BOWFLARE IMPACT LOADS ON CONTAINERSHIPS

D. Bajič, Lloyd s Register EMEA, Trieste, Italy J. Prpić-Oršić, A. Turk, Faculty of Engineering, University of Rijeka, Croatia BOWFLARE IMPACT LOADS ON CONTAINERSHIPS Summary Bowflare slamming describes dynamic wave impact on the bow side shell structure above the design waterline. During water entry the bow structure is subject to high pressure loads which sometimes lead to local damages that usually do not affect the survivability of the ship. Recently, the number of large container ships is increasing. In order to increase the convenience for cargo loading for containers and pure car carriers, the recent tendency has been to widen the bow flare angle. This raises the flare slamming pressure excessively and causes structural damages. The problem is recently exacerbated by modern trend to "drive through" bad weather. The bow flare slamming impact on such ships was investigated by applying contemporary rules. For practical purposes, the values of these pressures are usually obtained by simplified procedures suggested by classification societies. This paper addresses some basic aspects of the problem such as the influence of the design parameters as well as the problem of the uncertainty of the results. The comparison of various approaches suggested by different classification societies suggests significant difference in estimated values of pressure. The objective of the research is also to improve our understanding of the background of the various classification societies approaches related to bowflare slamming pressure estimation. Key words: bowflare slamming, slamming pressures, containerships OPTEREĆENJA IZBAČENE FORME PRAMCA NA BRODOVIMA ZA PRIJEVOZ SPREMNIKA Sažetak Udaranje izbačene forme pramca dinamički je udar bočnog dijela konstrukcije pramca iznad vodne linije o morsku površinu. Tijekom ulaska u vodu pramčana je konstrukcija izložena velikim opterećenjima koja ponekad mogu dovesti do lokalnih oštećenja koja uobičajeno ne ugrožavaju preživljavanje broda. U posljednje se vrijeme broj velikih kontejnerskih brodova značajno povećava. U cilju što boljeg iskorištenje prostora za smještaj tereta na novim se brodovima za smještaj kontejnera i automobila povećava kut izbočene forma pramca, što ima za posljedicu veća udarna opterećenja te lokalna oštećenja konstrukcije. Problem je dodatno izraženiji i zbog trenda "vožnje kroz" loše vremenske uvjete. U radu je udarno opterećenje na izbačenu formu pramca analizirano primjenjujući trenutno važeća pravila klasifikacijskih društava. Za praktične svrhe, vrijednosti tlakova uobičajeno se dobivaju pojednostavljenim postupcima prema preporukama klasifikacijskih društava. Ovaj rad obrađuje neke od osnovnih aspekata spomenutog problema kao što su utjecaj projektnih parametara te vjerodostojnost dobivenih vrijednosti. Usporedba različitih pristupa klasifikacijskih društava ukazuje na značajnu razliku izračunatih vrijednosti udarnih tlakova. Cilj istraživanja je i unapređenje razumijevanja pozadine postupka procjene tlakova uslijed udara izbočene forme pramca pojedinih klasifikacijskih društava. Ključne riječi: udaranje izbačene forme pramca, udarni tlak, kontejnerski brodovi

D. Bajič, J. Prpić-Oršić, A. Turk Bowflare impact loads on containerships 1. Introduction In order to increase area available for cargo loading, in particular on container and Ro- Ro ships, the recent trend is to widen deck forward. Consequently, bowflare angles are lowered. Bowflare angle is intended as an angle in a transverse section between side shell and horizontal axes. This raises the flare slamming pressure excessively and may cause structural damages. It has been observed that such ships often suffer damages in the fore end region while sailing in heavy weather conditions. It has been realized that damages are caused by dynamic wave impacts in these regions due to excessive ship motions. The problem is recently exacerbated by modern trend to "drive through" bad weather particularly in conjunction with high speeds. The major problem in this respect is to accurately predict values of these pressures in order to determine adequate reinforcements in the bow area. Various classification societies have developed in the past various sets of Rules. These Rules have been created in a different ways combining analytical techniques with results of direct calculations all calibrated with damages experienced in service. Intention of this paper is to investigate differences in their approaches in terms of predicted pressures in a bow flare area as well as sensitivity to different input parameters. Lloyd s Register (LR), American Bureau of Shipping (ABS), Bureau Veritas (BV), Germanischer Lloyd (GL), Korean Register (KR) and Det Norske Veritas (DNV) have been chosen for the purpose of this investigation. 2. Bowflare slamming phenomena As shown in Figure 1, the problem of bowflare slamming is the problem of a symmetrical body penetrating the water surface. The y and z axes are taken along the undisturbed free surface and along the body s centreline pointing upward, respectively. The problem corresponds to the vertical motion of a hull section in heading or following waves, in addition, the vertical speed V(t) in the figure can be considered the body s relative velocity with the water surface of incoming waves at the target section. The fluid domain is surrounded by the boundaries consisting of the free surface, the body surface, the centreline, the side walls, and the bottom. Assuming the fluid is incompressible and the flow is irrotational, the fluid motion is specified by the velocity potential. z V(t) WATERLINE y Fig. 1 Bowflare slamming Slika 1. Udaranje izbačene forme pramca

