Você está na página 1de 90
WOLFSON UNIT solange Sore FOR MARINE TECHNOLOGY & fel: #44 (0)23 8088 5044 INDUSTRIAL AERODYNAMICS TREE RE SOE: Email: wumtia@soton seul. Report No. 1779 February 2005 MARITIME & COASTGUARD AGENCY Research Project 509 HSC — Evaluation of Existing Criteria Final Report Executive Summary ‘This report describes Research Project 509, a study of the intact and damage stability criteria applied by the 2000 High Speed Craft Code. The aims of the study were fo determine the relative levels of safety provided by the monohull and multihull criteria and, if appropriate, suggest alternative criteria that would provide equivalence. ‘The work comprised tests in waves on models representing a range of high speed craft. A large number of variables affect vulnerability to capsize, and the limitations of the project precluded comprehensive study of all permutations, Simplified models of two monohulls, three catamarans and a trimaran were constructed, ‘with hull forms representative of contemporary vessels. Almost 800 test runs were conducted on these 6 ‘models, in 53 different configurations of damage and loading condition. ‘Tests were conducted in regular waves, ata range of frequencies and heights, for all model headings, to determine the wave conditions that resulted in capsize. Numerous capsize mechanisms were observed and categorised into five principal types. For some model configurations the capsize mechanism was dependent oon the heading and wave frequency. For each model configuration the minimum wave height to capsize was identified. ‘The minimom wave heights to capsize were within the range of operational seastates for most configurations that just complied with the Code requirements. This indicates that the Code does not ensure an adequate level of safety. There have been few stability casualties within the HSC fleet but this is believed to be due to substantial margins of stability over the minimum requirements. ‘The relationship between vulnerability to capsize and various stability parameters indicated the range of stability to be of greatest importance, with the maximum righting moment of secondary significance. Conventional stability criteria, including those of the HSC Code, do not generally assess these parameters directly, but regulate other parameters that provide approximate measures of them. ‘The scaling of model test data highlighted the effects of scale in the application of conventional stability criteria. Constant minimum GZ values are applied for all sizes of vessel, and regardless of the severity ofthe anticipated environmental conditions. The recommendations arising from this project include the adoption of criteria that take into account the size of vessel and operational seastate, and a method has been proposed for validation and impact assessment by the industry. ‘Another problem associated with many conventional criteria, including some of those in the HSC Code, is that the stability ofthe upright vessel is regulated, but not necessarily the residual stability after the application of heeling moments, such as those associated with wind heeling or passenger crowding. University of Southampton WOLFSON UNIT ‘The HSC Code incorporates different approaches to the assessment of stability of monohull and multihull vessels, The effects of heeling moments are considered in different ways, using different parameters, and the residual stability is not necessarily addressed. This makes any comparison of the level of safety of vessel types difficult, because individual cases will differ, depending on the nature of the heeling moments and the relationship between them and the stability characteristics, ‘The neglect of scale and seastate by the HSC stability criteria also introduces problems in making general ‘comparisons between vessel types. Itis possible to compare the levels of safety of specific vessels but not {generic types. The proposed method of assessment is designed to overcome these problems and enable equivalent levels of safety to be set forall types of vessel, with due regard to the size of the vessel and the anticipated environmental conditions. ‘The method incorporates the finding from the tests that all vessel configurations may be considered together 1s floating bodies defined only by their size and residual stability characteristics, This applies regardless of ‘whether the vessel is @ monohull or a multihull, intact or damaged, upright or heeled. WOLFSON UNIT CONTENTS Introduction . Background and Objectives. Selection of Sample Vessels. Model Descriptions... Model Construction and Ballasting, Computer Calculation of Stability .. Issues Associated with Modelling and Assessment of Raking Damage ‘Monohull Raking Damage - Parametric Study.. Catamaran Raking Damag ‘Trimaran Raking Damage Test Facilit Test Technique. Representation of Heeling Moments. Consideration of Irregular Waves ‘Summary of Capsize Mechanisms... Test Results 5 Relationships Between Stability Parameters . Relationships Between Stability Parameters an ‘The Importance of Residual Stability... Comparison of Monohuil and Multibull Criteria... Comparison of Vessel Types Using Existing Criteria. Implications of the Results for HSC Code Revision. Non-Dimensional Stability Parameters and Wave Height to Capsize .. ‘The Role of Displacement ‘The Importance of Range of Stability. Development of a Safe Boundary Development of an Alternative Method of Assessment Consideration of Downflooding ... es Consideration of Wind Heeling... Considerations of Passenger Safety Implications for the HSC Fleet ‘Summary of Findings: Response to the Objectives, Conclusions . Recommendation: References aps WOLFSON UNIT 1 INTRODUCTION “This report describes Research Project 509, an evaluation of stability criteria in the 2000 HSC Code. The project followed broadly the programme of work outlined in Wolfson Unit proposal no.2561BD, dated 26* ‘November 2002. The contract was issued by the MCA on 10" January 2003. the study comprised model tests conducted in an attempt to quantify the relationship between ship stability properties and safety against capsizing, and to evaluate the levels of safety of high-speed monohull and rmultihull craft in waves. The aim of the project was to provide information for a review of the IMO HSC Code, through the appropriate subcommittee, ‘A suggested work programme was issued by the MCA but, as might be expected in a project of 2 years uration, its scope and details were reviewed and adjusted in recognition of the implications of the intermediate findings. 2 BACKGROUND AND OBJECTIVES ‘Annex 7 of the 2000 HSC Code contains prescriptive requirements for the stability criteria of intact and damaged stability of multihull craft in the displacement mode. These intact and residual stability requirements were drafted with a minimum of time and resources to suit the time scale of the development of the 1994 Code and were subsequently carried forward to the 2000 Code. No theoretical or experimental validation of the adequacy of these requirements is known to exist and so not known whether the criteria provide adequate levels of safety against capsize ot sinking, It is also not known whether these requirements give comparable safety to the prescriptive requirements for monohull craft. Appendix I gives a summary of the stability requirements of the 2000 HSC code. ‘The objectives of the project were outlined by the MCA in their invitation to tender. 