Design for Safety and Stability Henning Luhmann, MYR WRFT henning.luhmann@meyerwerft.de Jörg Pöttgen, MYR WRFT joerg.oettgen@meyerwerft.de ABSTRACT Safety and stability are two key asects for the successful design of shis while keeing the balance between efficiency and erformance of the shi. In the ast the main drivers for safety imrovements have been catastroc accidents but a change of mind is needed to enhance safety and stability within the given enveloe of design constraints. This can only be achieved when beside comrehensive calculation tools basic design methods will be develoed and used in the daily design work. A method to redict the attained subdivision index has been develoed and has been shown here as an examle for a simlified design method. Keywords: design, safety, cruise shis, stability index 1. INTRODUCTION The design of comlex shis, like cruise shis, is an everlasting quest to find the right balance between the erformance of the shi, for cruise shis this is the satisfaction of the guests on board, efficiency of oeration and safety and environmental rotection. Obviously the comliance with rules and regulations are the basis for each design, but the develoment of technologies and new design ideas challenging the alication of regulations. 2. DSIGN TO SAFTY Shibuilding and design of shis has a very long tradition and is mainly built on exerience. Main drivers for design changes towards a safer shi have been in the ast mainly accidents or near-accidents and exeriences of the designers as well as oerational feedback. Very oular examles are the casize of the VASA, the sinking of TITANIC or the foundering of STONIA. In the ast such kind of accidents also influenced the rule making rocess and based on the IMO rules the current state-of-theart has been defined. Merchant shis are designed, built and oerated to be art of an enterrise to generate rofit. This main objective together with the challenge to find the right balance with rules and regulations is usually the motivation not to design to safety but to squeeze the rules and their interretation to the limits and maximizing the rofit for shibuilder and oerator. By maximising the nominal caacity of a shi and designing the shi for the date of delivery only by ignoring the life time of the shi and the oerational needs the strategy for design will fail on the long run. A change of mind is needed for the whole industry to maximize the safety within the given enveloe in close cooeration with the oerator and for the life-time of the shi. Another imortant factor for the design rocess is the available time. Decisions influencing the global safety of a shi, like the watertight subdivision, are defined at an very early stage of design and needs to be ket unchanged until delivery. Hence, the methods you may aly to determine the safety needs to be fast and robust. Comlex tools like arametric otimizations may be used from time to time to exand the level of exerience but they are un- 17
suitable for the daily design work. The industrialization of outcome of research rojects is very imortant to take new technologies into use, but it also worth to reconsider exeriences and knowledge from the old days. 3. STABILITY RLATD TOPICS FOR DSIGNRS There are many different toics which may influence the stability or general safety of a shi which needs to be considered during the design. The following figure illustrates a ossible accident scenario. Figure 1 Accident Scenario Although the best way to imrove the safety is the revention of any accident the focus of most of the designers and researchers is the mitigation of any accident. In articular the extensive discussion about stability after flooding during the recent years, which is still ongoing, is leading somehow in the wrong direction. In the daily work of shi designs some basic elements like a accurate estimation of light weight and centre of gravity is much more imortant than a fancy flooding simulation. Proer weight and COG estimations together with the reasonable account for future growth and service based loading conditions form the basis for the hull form and thus the stability behaviour of the shi during its life time. The constant verification of weight and intact stability, including dynamic stability behaviour, ensures that the shi will meet the requirements from the regulations as well as for the erformance. The detailed investigation for stability after flooding is the second focus during the design. To find the best subdivision is again a huge iterative rocess to align the different demands of sace requirements, oerability and survivability after damage. Also other safety rules, like escae routes are challenging arameters in this rocess. As exlained before this needs to take lace within a very short time frame and the following resentation of a method to judge on the damage stability caabilities for different hull forms in an easy way is a good examle how modern first-rincile tools together with basic knowledge can be combined to form a owerful design tool. During the develoment of a new hull form it was recognized, that the normally used hard oints for the hull form designer will not reflect all different demands a hull form has to fulfil. Therefore an algorithm has been develoed to comare different hull forms under secial interest of the demands of the damage stability calculation. 4. DSIGN OF A NW HULL FORM During the design rocess different hull forms are develoed to find the best for the given design. Hard oints for the hull designer are defined to reflect any constraints, which are the following: Geometry o L o Bmax o Design draught Hydrostatics o Minimum KM on design draught o LCB o Dislacement 18