Bowflare impact loads on containerships D. Bajič, J. Prpić-Oršić, A. Turk Traditionally the water impact problem is applied to the slamming analysis. So far enormous works have been done in this respect (e.g. Korobkin, (1996), Faltinsen (2005)). The water impact approach is classified roughly into analytical and numerical. The analytical approach is based on mathematical models. Although the Wagner theory is well known, the recent studies are devoted to deal with the slamming problem of the arbitrary shaped body (Korobkin (2005), Takagi et al. (2007)). The numerical approach is represented by the boundary element method (BEM). A certain level of success was already achieved previously, but the important progress has been actuated when the numerical simulation of the impact jet is realized, Sun et al. (2007). Moreover, the problem additionally become more complex because of stochastic sea nature and the problem of estimation of worst weather condition which the ship can suffer during exploitation. However, in practice the problem of design slamming pressures are usually obtained by empirical formulae given by classification societies. These formulas are simplified and take into account simplified form of ship defined by basic parameters (length, beam, draft, block coefficient etc.). The bow form is defined by relevant angles α (angle between axes x and section inclination in vertical longitudinal plane), β (angle between axes y and section inclination in vertical transversal plane) and γ (angle between axes x and section inclination in horizontal plane) shown at Fig. 2. z α β x y γ x Fig. 2 Relevant angles Slika 2. Relevantni kutovi The procedures proposed by classification societies are based on performed direct calculations and correlated by service experience. Consequently, these formulas are not necessarily valid for novel designs.

D. Bajič, J. Prpić-Oršić, A. Turk Bowflare impact loads on containerships 3. Classification societies requirements 3.1. Applicable ship type LR has two sets of relevant Rules (Lloyd s Register, (2009)). First one contemplated in the Part 3, Chapter 5 are intended for ships with moderate block and moderate speed hull forms. Since these requirements were not adequate for high speed hull forms with more significant bow flare angles a revised bow flare slamming pressure requirements were introduced in the Part 4, Chapter 2 for passenger ships, Ro-Ro ships, ferries and container ships. ABS has different sets of Rules (American Bureau of Shipping, (2010)) for oil tankers contemplated in the Part 5-1-3, bulk carriers contemplated in the Part 5-3-3 and Container ships contemplated in the Part 5-5-3. DNV has unified Rules (Det Norske Veritas, (2009)) contemplated in the Part 3, Chapter 1, Section 7, para E300 intended for all ship types. Normally only ships with well rounded bow lines and or flare will need strengthening. As regards BV (Bureau Veritas, (2009)), relevant requirements are set in the Part B, Chapter 9, Section 1, paragraph 4. These requirements are intended for ships other than passenger ships which makes them applicable to container vessels. GL (Germanisher Lloyd, (2009)) has separate requirements for container ships contemplated in the Chapter 5, in particular as regards bow flare slamming in the Section 4 B paragraph 2.2. As to KR (Korean Register, (2009)), relevant requirements can be found in the Part 3, Chapter 8, Section 1, paragraph 108. Basically these requirements are set for pure car carriers but may be considered applicable to container ships as well. 3.2. Applicable bow region ABS requires investigating bowflare slamming in the region above the waterline between 0.0125 and 0.25 Rule length (L) from the forward perpendicular (FP). As per DNV requirements, the effect of bow impact loads is in general to be evaluated in the region forward of a position 0.1L from FP and above the summer load waterline. On the other hand LR requires strengthening shell envelope against bow flare typically over the fore end side and bowing structure above the waterline in the areas where the hull exhibits significant flare. BV requires to reinforce bow area that is extending forward of 0.9 L from the aft end of L and above the summer load waterline up to the level at which a knuckle with an angle greater than 15 degrees is located on the side shell. As per GL, bow flare area to be specially investigated is that forward of 0.1 L behind F.P. and above the ballast waterline. Similarly, KR states bow flare positions considered to endure large wave impact pressures are those above the load line for 0.1 L forward. 3.3. Selected hull form Considering content of chapter 3.1. above, the container ship has been chosen as a referent ship type for this comparison. The main characteristics of chosen hull form and relevant bow sections are shown at the Fig. 3. As far as content of chapter 3.2. is concerned, this investigation has been limited to the area forward of 0.1L from FP above waterline. Selected frames 312, 320, 328, 336, 344 and 352 are located at approximately 91, 93, 96, 98, 100 and 103% of length from aft end