1. Toestablish critical modes of capsize for typical monohull and multihull HSC designs when at rest in the displacement mode. 2, To establish the critical or limiting environmental conditions for which the existing multibull requirements provide just sufficient intact and damaged stability, and the characteristics needed to give equivalent level of safety to monohulls. To propose (if appropriate) revised statical stability criteria requirements for multihull era. To investigate the effect of these revised requirements on a selection of designs. ‘To draft the relevant papers for IMO. 3. SELECTION OF SAMPLE VESSELS al target was to test two monohulls to provide a benchmark level of safety, against which about five ‘multihull models could be compared. It was hoped that data could be obtained regarding the margins of stability of different types of HSC relative to the minimum criteria. Such information would have enabled any relationships between vessel parameters and stability margins to be identified, and highlighted the iteria that tend to be critical. Designers, builders and other authorities were circulated with a request, but information was not forthcoming, Few vessels had been constructed for compliance with the 2000 HSC Code, and designers typically are unwilling to release data on their designs. Data supplied by Fast Ferry Information Ltd. enabled a simple statistical analysis to be carried out on the size and type of vessels in the UK and world fleets. The analysis yielded very different statistics for the two WOLFSON UNIT fleets, Since the aim of the research is to provide guidance for the IMO it was considered that the world fleet statistics were more relevant to aid the selection of representative vessels to model, 3.2. Monohulls and Catamarans ‘There are nearly three times as many operational catamarans as monohulls and the bar graphs, presented in Figure |, indicate that the majority of monohull vessels are in the range 20 to 40 metres. This length also dominates the catamaran flet, although there are a substantial number in the range 40 to 60 metres. Figure 2 presents the fleet in terms of their principal characteristics against length and superimposed upon these graphs are the vessels represented by the models used for this study. Figure 3 shows the variation of passenger capacity with length, Whilst the number of passengers carried is highly variable, the trend lines indicate that the average passenger capacities for monohulls and catamarans are very similar. Figure 4 presents what limited data were available for the maximum permitted operational sea states. There appears to be a linear relationship between the significant wave height and the length of the vessel, with no clear difference between monohulls and multihulls arly in the project one designer offered information on a monohull and a catamaran, and these were used as the basis for the first two models, albeit with the lines modified to facilitate construction, As tests on these models progressed, additional models were selected to provide a diverse range of forms. 3.3. Hydrofoils ‘There are over 200 hydrofoil vessels in the world fleet, not included in the above statistics, but only 20 have been built during the last 20 years. Because of this trend it was decided not to model any hydrofoils specifically. When stationary in waves, however, itis likely that their behaviour will be similar to that of conventional high speed monohulls and, by default, results from this project should be applicable, 34 Trimarans “Although there are very few operational trimarans, current interest in these vessels suggests that this situation ‘may change within a few years. Previous work on capsize of multihulls, Ref 1, demonstrated their stability and motion characteristics to be quite different to monohulls and catamarans and so it was decided to construct a trimaran model for this project. 4 MODEL DESCRIPTIONS 4.1 Presentation of Model Parameters Six models were constructed for the tess. Although this was less than the initial target of seven models, they ‘were tested in a total of $3 configurations. Figure 5 to Figure 10 give the principal dimensions, body plan and damaged compartments for each of the models. Photographs of the models are presented in Figure 11 and ‘igure 12. The configurations were varied by changes in displacement, VCG, damaged compartment and ial list. They were given identification numbers, IDI to ID53, to facilitate references to them in the analysis process. Appendix 2 identifies the configurations and principal stability characteristics. GZ, curves are presented in Figure 13 to Figure 19. ‘Appendix 2 lists the configurations in order of 1D number. The displacement, VCG and stability data are given forthe full scale vessel on which the model was based. The stability parameters listed are for the residual stability after heeling moments were applied, and correspond with the GZ. curves presented, The critical criterion number refers to the numbers assigned to the various criteria as tabulated in Appendix 1 Where different criteria apply to passenger and cargo eraft, the more stringent passenger eraft criteria have been used. In some cases the configuration failed to comply with the criteria, and in others the model condition provided a margin of stability over the minimum criteria, Because the GZ. properties listed are for the residual stability with heeling moments applied, the values do not necessarily correspond to the critical criteria values, which may refer to the stability in the absence of moments. Furthermore, the GZarcas listed WOLFSON UNIT are the total areas for the range of positive stability, rather than the areas addressed by the Code, which may be limited to a smaller range of angles. 4.2 Monobull 1 ‘The hull was a wide, shallow, hard chine form based on a vessel of 41 metres overall at a scale of 1:25. See Figure 5, The actual model length represented 38.8 metres because jet drives and stern platform were excluded, as was the case with other models. It was tested at two displacements, the lighter one being typical for vessels ofthat size, and the heavier one representing the upper boundary based on the flect statsties. For some of the tests a simple open topped box superstructure was added, Tests were conducted with the model intact and with side damage. Intact tests were carried out with varying initial angles of heel. This model was tested in twenty one configurations. 4.3 Monohull 2 ‘The hull was a double chine hull form based on a vessel of 96 metres overall ata scale of 1:40, See Figure 6, Twas tested in six configurations, intact and with side damage, all with high bulwarks equivalent toa first tier of accommodation. A deck was fitted at main deck level but the upper deck was not incorporated. When ‘a compartment was flooded, the bulwarks were also open to the sea, allowing water to wash onto the main deck, which represented an open vehicle deck without bulkheads. 44° Catamaran 1 ‘The hull was a round bilge symmetric eatamaran of moderate beam, based on a vessel of 32.7 metres overall at a scale of 1:20, See Figure 7. It had no superstructure or bridge deck, the hulls being connected with a pair of rectangular tube cross beams. Four configurations were tested: intact, with side damage, and with simulated 100% raking damage. The raking damage was achieved by adding an appropriate weight of ballast at the calculated centre of buoyancy. 4.5 Catamaran 2 Catamaran 2 used the same hulls as Catamaran 1 but the cross beams were extended to increase the overall ‘beam by 33%, from 8.32 metres to 11.07 metres, See Figure 8. Two intact configurations were tested in addition to eight side damage configurations 4.6 Catamaran 3 Catamaran 3 was a wave piercing form based on a vessel of 81 metres overall at a scale of 1:40. With jet drives and stern platforms excluded, the model represented a length of 70.8 metres. It incorporated an integral bridge deck and bow pod, which were kept watertight in all configurations. See Figure 9. Five damage configurations were tested, one with an open topped, full width, box superstructure added. The model was not tested intact because an extremely high centre of gravity would have been required to achieve low stability. This would have been impractical on the model and unrepresentative. 