur due o collision. Similarly, sudies on ship capsizing exhibi he disasers are recurring quie frequenly around he world as well. number of Norh Sea rawlers from he UK and oher EU counries have capsized in heavy seas in he pas hiry years. leas one RO-RO vessel becomes a casualy in each week [, 3]. In Canada he average number of accidens due o collision and capsizing is around 45 per year [4].. Background and Scope of Sudy The problem of capsizing has been generally reaed as a phenomenon originaing from wave and wind forces. number of in deph research has been conduced over he years o undersand he capsizing mechanisms and ways o preven i [5, 6]. Lin & Yim [7] used he new subjec of chaos o analyse he non-linear equaions devised o represen he moion of ships in roll-sway coupled moions. They showed four ypes of capsize: Non-oscillaory capsizing in which he resoring momen is small compared wih he momens of wind and waves exered on he ship. Oscillaory sudden capsizing in his case resoring momen should be sufficien bu insabiliy is caused by successive series of waves. Oscillaory symmeric build-up capsizing, here ampliudes of rolling moion increase
rapidly afer only a few cycles similar o linear resonance. The build-up is likely o be caused by a series of waves. Oscillaory ani-symmeric build-up capsizing. In some cases he rolling moion appears o be anisymmeric wih respec o he axis of symmery abou he ime axis. This again appears as he resul of passing hrough a succession of waves producing oscillaions, which are so large ha recovery is impossible. One of he mos serious incidens was illusraed by Car Ferry Wahine disaser in 968, as described by Conolly [8]. The ype of disaser is referred o as broaching. In his case he ship is ravelling wih a sern sea slighly o one quarer. The ship experience difficuly wih he rudders being increasingly ineffecive. Large yaw angles will be experienced and he ship will roll hrough a large angle o leeward. The ship is said o be broachedo and he breaking waves over he ship and he wind effecs may be sufficien o capsize. Spyro [9] has analysed a phenomenon known as surf-riding where he ships is saionary relaive o he wave rough. The simulaion is an enrapmen of he vessel for prolonged periods a exacly zero frequency. The auhor showed ha how wih he fixed conrol sysem he ship becomes unsable and resuls capsizing. However, sudy on capsizing due o collision has ye o be fully undersood and published lieraures on his opic are rare. Recenly, wal [] and wal & e. al [] has conduced a research where such an aemp has been aken, ye he furher developmens are needed. Therefore, in his sudy i has been aemped o invesigae capsizing behaviour of a ship which is encounering a collision wih anoher similar vessel. The sudy searches he survivabiliy boundaries in erms of sriking velociy, coefficien of resiuion, collision angle, collision ime and verical posiion of hiing poin. 3. Theoreical odel for Ship Collision Dynamics Considering a collision scenario, as shown in Fig., where Ship srikes Ship B, wo co-ordinae sysems may be assumed for each ship such as X-Y for sriking ship and I-J for sruck ship. Ship Y Ship B X J I C hea Fig.. Co-ordinae sysem of a ship-ship collision. phi Using simple rigonomeric relaions he collision forces in he respecive axes on boh he sruck and sriking ship may be compued. For example, forces on Ship in X-axis and Y-axis direcion are, F X = F cos + F cos (9-) () F Y = - F sin + F sin (9-) () Similarly, forces on Ship B in I-axis and J-axis direcion are obained as, F I = F cos (-) + F cos (-) (3) F J = F cos (-) + F cos (-) (4) Here, forces F and F are perpendicular forces acing a he conac poin C developed from he impac beween he wo bodies. I is known ha impac force a a paricular direcion is equal o change of linear momenum in ha direcion, i.e. F equals he change in momenum in -axis direcion and F equals change in momenum in -axis direcion Therefore, by using hese expressions he forces may be obained, For Ship, afer F Tcol (5) afer F Tcol (6) For Ship