A new kind of hard oint has been searched for the hull designer that guarantees the same level of the attained index. 4.1The Stability nergy Index The fundamental idea was formulated by RAHOLA already in 1923. He invented the stability energy of a vessel which was used for the stability rating of different vessels. Based on these rinciles the following algorithm was develoed. Contributing Factors The area under the righting lever arm curve is calculated from the uright to a certain range of heel. This area is been called. To reflect the influence of the damage stability calculation is only calculated for the design draught of a vessel but for all three draughts relevant for the calculation of the attained subdivision index: Light service draught (D l ) Deeest subdivision draught (Ds) Partial subdivision draught (D ) Figure 1 Area under the GZ curve comared with the Attained Index A i As the shi is not floating on the three initial draughts after damage anymore, an additional draught has been considered to reflect the situation of the vessel after flooding. This over draught (D o ) is the deeest subdivision draught D s lus 40% of the difference between D s and D l. In addition a weight factor 0.5 for D l is used to adjust for the minor influence of this draught. Figure 2 show the imrovement driven by these decisions. Basic Calculations A variation of different hull forms with the same KG on the different draughts is calculated according the above mentioned rinciles. The watertight subdivision for the calculation of the attained index has been the same for all four hull forms. The below diagram show the resulting attained index in comarison with the comuted area under the GZ-curve from uright to 22 of list. Figure 2 Area under the GZ curve comared with the Attained Index A i with an additional draught Do Calculation Rule for the Stability nergy Index Based on the findings an easy algorithm for the hull form designer has been develoed to verify if his hull form will reach the Stability nergy 19
Index and to calculate the Required Stability nergy Index as a hard oint for the hull for designer based on a given Attained Index reached in the damage stability calculation The hull form designer will get the draughts D l, D, D s and D o with their corresonding KG values. For each draught the corresonding area under the GZ curve has to be calculated from 0 to 22 list and summed u according the following formulae. 0,78 0,76 0,74 0,72 0,70 0,68 0,66 0,64 0,62 0,60 Real Se Samle Shi B (seshi=1.00) Calculated Se Samle Shi B (seshi=0.98) A B C D F S with : l s o 0.5 ( D ; KG ;0 22 ) ( D ( D l l S ; KG ; KG ;0 22 ) ;0 22 ) ( D ; KG ;0 22 ) o l S o s o [1] Stability nergy Index versus given Attained Index Based on further calculations a simle calculation rule for S at a given Attained Index could be derived statistically. S ( RAI ) 2 RAI [2] se shi with: RAI = Required Attained Index and se shi = correction factor for different shis [arox. 0.96-1.06 1 ] The following diagram shows the results by using the above introduced formula. For the same KGs and watertight subdivision the attained index has been calculated as well as the S indicated as the Real S in the diagram. A very good correlation has been found and with this rove this method has been used during arametric otimizations of hull forms resulting in the otimum comromise between hydrodynamic erformance, sace requirements and sufficient stability after flooding. Figure 3 congurence between the real and the calculated S 5. XAMPL DSIGN TO SAFTY One other examle for design to safety is the arrangement of watertight doors in a assenger shi. The sace below the bulkhead deck is subdivided into watertight comartments and on cruise shis, each square meter is used for the accommodation of the crew and technical saces like workshos and laundries or storage areas. ach of the watertight comartments requires two means of escae, one of them needs to be a vertical stair or escae leading to the embarkation deck, the second one is usually a watertight door leading into the adjacent comartment. If oerational needs are not considered in the right way at an early design stage the urose of the saces may cause that watertight doors are required to be oen during normal service and not only as an emergency escae. Tyical examles are the laundry and the connected linen stores. In the ast laundry and linen stores have been located in adjacent watertight comartments, but recent designs have shown that this can also be laced on to of each other. With this vertical flow the watertight doors may be ket closed during normal oeration and this really increases the safety level. 1 To be further investigated 20
6. RISK MANAGMNT AND FUTUR CHALLNGS of modern tools with old exeriences can be used in the daily design rocess. The safety related design rocess requires a high degree of transarency and close cooeration between the stake holders. Not only shiyard and oerator are required to cooerate, also the regulatory bodies, like flag administration and classification societies, and technical exerts need to be art of the team. This aroach has a number of ositive effects. One is of course that the design is of outstanding quality, usually with a roven higher safety level than required by the rules and regulations, on the other hand the lack of knowledge about the secial challenges for large cruise shis can be communicated in a better way to a wider audience. A basic challenge however remains new designs and also new rules and regulations imrove the safety of new shis significantly in a continuous way, however it takes about 30 to 40 years to get a whole fleet renewal. The question how to ugrade the safety of the existing fleet is one of the major tasks for the industry and the regulatory bodies in the coming years. Otherwise the ga in safety level between old and new shis will become unaccetable. The introduction and quantification of active safety measures may be one ossible way to solve this roblem. 7. CONCLUSIONS Shi design always focus on safety and stability, however instead of interreting gien rules and regulations to their limits a change of mind is needed to maximize safety within the given design constraints. A roer holistic aroach based on close cooeration between regulators, designers and oerators is the way ahead, while using highly sosticated calculation tools together with exerience and traditional simle design methods to avoid the reetition of mistakes which have haened in the ast. A method has been shown how this combination 21
This age is intentionally left blank 22