4.7 Trimaran 1 ‘The trimaran, with an integral bridge deck and sponsons, represented a vessel of 53 metres overall at a scale of 1:25. The model was tested at two displacements, Each sponson was capable of supporting 65% of the displacement at a total displacement of 175 tonnes, and 38% at 300 tonnes. See Figure 10. It was tested in seven configurations, two intact, four with damage to a sponson and one configuration with simulated 100% sponson raking damage, When the sponson was damaged, the bridge deck remained intact 5 MODEL CONSTRUCTION AND BALLASTING 5.1 Hull Design For each model a lines plan was drawn by the Wolfson Unit, having regard to hull forms and dimensions typical of high speed craft of the size to be represented. The fleet data were used to select a spread of sizes, and displacements. Whilst each model was based on a particular size of vessel, the data can be re-scaled to represent any size of vessel. The re-scaling process can be used to change the stability, of the new full size vessel, to any ratio of the code criteria. 5.2. Construction ‘The model hulls and bulkheads were constructed from G.R-P., using the minimum thickness commensurate with the structural requirements, typically 2mm, This light construction maximised the quantity of moveable WOLFSON UNIT ballast and minimised the stability contribution due to the hull shell in way of flooded compartments. The ‘decks were constructed of Perspex to enable the internal compartments to be checked for ingress of water due to leakages. Rectangular damage openings were provided in the hulls and decks, together with panels to close them when not in use. In general the size of the openings corresponded to the damage extents specified in the Code, although in some cases smaller openings were used for practical reasons. Prior to testing each configuration, all removable hull panels and decks were sealed with silicone sealant or tape and tested for watertight integrity. The construction of some models was complex however, and perfect sealing proved virtually impossible. Small drain holes with rubber bungs facilitated drainage of any ‘unwanted water from intact compartments. 5.3 Ballast ‘The models were adjusted to the required displacement and centre of gravity using movable ballast weights. In order to achieve the required VCG, and to facilitate fine adjustment, some ballast weights were mounted ‘on masts comprising threaded studs. In some cases the high location of ballast meant that the roll, pitch and. yaw inertias may have been unrepresentative. Only a small minority of capsizes appeared to be the result of a ‘gradual increase in roll angle due to synchronicity between wave petiod and natural rol period where rol inertia would be important. In these cases, where roll inertia is important and it was modelled incorrectly, it ikely that capsize will sil have occurred but a a different wave period. ‘The models had adjustable, vertical, pointed screws fixed at the fore and aft ends to facilitate swinging the model to adjust the ballast to a desired vertical centre of gravity location. A knife edge was used for determining the longitudinal centre of gravity location. Ballast was moved across the deck transversely to ensure the model was at the correct hee! angle prior to ‘each test when the model was floating, ‘The desired initial hec! angle was measured using a digital inclinometer accurate to 0.01 degrees. ‘The angle of vanishing stability, calculated previously, was also ‘checked at this siage by heeling the model by hand and measuring the angle with the inclinometer. Several attempts of just capsize to just not capsize enabled assessment of this angle to within 4 degree. 6 COMPUTER CALCULATION OF STABILITY 6.1 Calm Water ‘The stability calculations were carried out using Wolfson Unit software. Using a given displacement and centre of gravity location the stability was determined for a range of heel angles. At each heel angle the equilibrium waterline was determined with the vessel free to trim. In the damage configurations, the equilibrium heel angle was calculated. The specific gravity of water was assumed to be 1.025. An allowance ‘was made to the model displacements to take account of the difference in specific gravity of sea water and the water in the test tank, ‘The calculated static stability curves for all 53 configurations tested are presented in Figure 13 - Figure 19. ‘The curves represent the model at full scale. In the damage configurations the defined compartments ‘modelled those in the ‘as built” models. The models had bulkheads and deck stiffeners but no other outfit, so their permeability was close to 1.0, The models as tested, and stability curves as calculated, therefore had ‘more floodwater when damaged than would be the case for a typical full scale vessel at sea. This is particularly so in the case of engine room damage, where its permeability is normally assumed to be 0.85. Great care had to be taken to ensure that the computer model corresponded to the physical model in all respects, Early tests highlighted a number of possible sources of error, when discrepancies were measured between the calculated equilibrium heel angle or the range of positive stability, and those measured on the ‘model. This test programme, where very marginal stability characteristics were used, highlighted the WOLFSON UNIT potential for such errors, but the same possibilities exist with model tests on any vessel, particularly where damage cases are studied. The following items were taken account of: 1. The model dimensions were checked against the drawing and the computer model adjusted to correspond to the model if necessary. 2. Models were constructed to moulded lines and so no shell thickness was added to the computer model for the full scale vessel. 3. All flooded compartments were defined to the inside of the model hull and deck, so that correct, account was taken of the buoyancy of the structure, 4, All structural components, including bulkhead thicknesses, deck stringers, and deck beams were defined in the computer model. 5. Ballast weights were confined to the intact compartments or to locations above the waterline so that they had no buoyant contribution in addition to that included in the computer definition, 6. Permeability of all flooded compartments was taken to be 100% in the calculations to correspond to the model. 