B, Bafer B FB B Tcol (7) Bafer B FB B T (8) 3. pplicaion of Coefficien of Resiuion The mos fundamenal approach owards solving he problem is considering he ship velociy afer collision as a funcion of coefficien of resiuion and he ime required o resiue or simply he collision ime. The applicaion of hese wo variables are however, very criical and requires careful assumpion o model a poenially realisic scenario. The coefficien of resiuion is a measure of he elasiciy of he collision beween wo objecs. Elasiciy is a measure of how much of he kineic energy of he colliding objecs he collision remains as kineic energy of he objecs afer he collision. There are hree ypes of collision: Perfecly Elasic, Perfecly Inelasic and Elasoplasic collision. perfecly elasic collision has a coefficien of resiuion of. Example: wo diamonds bouncing off each oher. perfecly inelasic, collision has E =. Example: wo lumps of clay ha don' bounces a all, bu sick ogeher. On he oher hand an elasoplasic collision, some kineic energy is ransformed ino deformaion of he maerial, hea, sound, and oher forms of energy. For his ype he coefficien of resiuion varies be beween zero and one. Now when a collision sars aking place he change in momenum is equal o he impulse inegral and he col
common velociy a he beginning of he resiuion ime reaches he maximum level or in oher words he velociy reaches maximum a he end of compression. Therefore, according o he impulse momenum heory he following may be obained for ime beween sar of collision (=) and maximum compression (= ), For Ship : F d (9) B F d () For Ship B: B B B B F d () B B B F d () However, according o impulse momenum heory he following relaions mus saisfy along he axes, F d FBd (3) F d FBd (4) Thus operaing he above relaionships he common velociies are obained along he wo axes, BB B (5) B B B B (6) B Similarly beween he maximum compression and full separaion of he ships he followings relaions are obained for common velociies, B B afer afer B (7) B B B afer afer B (8) B I is now possible o esablish a relaionship beween impulse inegrals wih he help of Coefficien of Resiuion (E). This relaion can be expressed as he following: Fd E Fd (9) Using equaion (9) o (9) i is now possible o obain he expressions of velociies afer collision for boh he ships. Such as, afer E E () E E E E B E E B afer () Bafer B () Bafer B (3) 3. Loss of Kineic Energy The loss of kineic energy is herefore obained as, Ship : KE afer (4) KE afer (5) Ship B: KEB B B Bafer (6) KEB B B Bafer (7) 3.3 Soluion of he Equaion of oion The equaion of moion is required o be solved wih necessary boundary condiions in order o find he ships responses due o collision forces. During a collision he equaion of moion may be expressed as he following, d xi dxi b c x Fc ( ) ij ij i i (8) d d Therefore, he general soluion of he equaion may be expressed as, x i e sin cos (9) Where, and are consans which are needed o be deermined using appropriae boundary condiions. ssuming an iniial condiion when he collision force is maximum a ime = max =, he displacemen is x i = x i. ccording o he heory of simple harmonic moion, his ampliude or displacemen is maximum when he velociy reaches o zero and he velociy becomes maximum when he ampliude becomes zero unis. Therefore, assuming x i is he maximum ampliude due o collision force a he ime =, he following unknowns are obained from he equaion of general soluion as derived above. x i and x (3,3) i nd herefore, he general soluion becomes, e x i x i cos sin (3) The above equaion is similar o he damping par of any equaion of moion where x i resembles he maximum ampliude due o an exciaion and e resembles exponenial decay of he moion. I is, however, imporan o menion ha in his paper only rolling moion (x 4 ) is being invesigaed o sudy he capsizing phenomena of he vessels under collisions. 3.4 Force as an Exponenial Funcion of Time The ime hisory of force is considered vial for solving he equaion of moion in ime domain. However, experience sugges ha in mos of he pracical cases he force-ime daa is exremely