Bowflare impact loads on containerships D. Bajič, J. Prpić-Oršić, A. Turk respectively. Pressures have been calculated at each frame for three different levels, i.e. at 11 m (Level 1), 13.8 m (Level 2) and 16.6 m (Level 3) above base line. L = 240 m Breadth = 36 m Draught = 8.5 m Block coefficient = 0.71 Service speed = 22 kn Fig. 3 Bow form Slika 3. Forma pramca 4. Bowflare slamming pressure estimation For selected positions six different Rule calculation procedures have been applied and following values of bowflare slamming pressures, shown in Fig. 4, have been obtained. For the time being analysis is limited to pressure estimation only. Previous work (Bajič&Prpić-Oršić, 2009) has shown that, with respect to thickness and at least considering ABS, DNV and LR, formulas for scantlings determination were quite similar. Therefore, conclusions made only on the basis of the pressures were still approximately valid for the final scantlings too. However, intention is to continue with this analysis in the future and also elaborate scantlings in terms of plating thickness and section modulus of primary and secondary members. As can be seen from the Fig. 4, BV, DNV and LR values are roughly similar. GL and ABS values, apart from few points, are significantly lower especially where bow flare angles are lower. Great oscillations in calculated pressures have been noted in case of KR. As explained later, these oscillations are mainly the function of the oscillations of hull form coefficients.

D. Bajič, J. Prpić-Oršić, A. Turk Bowflare impact loads on containerships Fig. 4 Pressure distribution over decks Slika 4. Razdioba tlakova po palubama 5. Bowflare slamming formulas decomposition Generally, slamming pressure is considered as a proportional to the square of the relative velocity v r between the wave and ship bow at the instant of impact expressed in broad terms as

Bowflare impact loads on containerships D. Bajič, J. Prpić-Oršić, A. Turk where ρ is a water density and k is a non-dimensional hull form coefficient. In the following paragraphs, for various classification societies, calculated pressure has been decomposed in order to extract relative velocities and hull form coefficients as estimated by each society. Furthermore, influence of each on these two parameters on the final value of calculated pressure has been analysed. ABS has not been included since in this instance it was not feasible to extract above mentioned two parameters from pressure estimation formula. As regards to LR, in order to simplify the analysis, part of the pressure caused by forward ship speed is neglected since it generally constitutes for only up to max 3% of total pressure. (1) 5.1. Hull form shape coefficient When comparing hull form coefficients determined by classification societies with theoretical ones a following can be exhibited from the Fig. 5 Hull form coefficients as determined by LR are very well corresponding with those predicted by Von Karman s theory. On the other hand, coefficients as estimated by BV are close to those predicted by Ochi s theory. DNV appears to follow also Ochi s theory but with some additional safety margin. This may be expected since coefficients foreseen by the societies can usually include some safety factors obtain by calibration process on the damaged ships. No correlation has been found in case of GL while large oscillations are present in the case of KR. As seen from the Fig. 6, GL generally have the lowest coefficients. LR values are slightly above. Greater values are used by BV while the greatest values of hull form coefficients are those foreseen by DNV. As said previously, large oscillation of hull form coefficients have been noted in case of KR. As per separate analysis this phenomena is mainly caused by coefficient K p that has large step when corrected flare angle reaches 18 degrees. Fig. 5 Hull form coefficients Slika 5. Koeficijenti forme trupa

D. Bajič, J. Prpić-Oršić, A. Turk Bowflare impact loads on containerships Fig. 6 Hull form shape coefficient over decks Slika 6. Koeficijent oblika forme po palubama 5.2. Impact velocity estimation With respect to relative impact velocities, diagrams in Fig. 7 show less difference between various classification societies when compared to differences found for hull shape coefficients. In case of velocities, the lowest values are those estimated by KR. Slightly higher are those calculated by BV and DNV having the same formula for relative impact velocity. GL in this case has a constant value generally slightly above BV and DNV estimation. The highest values are those as calculated by LR. Both, KR and in particular LR

Bowflare impact loads on containerships D. Bajič, J. Prpić-Oršić, A. Turk show trend of increase of relative velocity towards the fore end that is perfectly to expect since these velocities mostly influenced by pitching. Fig. 7 Impact velocity over decks Slika 7. Udarna brzina po palubama