6.2 Waves ‘The stability was also calculated for Monohull 2 in waves. These were static calculations assuming the vessel stationary on a trochoidal wave down the length of the vessel. Again, for each hee! angle, an equilibrium waterline was determined given freedom to pitch and heave. The resulting stability curves are shown in Figure 20, for the vessel positioned on a wave height and length corresponding to a point on the capsize ‘boundary, that is at which the model just capsized in stern seas. 7 ISSUES ASSOCIATED WITH MODELLING AND ASSESSMENT OF RAKING DAMAGE Modelling such damage is complicated by the extent of venting of the compartments. If damage is restricted to below the waterline and a compartment is effectively sealed, there may be little or no flow of water through the damage opening. If damage is extensive, or if a compartment isnot ai tight, flooding will ake place up to the equilibrium waterline or the top of the compartment, which ever is lower. The presence of large open vents, doors or hatches to the compartment will increase the likelihood of water flowing through the damage opening in response to wave action and vessel motions. Raking damage might result in double bottom compartments becoming pressed full, so that the stability is equivalent to the intact stability in a different loading condition. When considering such a case it must be ‘understood thatthe stability is defined by the righting moment of the vessel, and the GZ. curve is determined by dividing the righting moment by the displacement. It is the convention in stability assessment that the displacement used is the intact displacement of the vessel. Whilst the righting moment of the vessel would be the same whether the double bottom were filled with water ballast or flooded as a result of damage, the GZ. ‘values for the latter would be greater because the displacement used would be lower. The range of residual stability, however, would remain unchanged, See Figure 16, ID 31 and 31 1D3I and ID3ta represent the same case treated as raking damage, and intact with a double bottom full of ‘water ballast, respectively. When assessing damage, the criteria are less stringent than for intact conditions, and so in this instance the raking damage case will benefit from a combination of apparently greater GZ, values and lower requirements. 8 | MONOHULL RAKING DAMAGE - PARAMETRIC STUDY 8.1 Objectives ‘To examine the effects of raking damage on monohull stability, calculations were conducted on a simple parametric series, derived for the single chine hull form of model Monohull 1. ‘The aim of this study was to determine what damage extents might be most onerous in terms of stability, and whether it might be necessary to conduct model tests with raking damage. WOLFSON UNIT 8.2 Hulls and Damage Configurations ‘The beam was varied maintaining all other dimensions constant, These beam variations represent a 25% increase from the narrow to the mid width hull, and a further 25% increase to the wide hull. Calculations were carried out for two intact draughts. For each beam variant the draught was kept constant and the displacement adjusted accordingly. Five damage cases were considered! 1. Raking damage to the double bottom forward, extending from the forward end of the waterline to 55% LWL aft 2. Raking damage to the douile bottom af, extending from the transom to 35% LWL forward, 3. Asin2, but with no double bottom in the engine room, so that the engine room was free to flood to the fall depth. 4, Raking damage to the double bottom over its full length. 5. Asin4, but with no double bottom in the engine room, so that the engine room was free to flood to the fall depth, In each case the double bottom height was assumed to be level with the intact waterline, The engine room ‘was assumed to extend from 10% to 35% LWL forward of the transom. Petmeabilities assumed were 85% in the engine room and double bottom beneath it, and 95% elsewhere. For simplicity, the extents of flooding ‘were taken as equivalent to the damage extents specified in the Code, In practice, with some bulkhead locations, the extent of flooding could be significantly greater, but that possibility is within the control of the designer who would locate bulkheads to maintain compliance. The use of 35% and 100% flooding therefore represent the extremes of flooding extent, ‘The KG used in each case was the maximum allowable for compliance with the intact stability criteria, The ‘weather criterion and the effects of wind heeling were not included because superstructures were not considered. ‘The hull dimensions are presented below as non-dimensional, together with the KG used in each case and the intact stability criterion that was critical in its determination, Light Draught Deep Draught Narrow | Mid Wide | Narrow id Wide width Width 677 Sal 433 677 5.41 433 1.06) 1.06 1.06 1.05) 105 1.05 Tat 144 14a 144 Lad 144 35.2 35.2 35.2 25.8 25.8 25.8 441 5.51 6.89 3.35 4.19 5.24 948 759 607 508 406 325 0.465 0.479 | 0.524 [0.469 [0.483 0.509 ‘Area ‘Area | Angle of | Area ‘Area ‘Area Criterion 30%-40° | to 30%r | GZmax | to 30%or | to 30%or | 30%. 40° GZmax GZmax_| _GZmax 8.3 _ Stability Calculations Stability curves are presented in Figure 21 to Figure 26 for the calculations, with a single page for cach hull ‘and draught variant. On each page are five sets of curves, one for each damage scenario, Each set of curves includes the intact stability curve as a reference. WOLFSON UNIT ‘The stability of the vessel in the equilibrium condition represents flooded level with the sea, orto the top of the compartment if that is below the external waterline, ‘The stability curves for three intermediate stages of flooding are included. These were calculated assuming 25%, $0% and 75% of the volume of floodwater contained in the vessel when upright in the equilibrium condition. Unlike the equilibrium cases, this volume remains constant at all heel angles. ‘There is a fundamental difference between the equilibrium and intermediate stage calculations. The equilibrium calculation represents the stability assuming that the floodwater is able to flow in or out of the compartments at such a rate that itis always in equilibrium, whatever the angle. This is representative of a gradually changing situation, such as when passengers move across the deck and a nev equilibrium angle is reached. The intermediate stage calculation assumes that the volume in the compartments remains ‘unchanged as the vessel heels, and is more representative of a small damage opening, or a rapid change of hoc! angle as in a rolling vessel. A real situation may Iie between these assumptions, such as when struck by «wind gust for example. ‘At small angles of heel, the intermediate stages generally indicate lower stability than the equilibrium cases. ‘At higher angles, in most cases, they indicate greater stability than the equilibrium case but the volume of floodwater and flooded draught are greater. The curves for intermediate stages at high angles therefore may not be particularly representative of a raking incident, where the damage is likely to be extensive and flow rates relatively high. 8.4 Summary of Results In most cases the stability in the damaged equilibrium condition was greater than in the intact condition, The exception, for all hulls, was the case for 35% raking aft with the engine room free to flood. In the deep draught cases there was a tendeney for the initial stability to be negative in some intermediate stages of flooding, so that the vessel would be likely to take up a temporary angle of loll during the flooding process. In some cases this was about 10 degrees, In the light draught condition a high GM value was ‘maintained throughout. ‘The lowest stability was exhibited by the mid width hull atthe deep draught, with 35% raking damage and the engine room free to flood. In the equilibrium condition the GZ reached a maximum value of 0.12m at 20 degrees, with an area under the curve of 0.04m.rad up to 27 degrees, and a positive range of 40 degrees. ‘These values compare favourably with the minimum requirements of 0.1m, 0.015m.rad and 15 degrees respectively. In the intermediate stages there would be a temporary angle of loll of perhaps 10 degrees, but the GZ. values and range would remain similar to the equilibrium condition, These values are well in excess ‘of the minimum requirements for intermediate stages, which are a righting lever of at east 0.05m, and a range of at least 7 degrees. During validation work, conducted by others during the development of the