difficul o predic since i involves complicaed inernal srucural arrangemen, including he exernal fenders, of ship hull ha are subjec o progressive srucural deformaions/failures by buckling, shearing, earing, crushing, bending and wising of plaes, sringers, panels ec during a collision. wal [] proposed several force funcions in his aspec bu he formulaions are ye o be experimenally verified. In his paricular sudy he force is assumed o be an exponenial funcion of ime where he force increases exponenially from ime hi o ime max and hereafer i reduces exponenially again from ime max o ime sep as shown in Fig.. Force F max exceed he comparaive limis and hus i may be concluded ha he developed model is in good agreemen. Table. Comparison of loss of kineic energy a (m/s) b (m/s) (deg.) Peersen (98) KE (J) Hanhirova (995) Zhang (999) Pzresen Sudy Peersen (98) KE (J) Hanhirova (995) Zhang (999) Presen Sudy 4.5 9 69.6 54.4 7. 78.5 4.5 4.5 9 4.7 4.5.4 6.7 64. 54.4 7. 78.5 4.5 4.5 6 5. 5.8. 3.7 8.8 8.3 35.3 58.89 4.5 4.5 3 49.3 7..6 7.9 4 7.4 9.63 4.5 9.8 4 5 9.63 54 4.9 5. 58.89 4.5.5 4.7 5.5 45. 6.7 6.3 4.8 57.5 58.89 3.5. The Hydrodynamic Coefficiens hi max sep Time Fig.. Collision force is an exponenial funcion over he conac period. The paricular inegral of he funcion Fci ( ) Fmax i f ( e ) may be obained as he following, max Fmax e xi [ = hi o = max ] (33) a b c ij ij 3.5 alidaion of he odel ij The developed model has been compared wih a number of published research works which are described in he following paragraphs. 3.5. Comparison of Los Kineic Energy The comparison of loss of kineic energy has been compued using wo similar ships of lengh 6 meer. The pariculars are breadh 9. meers, draf 6.9 meers, displacemen,34 ons and coefficien of resiuion zero. The collisions were a various angles of aack and speeds as well. For validaion i is considered ha he hiing akes in place a he midship of he sruck ship and he collision is enirely plasic. plasic collision means ha he ships remain in conac afer he collision and all he kineic energy is being used in deforming he ships hull srucure and dynamic movemen of he ships. The resuls are compared wih he loss of kineic energy along -axis (KE) and -axis (KE) direcions and he unis expressed here are in mega joule. The resuls are compared wih published daa of Peersen [], Hanhirova [3] and Zhang [4] as shown in Table. The comparison suggess ha here are noiceable variaions among differen mehods adoped by differen researcher; neverheless, he resuls do no The hydrodynamic coefficiens a ij, b ij and c ij depend on he hull form and he ineracion beween he hull and surrounding waer. The coefficiens may also vary during a collision as well and he range of variaion is even wider considering open or resriced waer condiions. However, for simpliciy inorsky [5] proposed o use a consan value of he added mass coefficiens of ships for he sway moion, m ay = a =.4. The added mass coefficien for rolling is suggesed by Bhaacharyya [6] o be in beween o percen of he acual displacemen of he ship. However, in his sudy he hydrodynamic coefficiens were deermined using he 3-D source disribuion mehod [7] and he values are compared wih exising resuls expressed in range of virual mass (Table ). I is observed from he comparison ha he hydrodynamic coefficiens for surge sway and yaw fairly maches wihin he range excep a few discrepancies in he sway moion. This is probably because he range is deermined on he basis of ships ha are relaively large and ocean going in comparison o he small vessels designed for inland ransporaion. Table. Comparison of virual mass (non dimensional). Hydrodynamic Coefficiens Range of irual ass 46 m essel (3-D ehod) 3 m essel (3-D ehod) Surge, a..7.. Sway, a..3.5.4 Roll, a 44...6. Yaw, a 66..75.4.53 4. Resuls and Discussions 4. The model ship This sudy considers ypical inland vessel of lengh 3.64-meer and 46.8 meers for sruck vessel and sriking vessel respecively. Fig. 3 depics a ypical collision scenario generaed using 3D mesh for his paricular sudy.