D. Bajič, J. Prpić-Oršić, A. Turk Bowflare impact loads on containerships 6. Sensitivity study Differences found in the comparison of bowflare pressure values are significant. Analyzing the different empirical formulas as proposed in the Rules by each classification society it is not possible to get insight into the background of theirs procedures. For example LR requirements originally were developed from a comparative study carried out on cargo ships which experienced fore end damages from heavy weather and on undamaged ships having a similar configuration and service speed (to ensure that strengthening for bow slamming was not carried out unnecessarily). This method was further calibrated using general cargo ship type hull forms, i.e. moderate block and moderate speed hull forms. Later, it was found that these requirements were not adequate for high speed hull forms with more significant bow flare angles. For this reason a revised bow flare slamming pressure requirement was introduced for passenger ships, Ro-Ro ships, ferries and container ships. This method is based on the theoretical approach using ship motions correlated with service experience. The summary of input parameters for calculation of slamming pressures has been listed at Table 1. Therefore, at this stage, it was decided to examine the differences in the results through influence of various input parameters. Only parameters required by all societies are considered. Among them ship speed V, distance of considered point from waterline z P (calculated as difference of distance of considered point from baseline z P and draft T) and angle between side shell and waterline in a transverse section β have been selected as the most influential ones. Moreover, the influence of block coefficient C B and angle between axes x and section inclination in horizontal plane, γ, on pressure values have been examined. It is important to point out that medium increment of pressure is calculated as an average for the entire bow region. The results are shown in the Figures 8 to 12. respectively. Table 1 Input parameters Tablica 1. Ulazni parametri LR ABS DNV BV GL KR L C B - - V T - β γ - α - - - - - x P - - - z P - - - - - - Stem angle Half breadths - - - - - B - - - - D - - - - - As regards ship speed, lowest influence has been noted in case KR. Slightly larger and very similar dependency has been noted in case of LR, DNV, BV and GL. The biggest influence of speed has been noted for ABS.

Bowflare impact loads on containerships D. Bajič, J. Prpić-Oršić, A. Turk Fig. 8 Ship speed influence on slamming pressure Slika 8. Utjecaj brzine broda na tlak udara Fig. 9 Ship draught influence on slamming pressure Slika 9. Utjecaj gaza broda na tlak udara Apart KR, very small dependency on the distance of considered point from WL has been noted. Fig. 10 Angle β influence on slamming pressure Slika 10. Utjecaj kuta β na tlak udara

D. Bajič, J. Prpić-Oršić, A. Turk Bowflare impact loads on containerships Variation of flare angle showed generally to have large impact on the calculated pressure. This has been particularly noted in case of KR and LR, and partially GL. Fig. 11 Block coefficient influence on slamming pressure Slika 11. Utjecaj koeficijenta punoće na tlak udara Apart from KR, very small dependency on the block coefficient has been noted. Fig. 12 Angle γ influence on slamming pressure Slika 12. Utjecaj kuta γ na tlak udara Apart KR, very small dependency on the WL angle has been noted. 7. Conclusion This analysis has shown that various classification societies have very different approach when calculating bowflare pressures. Bigger differences have been noted in case of hull form coefficients while relative impact velocities were closer. Sensitivity analysis has shown that the most governing criteria in the pressure calculation is flare angle.

Bowflare impact loads on containerships D. Bajič, J. Prpić-Oršić, A. Turk The further investigation, still in terms of pressures, will be directed toward analyzing novel forms extreme bowflare slamming pressures by direct calculation and comparing it with values estimated by empirical formulas. REFERENCES [1] American Bureau of Shipping, 2006, Rules for Building and Classing Steel Vessels, Houston [2] D. Bajič, J. Prpić-Oršić, Comparison of various classification societies requirements, ASDEPP Workshop, Split, Croatia, 2010. [3] Bureau Veritas, 2009, Rules for the classifications of steel ships, Part D- Containerships, Neuilly-sur- Seine Cedex [4] J. Daidola, and V. Mishkevich, 1995, Hydrodynamic impact on displacement ship hulls, Ship Structure Committee SSc-385, Washington [5] Det Norske Veritas, 2004, Rules for Ships, Hovik [6] O.M. Faltinsen, 2005, Hydrodynamics of High-Speed Marine Vehicles, Cambridge University Press [7] Germanischer Lloyd, 2009, Rules for the classifications and construction of seagoing ships, Hamburg [8] A.A. Korobkin, 1996, Water impact problem in ship hydrodynamics, Advance in Marine Hydrodynamics, M. Ohkusu, Ed (Chap.7), Computational Mechanics Publishing, Southampton, Boston [9] A.A. Korobkin, 2005, Analytical Models of Water Impact, Euro. J. Applied Mathematics, 16, pp.1 18 [10] Korean Register, 2009, Rules for the classifications of steel ships, Seul [11] Lloyd s Register, July 2009, Rules and Regulations for the Classification of Ships, London [12] H. Sun and O.M. Faltinsen, 2007, Water Impact of Horizontal Circular Cylinder and Cylindrical Shells, Applied Ocean Research, Vol.28, pp.299-311 [13] K. Takagi and Y. Ogawa, 2007, Flow Models of the Flare Slamming, Proc. of International Conference on Violent Flows, pp.173-179