raking damage requirements for the 2000 HSC Code, it was found that some smaller vessels could not comply with the requirements with a 35% damage length aft. Compliance could only be achieved with a reduced damage length. This suggested ‘that a damage length greater than 35% might prove more onerous. For this reason the wider two hulls, at the deep draft, were investigated further with 45% raking aft. The results are compared with the 35% raking ‘cases in Figure 27. The stability in the equilibrium case benefits from the additional flooding and the 35% flooding remains the worst case. Iti believed thatthe vessels that failed to comply during the validation ‘exercise did so because of lack of flotation rather than stability, ‘The minimum requirements in the event of 100% raking damage are lower than for restricted raking damage. ‘There is no requirement for the GZ. value, and the area under the GZ, curve is required to be 0.015m.rad up to 30 degrees rather than 27 degrees. In these example calculations, the 100% raking damage cases exhibited greater stability than the restricted damage cases. WOLFSON UNIT In conclusion, these results imply that the stability requirements in the event of raking damage are not likely to be more onerous than the intact requirements. (On the basis ofthis parametric study, it was decided not to extend the model testing programme to investigate raking damage to monohull. 9 CATAMARAN RAKING DAMAGE Raking damage for catamaran 1, ID 31, was carried out using added ballast to simulate a double bottom 1.8 metres high full of water. Figure 16 shows the stability curve for the damaged vessel together with that of the undamaged catamaran, ID 28, At the assumed seale of 1:20 the displacement was 93 tonnes and the ‘corresponding draught is 1.14 metres. It can be seen from the stability curve that inclination angle was at the allowable maximum of 20 degrees but the range of positive stability is less than the requirement at just over 11 degrees. In shorter waves, it was possible to capsize the catamaran in wave heights of around 1.2 metres. In Longer waves, the wave height necessary to capsize the catamaran increased to around 3 metres. It was most vulnerable in beam seas with the damaged hull away from the waves. As described is Section 7, this model also represents an intact catamaran of greater displacement, with a heeling moment applied, if the additional ballast is assumed to represent deadweight rather than floodwater. Since the stability is defined by the righting moment, the increased displacement, from 93 to 170 tonnes, results in a corresponding reduction in the GZ. values. The result of this is thatthe intact vessel would have the GZ curve presented on Figure 16 as ID 31a, 10 TRIMARAN RAKING DAMAGE Raking damage on the trimaran was considered for just the centre hull damaged or just a sponson damaged. 10.1 Centre Hull Damage Figure 28 presents computer calculated GZ, curves for two configurations of raking damage for the trimaran with a displacement of 300 tonnes at a draught of 2.13 metres, together with the intact GZ, curve. For the small engine room case it was assumed that the centre hull was fiee flooding from the transom to 8.5 metres forward, with the rest of the centre hull completely flooded to the double bottom at 1.5 metres above base. ‘The large engine room case is similar to the previous case except that the engine room extends from the transom to 13.5 metres forward. ‘The two damaged GZ curves are significantly greater than the intact GZ curve. On testing ID 48, the intact configuration, it was found to be impossible to capsize the trimaran in waves that ranged from 2.2 metres high at 4 seconds, to over 5 metres at periods of 6 seconds and above. Clearly, since the intact trimaran did not capsize, the damaged configurations would not capsize either. Thus, physical testing was not carried out for these configurations. 10.2 Sponson Damage Calculations showed that when the sponson was subjected to 100% raking damage and allowed to free flood to its full height, then the trimaran would capsize in calm water. For this reason, it was assumed that the sponson must have a double bottom. When a trimaran is rolling in a seaway the damaged sponson will have ‘various draughts as it immerses and emerges in the waves. Iti likely that at some stages the double bottom will be full of water and at other times less than full. With a damaged sponson, an increase in double bottom height will reduce the stability of the trimaran. ‘Thus, a sponson with a given double bottom height that is partially full of floodwater represents a smaller sponson that is full and thus greater stability. Tests were carried out with a change of ballast weight and location to represent raking damage equivalent to a double bottom being completely flooded at all times, ‘When complying with the raking damage criteria with regard to the minimum double bottom height, i. 2.13 metres high at the bow and 1.64 metres high aft of midships, the trimaran fails to meet the stability criteria of the code. The GZ. curve for the trimaran with 100% raking damage to a sponson, with a double bottom 1.3 metres above base, is presented in Figure 19. As can be seen in Appendix 2, this configuration has GZ u WOLFSON UNIT properties well above the code criteria due to the double peak in the GZ curve, with a local minimum at ‘about 25° being just positive. small increase in VCG woutd cause this local minimum to be negative, and the configuration would fail the criteria. In this respect, therefore, the configuration can be considered to be only marginally above the code requirements. This test was designated ID 53 and the trimaran failed to ‘capsize in waves greater than 2 metres with a period of 4 seconds, and waves greater than S metres with periods greater than 6 seconds, AAs for the catamaran described in Section 9, the damaged trimaran ID $3 could also represent an intact trimaran of greater displacement. With the additional ballast representing deadweight rather than floodwater, the displacement is increased from 300 to 307 tonnes, and the GZ, values are decreased in proportion. This alternative configuration is denoted ID 53a, and its GZ, curve is included on Figure 19. The displacement increase is small and so the GZ. curve remains similar to that of ID 53, in contrast to the catamaran configurations, where the large quantity of floodwater gave a very different GZ, curve to the heavier intact configuration, 11 TEST FACILITY 11 Tank ‘The majority of tests were conclucted in the No.2 towing tank operated by GKN Westland Aerospace on the Isle of Wight. The tank is 76 metres long by 3.7 metres wide by 1.7 metres deep. It is equipped with a plunging wedge wavemaker. The last few tests were conducted in the GKN No. 3 tank duc to the decommissioning of the No. 2 tank, No, 3 tank is similar to No. 2 tank but larger, being 200 metres long by 4.7 metres wide by 1.7 metres deep. I is equipped with a hinged flap wavemaker. 