Fig. 3. ypical 3D esh collision of wo ships The breadh and deph of he sruck vessel are 6.7 meer and 3.5 meer respecively. full load he displacemen of he sruck ship is 498 ones and he angle of vanishing sabiliy is around 63 degrees in sill waer condiion as shown in Fig. 4. collision angle and zero o around.8 second of collision ime. Therefore, any surface excluding hese boundaries represens a safer siuaion while oher variables are kep consan. In his case he ship may survive a collision if i operaes or designed wih he following parameers: (a) Collision angle is in beween zero o 4 degrees, (b) collision ime roughly greaer han.8 second. Fig. 6 represens a 3D surface char for a range of coefficien of resiuions (E) and speeds of he sriking ship ( b ). I is observed ha boh he variables influence he rolling ampliude linear proporionally. The range for safe surface is observed o exis in beween he following variables: (a) The coefficien of resiuion in beween o.6 and (b) sriking speed less han.5 meers/second (5 knos). 75. 6.-75. 45.-6. 3.-45. 5.-3..-5. Fig. 4. Hydrosaic roll sabiliy of he sruck ship 4. Resuls on aximum mpliude emp has been aken o invesigae he capsizing phenomena considering wo variables a a ime as shown in he following figures. Fig. 5 represens he maximum rolling angle agains wo differen variables namely, he collision ime and collision angle. Collision ime refers o he ime required for conac, compression, resiuion and separaion of he conac surfaces of he ships. I has been observed from he char ha he relaion beween collision ime and rolling ampliude is exponenial and he relaion beween collision angle and roll ampliude is rigonomeric. Roll ampliude (Deg). 9. 6. 3...5..5..5 3. Colime. 5. 3. 45. 6. 75. 9. 9.-. 6.-9. 3.-6..-3. Colangle (deg) Fig. 5. aximum roll ampliude agains collision ime and collision angle. The surfaces, as shown by separae colours, represen a paricular region for range of rolling ampliude where collision ime and collision angle are he variables. In order o find a safe condiion, he firs and mos imporan parameer is ha he rolling ampliude doesn exceed he angle of vanishing sabiliy. For example in his figure he op wo surfaces (roll angle 6 o 9 and roll angle 9 o ) represen he unsafe condiion which exiss roughly in he range of 4 deg o 9 deg of Rolling mpliude (Deg) 6. 45. 3. 5....5.5.75. E.5.3.54.6.57 3.9 b (m/s) Fig. 6. aximum roll ampliude agains coefficien of resiuion and sriking ship s speed. Fig. 6 shows a surface of collision angle and heigh of conac poin above he cenre of floaion. I is observed ha he higher he posiion of he conac poin he higher is he ampliude of role while ohers variables are kep consan, The ampliude, however, reduces quie significanly wih he decrease in he verical posiion of he conac poin. However, he range of safe surface exiss beween he following ranges: (a) Collision angle less han 35 degrees and (b) heigh of collision conac poin below.75 meers measured above cenre of floaaion. 9.-. 6.-9. 3.-6..-3. 3.5.75 Heigh above CF (m).7.88. 5. 3. 45. 6. 75. 9..7 Collision ngle (Deg) Fig. 7. aximum roll ampliude agains collision angle and heigh of conac poin 4.3 Time Domain Simulaion Fig. 8 represens he ime domain roll simulaion of he sruck ship and he decay of. 9. 6. 3.. Rolling mpliude (deg)
moion afer collision. I is observed ha higher sriking speeds cause higher he momen for rolling and hus higher rolling ampliude. lhough his phenomena is noneheless a common fac bu he key aspec is o observe he ampliudes which are being reduced significanly by aleraion of he coefficien of resiuion. I is observed ha up o 83 percen of he rolling ampliude may be reduced if zero resiuion maerials are being used. This is indeed, a very imporan aspec of he research findings ha as excessive rolling causes ships o capsize and such capsizing could be prevened by applying he lower resiuion shock absorbing maerials. This phenomena is simulaed he figure. where he ship is sruck a 6 knos and rolled over he angle of vanishing sabiliy considering he value coefficien