11.2. Wavemaker ‘The wavemakers were controlled using computer software that enabled generation of the chosen wave provided it did not exceed the capabilities of the wavernaker. Owing to the different nominal scales of the models tested, the characteristics of the wavemaker and the physical properties of gravity waves the ‘envelope of wave height with wave period varied between models, Broadly, regular waves of about 1% -2 metres high could be generated at a period of 4 seconds full scale, and about 6 metres high at 10 seconds full scale, 12 TEST TECHNIQUE, 12.1 Handling At the start of each test, the model was located approximately 20 metres from the wavemaker, Its orientation ‘was controlled by hand, with light strings secured to the bow and stern. It was important to ensure that no force was applied to the strings at critical times, otherwise a capsize could be induced, If this occurred the result was ignored and the test repeated, Owing to the characteristics of the generated waves, the first couple of waves in a regular wave train were higher than those requested and, during the passage of these waves, the model was restrained from capsizing. In some of the tests it was found that the model moved rapidly down wave and capsize could be induced artificially by preventing this movement, particularly in beam seas. It was also noted that, with these motion characteristics, the first waves of a sequence applied greater force to the model than later waves that were encountered after the natural drift had become established. Capsize might occur in the initial encounters, but could be avoided if the model were assisted manually until it had established its natural rate of drift. Once this had been recognised, care was taken to ensure that the model was prevented from capsizing during the transition phase at the start ofa test. It was considered that this method more closely represented the situation ofa vessel drifting freely in a seaway, but it should be noted that an encounter with a severe wash might better be represented by the initial, more onerous, encounters, WOLFSON UNIT 12.2 Orientation With cach new configuration, tests were carried out at all orientations to the waves, that is, head, bow, beam, quarter and stern seas, with damage towards and away from the oncoming waves. If a model configuration exhibited a clear vulnerability ata particular heading, this enabled the minimum wave height to be determined quickly without the need to conduct subsequent tests at all headings. Advantage could only be taken of this feature for tests ata constant wave frequency because, ata different frequency the critical capsize mode and heading could be quite different. ‘Some configurations had a tendency quickly to adopt some particular orientation to the waves. Whilst every effort was made to ensure that all orientations were tested, no attempt was made to hold the model at a given orientation even where it seemed possible that this was the worst case. This was necessary since the forces required to restrain the model could themselves induce a capsize. Repeated attempts were made, however, to determine the vulnerability of the model at each desired orientation. Generally, provided capsize did not occur sooner, the model was kept at a given orientation for several ‘waves until it was felt that capsize would not occur. In some marginal cases, this period was extended to some minutes where it appeared that capsize might occur. 12.3. Checks on Condition Prior to each test, the intact compartments and deck were cleated of any residual water. In some cases it was necessary that the deck of the model was wiped free of water droplets, since it was found that these could influence the capsize boundary. In general, once a test was underway as many orientations as practicable were tried before stopping the waves and restarting. It was, therefore, not practicable to have dry decks at the start of each orientation, For this reason, as well as in the interests of efficiency, it was always the intention to begin each test at the orientation most likely to cause capsize. At intervals, and following a capsize, the condition of the model and ballast locations were checked by ‘monitoring the freeboard and equilibrium heel angle. 12.4 Regular Waves Because of the large number of variables and model configurations, it was not possible within the budgetary constraints to conduct tests of long duration in sea spectra. Most tests therefore were conducted in regular waves. The wave periods represented between 3 and 10 seconds at full scale, and were varied in 1 second increments, although, in some circumstances non-integral values were used So that a more onerous wave height could be set. The wave heights set, generally, were in % or Imetre increments until capsize was achieved. The smallest incremental value set was around 4 metres. 12.5 Wave Spectra Some tests were conclucted in irregular waves. The spectra used were all of the JONSWAP type, representative of waves with a limited fetch. These tests were carried out in a similar manner to the regular ‘wave tests. Figure 29 shows a typical spectrum measured during one of the tests, Sea spectra are discussed further in Section 14, 12.6 Wave Measurement ‘A resistance wave probe, located up wave of the model, was used to measure the generated waves, Time histories of the waves were recorded by proprietary software. In most cases the model drifted away from the wave probe and interference between them was negligible. In some cases, in particular a heeled catamaran with the solid bridge deck immersed, the model motion ‘generated substantial waves, which propagated towards the wavemaker and polluted the wave record, Care ‘was required, therefore, in interpreting the wave record, to ensure that the measured waves were representative of those encountered by the model. In extreme cases such as with the catamaran described above, it was necessary to repeat and measure the wave generation in the absence of the model, WOLFSON UNIT 12.7. Behaviour Monitoring ‘While one experimenter controlled the model, a second operated the wavemaker, wave measurement system, and video recording system. The two experimenters were located such that their views of the model enabled the model motions, water ingress and deck wetting to be monitored. Details ofthe behaviour could be examined following the tests with the aid of the video record. 12.8 Test Results ‘The tests yielded data for each model configuration consisting of capsize or no capsize at a given wave height and period. For some configurations the capsize wave height was determined over a range of wave . Figure 30 shows typical capsize boundary curves with the defining points plotted, Figure 31 to Figure 38 show all the boundary curves but the defining points have been omitted for claity. For other configurations the minimum wave height to result in capsize was determined without defining the full boundary. Appendix 2 presents the minimum wave height required to cause capsize for all configurations. It should be noted that it was not possible to cause a capsize in some configurations, and in these instances the ‘wave height column has been left blank. For the cases where capsize was not achieved it is irelevant to indicate the highest wave that failed to achieve capsize, because the critical orientation and wave frequency are unknown, 13. REPRESENTATION OF HEELING MOMENTS ‘The HSC Code requires the application of heeling moments to be taken into consideration when assessing the stability. Passenger crowding, wind heeling, lifeboat launching and high speed turning moments are included in the assessments, singly or in combination, Because the tests were conducted with models stationary, the heeling moments due to high speed turning ‘were not considered. Passenger crowding and lifeboat launching moments are represented correctly at model scale by transverse movement of ballast. ‘Wind hecling moments were the subject of MCA Research Projects 503 and 537, where wind tunnel tests ‘were used to measure heeling moments on a range of HSC forms, upright and heeled. It was found that, for monohull vessels of typical HSC proportions, the heeling moments do not vary substantially with heel, For vessels with a very large beam to height ratio, the heeling moment when heeled may be greater than when upright. For a beam