of resiuion around.. The roll ampliudes, however, are significanly less in heir respecive cases if he coefficien of resiuions were considered zero or close o zero. Therefore, he facs revealed here could be a maer of life and deah and indeed requires recogniion o be looked ino while consrucion ship fenders and oher similar proecive devices. mpliude (Degree) mpliude (Degree) mpliude (Degree) 5 5-5 - -5-5 5-5 - -5-5 5-5 - -5-5 5 5 3 35 4 5 5 5 5 5 5 Time (sec) Fig. 8. Rolling of sruck ship hi a 9 deg. a speeds (op), 3 (middle) and 6 (boom) knos. 5. Conclusions Based on he resuls, i could be concluded ha vessels of his paricular ype plying in relaively calm waers may use fenders made of maerials having coefficien of resiuion less han.5 and resiuion ime greaer han second o avoid consequences due o collision wih a similar vessel. In addiion, he collision angle has o be less han 33 degrees, he relaive sruck speed has o be less han.5 meers per second and he verical collision conac poin has o be less han.75 meers above he cenre of floaaion. E. E.5 E.5 E.75 E. The research on sudying he capsizing of ships due o collision is sill in he iniial developing sage. The applicaion of he parameer of coefficien of resiuion of he hull maerial for collision capsizing analysis is considered fundamenally new. So far limied knowledge is available o he researchers abou is affec on ship s dynamic behaviour. Furher research on his model is herefore recommended, as i seems highly poenial in erms of suggesing survivabiliy boundaries for various hazardous operaing condiions and hereby, saving invaluable human lives and resources. References. wal, Z.I., Islam,.R., and Hoque,.., arine ehicle cciden Characerisics in Bangladesh: Sudy on Collision Type ccidens, Proceedings of he 7h In. Conf. on ech. Eng. (ICE 7), ICE--46, Dec 7, Dhaka.. Whie,.S., e.al., Conrol of ship capsize in quarering seas, In. Journal of Simulaion, ol. 8, No., pp. -3. 3. assalos, D., e.al., Numerical and Physical odelling of Ship Capsize in Heavy Seas: Sae of he r, 6 h Inernaional Conference of Sabiliy, pp 3-5. 4. Transporaion Safey Board of Canada, Saisical Summary arine Occurrences, 5. 5. Hammamoo,., e. al., 996, odel Experimens of Ship Capsize in sern Seas, Japan Sociey of Naval rchiecs, 79, pp 77-87. 6. King, D., and Scalvovnos, P.D., Numerical sabiliy analysis for ime domain ship moion simulaions, J. of Ship Res., ol 39(4), pp 33-3. 7. Lin, H. and Yim, S., 996, Chaoic roll moion and capsize of ships under periodic exciaion, wih random noise, Journal of pplied Ocean Research,.7, pp 85-4. 8. Conolly, J.E., 97, Sabiliy and conrol in waves a survey of he problem, Journal of echanical Engineering Science, Insiuion of ech. Engineers, ol 4, No. 7, pp 86-94. 9. Spyro, K.J., 996, Dynamic insabiliy in quarering seas: The behaviour of a ship during broaching, Journal of Ship Research, ol 4, No., June, pp 36-46.. wal, Z.I., Developmen of a Collision Dynamics odel for Ship o Ship Collision in he Inland Waerways of Bangladesh,.Sc.Engg. Thesis, Dep. of Naval rch. & arine Eng., Bangladesh Univ. of Eng. & Tech. (BUET), Dhaka.. wal, Z.I., Islam,.R., and Baree,.S., Developmen of Collision Dynamics odel for Ship o Ship Collision, Proc. of he 7h In. Conf. on ech. Eng., ICE--47, 7 Dhaka.. Peersen,.J., Dynamics of Ship Collision, Ocean Engineering, ol. 9, No. 4, pp. 95-39, 98.
3. Hanhirova, H., Exernal Collision odel, Safey of Passenger/RoRo essels, Helsinki Universiy of Technology, Ship Laboraory, Ocober, 995. 4. Zhang, S., The echanics of Ship Collisions, Ph.D. Thesis, Deparmen of Naval rch. & Offshore Eng., Technical Universiy of Denmark, DK-8 Lyngby, Denmark, January 999. 5. inorsky,.u., 959, n nalysis of Ship Collisions Wih Reference o Proecion of Nuclear Power Plans, J. of Ship Res., ol. 3 No., pp. -4. 6. Bhaacharyya, R., 978, Dynamics of arine ehicles. John Wiley & Sons, New York. 7. Islam,.N., Islam,.R., Baree,.S., 4, Compuaion of Ship Responses in Waves Using Panel ehods, Journal of Naval rchiecure and arine Engineering, ol, No., pp. 35-46.