to height ratio of 2.75, for example, the wind heeling moment, when heeled by the wind to 10 degrees, may be 50% greater than when upright. This factor does not apply to the equilibrium heel angle due to asymmetric damage, because it results to a large extent from the transverse movement of the centre of buoyancy as a result of an applied moment. Any attempt to model such variable moments would have introduced a high level of complexity to the tests, and was not considered worthwhile. Because the models were not of specific designs of vessel, it was not possible to calculate the moments in the usual way. Any size of superstructure might be added to the hull modelled for example, with the result that wind heeling moments could be wide ranging. With heeling moments applied, the Code requirements are for ‘maximum heel angles, minimum residual GZ values or minimum GZ. curve areas. For a given vessel configuration, therefore, i is possible to calculate the heeling moment corresponding to the minimum requirements, The assumed cause of the heeling moment was not of concern in these tests and therefore, in representing these moments, it was decided to use simple ballast transfer to obtain the required equitibrium. heel angle, 14 CONSIDERATION OF IRREGULAR WAVES, 14.1 Spectra ‘Vessels at sea invariably operate in a spectrum. A sea spectrum is defined by type, significant height and period. In waters with limited fetch the spectrum is often assumed to be of the JONSWAP type, whereas in the open ocean itis often assumed to be an ITTC spectrum. The JONSWAP spectrum has more of its energy close to the modal period, that is it has a more pronounced peak. WOLFSON UNIT Wave statistics are available for most parts of the world. Ref. 2 presents the data for these as the number of ‘occurrences in 1000 observations, for bands of significant wave height and zero crossing period. These data have been manipulated and presented graphically in Figure 39 to show contours, where increasing probability of occurrence is given a darker shade, It can be seen from the three areas presented that the probabilities of occurrence for a given wave height and period are significantly different 14.2 Encountered Wave Heights The probability of encountering a wave larger than the significant height increases as the total number of ‘encounters increases, whilst operating in non-changing spectrum, For example, it ean be shown that, approximately, | in 2000 waves in a Rayleigh distributed spectrum will be twice the significant height. The ‘expected maximum height can be calculated based on the number of encounters and the significant height from the following formula: Expected Maximum Height = C*Significant Height/1.414 Where values for C, derived from Ref. 3, and the corresponding expected maximum heights are: C_| Maximum/Significant Height [ Occurrence 7.708 121 Tin 10 2.280 LOL Tin 100 2.738 194 Tin 1,000 3.130 221 1 in 10,000 3.478 2.46 Tin 100,000 In practical terms, a vessel on a sheltered route, operating in a spectrum with a modal period of 4 seconds, right expect to encounter a wave of twice the significant height about once in 2% hours. In exposed conditions, ina fully developed spectrum with a modal period of 10 seconds, one might expect to encounter a wave of twice the significant height about once in 5% hours. It must be understood, however, that if such @ wave is expected to occur once in 2000 waves, it may occur asthe first wave, or may not occur at al. 143 Model Tests Some tests were conducted in a JONSWAP spectrum to try to correlate the behaviour in a sea state with that observed in regular waves. A typical wave height record is presented in Figure 29, In some instances, the ‘models capsized quickly, sometimes after a prolonged period, and, on other occasions, they did not capsize for the duration of the test. ‘These results are entirely consistent with what might be expected. ‘An examination of the wave record, measured some distance up wave from the model, showed that waves large enough to capsize the model during regular wave tests, existed during the duration of the test, when the ‘model did indeed capsize. Owing to the position of the wave probe, it cannot be known with any certainty ‘what heights of individual waves were encountered by the model, because the wave profile changes continuously as the individual waves in the spectrum travel at different speeds. It is known, however, that the ‘model encountered the same spectrum as the wave probe. 14.4. Discussion ‘The nature of the capsize mechanism has a bearing on the probability of encountering waves that will eause the ship to capsize. A ship may require a train of several large waves, of similar period, to roll it to progressively greater angles until it capsizes. Another vessel may require just one such wave to cay ‘The former will be much less likely to capsize in a given sea state than the latter, since the probal ‘encountering several large waves consecutively is relatively low. In the model tests with regular waves, very few capsizes, ifany, were due to synchronous rolling, where a gradually increasing roll angle with each successive wave results eventually ina capsize. If itis assumed that eapsize is primarily due to encountering a single large wave, and not synchronous rolling, then it follows that itis necessary only to perform tests in regular waves, and to extend the results, using statistics, to allow for a sea state if desired, ‘Taking account of the variability of the occurrence of a sea state in a given geographical region, and the probability of encountering a specific wave in that sea state during a voyage, itis possible to assess the risk WOLFSON UNIT of capsize using existing statistical methods. Clearly, this risk will change with season, region and voyage duration, 15 SUMMARY OF CAPSIZE MECHANISMS ‘The 53 configurations tested resulted in 39 being capsized and 14 remaining upright. In general several capsizes were achieved for each configuration. It was noticed that in some configurations the mode of capsize depended on wave frequency. In other configurations, there was no evidence of this but for the later tests a minimum wave height to capsize was determined, without defining the complete boundary, and here, changes in capsize mode might have existed. Photographs of two runs are presented to illustrate the tet facility and typical model behaviour. See Figure 40 and Figure 41. ‘During early tests on Monohull 1, 5mm square battens were fixed to the deck in the form ofa rectangular coaming, to facilitate fitting of a simple superstructure. These were found to retain a small amount of water con deck that was sufficient to cause capsize. Because details of structure and deck arrangements are too varied to be incorporated into this study, the battens were removed and the model tested with a flush deck, It was very difficult to configure the catamarans close to the Code criteria when intact and upright, owing to the high VCG required. Catamaran 2 was tested intact and upright with, abnormally high, VCGs of 9 and 11 metres, and capsized. It was also possible to capsize an intact catamaran when subjected to an initial heel angle with a reasonable VCG. The intact trimaran was the only multihull model that could be tested upright close to the code limit, but it failed to capsize. It was also tested with an initial heel but again failed to capsize, Other mode! tests on the stability and capsizing of catamaran vessels, Ref. 1, showed that the range of stability is the important parameter in a breaking wave capsize. There have been few instances of catamarans being capsized by breaking waves, but those model tests showed that it isa possible occurrence, given a breaking crest of sufficient height. The capsize mechanism entailed rolling of the mode! toa large heel angle following the impact of the breaking crest on the windward hull, ‘The model then tended to drift sideways in the wave, at the extreme heel angle, and its subsequent return to upright or capsize was dependent on ‘whether the heel angle was greater or les than the range of stability. While breaking wave impact was not within the scope of this study, observations of the models suggested that, in many cases, a similar mechanism was involved. That is, the model was rolled by wave action to some angle and, if that angle was ‘greater than the range of stability, it subsequently capsized. When the angle was about the same as the range of stability, the model frequently was observed to remain at that angle fora finite period, perhaps terminated by the encounter with the next wave in the sequenice. This suggests that the dynamic effects at that stage of the process were negligible, and the behaviour was largely dependent on the statical stability properties. The carly, dynamic phase of the event is determined by the combination of vessel configuration, stability, heading and wave properties, whereas the result of the latter phase appears to be determined by the statical stability alone. 16 TEST RESULTS 16.1 Capsize Mode Definitions Each model configuration was tested over a range of wave heights and/or wave periods and, in general, several capsizes occurred, In some configurations the mode of capsize varied for different wave periods. In this section the eapsize mode for each configuration is defined as the mode that caused capsize at the lowest ‘wave height. In the following descriptions of the modes, windward is defined as towards the wavemaker and leeward is away from the wavemaker. ‘The capsize modes have been grouped together to form 5 generic ‘modes, denoted A to E: A: Lack of stability on a wave in stern seas, 16 WOLFSON UNIT B: Beam seas induced a heel to windward, this heel gradually inereased, sometimes with a small amount of rolling, until the model capsized to windward C: Hecled to leeward, down by the stern. In bow seas, gradually increased the heel and trim with some rolling and pitching until the model capsized diagonally. D: Heoled to leeward. In beam or bow seas, increasing roll angles until the model capsized to leeward. E: Water trapped inside superstructure, with gradually increasing heel as more water was trapped until the ‘model capsized. When the damage was to leeward it capsized to leeward, and when the damage was to ‘windward it capsized to windward 16,2 Implications of Models that Did Not Capsize ‘Some models did not capsize in any of the wave conditions tested. It is not straightforward to indicate the maximum wave height that the models survived because the maximum height that could be generated was dependent on the wave frequency. Figure 42 presents tested wave conditions for ID 42, which is an example ‘of a model that did not capsize despite tests in a wide range of wave frequencies, In this example the model survived waves of 7 metres with a period of 10 seconds and 6.8 metres at 8 seconds, but, at this scale, waves of only 2.2 metres could be generated with a period of 4 seconds, ‘Comparison of these values with the seastate charts presented in Figure 39 shows that larger waves with these periods would be unlikely to occur. Waves of greater than 2.2 metres with a period of 4 seconds would be of extreme steepness and likely to break, and it was the steepness limit that restricted the wave generation at the higher frequencies. ‘The limitations ofthe test facility therefore are unlikely to have prejudiced the results 16.3 Results for Each Configuration 16.3.1 Monohull { intact at a displacement of 123 tonnes and VCG 4.684 metres, variation in initial list IDI Heeled to 5°to leeward atthe criterion 4 limit. Beam seas caused slowly increasing roll angles until the last wave, when it failed to roll back, and on the next wave it capsized in 2.1 metre waves. Category D. ID2_—_Heeled 10° to leeward at the criteria 4 and 9 limits. It capsized in the same way as IDI in 0.8 metre waves, Category D. 16.3.2 Monohull 1 intact at a displacement of 200 tonnes, variation in initial list ID3 Upright at a VCG of 4.47 metres with a margin over the Code requirements. Did not capsize. IDS VOG 4.65 metres heeled 4.9° to port, failed criterion 5, Stem to the waves, ran down to leeward as it rolled over, veered to starboard and capsized in 3.3 metre waves to leeward, port. Photographs of this configuration are presented in Figure 40. Category D. IDS VCG 4,505 metres heeled 10° to leeward at the criteria 5 and 9 limits. Capsized in a similar way to IDI in 1.65 metre waves. Category D. 16.3.3 Monohull 1 damaged between 23% - 76% at a displacement of 123 tonnes, variation in list ID6 Upright at a VCG of 4.95 metres with a margin over the Code requirements. Damage to leeward in {quartering seas. Gradually inereasing hee! and capsized to windward in 1.1 metres waves. Category B. ID7 — VCG 5.012 metres, heeled 10° to windward with a margin over Code requirements. In quartering, seas, gradually increasing heel and capsized to windward in 0.77 metre waves. Category B. 7 WOLFSON UNIT ID8 _VCG 5.035 metres, heeled 13° to windward atthe criteria 2 and 10 fimits. In beam seas, gradually increasing heel and capsized to windward in 0.5 metre waves. Category B. 163.4 Monohull 1 damaged between 5% - 35%, with a superstructure, ata displacement of 200 tonnes, upright, vari vcG 1D VCG 3.799 metres at the criterion 2 limit, Did not capsize, IDIO_ VCG 3.97 metres failed criterion 2. Did not capsize. 16.3.5 Monohull 1 damaged between 5% - 35% metres IDII At the criterion 2 limit and then heeled 6° to port to be at the criterion 10 limit. With waves on the port quarter, gradually veered to starboard with increasing heel angle and capsized in 1.4 metre waves. Category A! at a displacement of 200 tonnes and VCG of 3.799 16.3.6 Monohull 1 damaged between 23% and 76 metres, upright IDI2__ With a margin over the Code criteria, Did not capsize. displacement of 200 tonnes and VCG 4.05 16.3.7 Monohull 1 damaged between 23% and 76%, at a displacement of 200 tonnes and VCG 4.13 ‘metres with a margin over the Code criteria, upright, variation of geometry IDI3_Deck intact with a superstructure, Damage to windward in beam seas. Waves washing into superstructure with gradually increasing heel and roll, and capsize to windward in 1.6 metre waves. Category E ID14 Without a superstructure and deck damaged. Did not capsize. 16.3.8 Monohull 1 damaged between 23% and 76% at a displacement of 200 tonnes and VCG 4.347 metres at the criterion 2 limit, upright, variation of geometry IDIS Deck intact without superstructure in beam seas. Waves caused it to heel to windward. As it rolled the heel increased and the model capsized to windward in 0.73 metre waves. Category B. IDI6 Deck damaged without superstructure in beam seas. Same as IDIS, capsized in 0.65 metre waves. Category B, IDI7 Deck damaged with superstructure, Same as IDI3 and capsized in 0.86 metre waves. Category B. 16.3.9 Monohull 1 damaged between 23% and 76% with the deck intact, at a displacement of 200 tonnes and VCG 4.13 metres with a margin over the Code criteria, heeled 5.5°, variation of geometry IDL No superstructure, heeled to windward in beam seas. Gradual increase waves aft of the beam and capsize to windward in 0.5 metres waves. Category B. heel, slight veer to put the ID19 With a superstructure heeled to windward in beam seas. Same as ID13, capsized in 1.15 metre waves, Category E. 16.3.10Monohull 1 damaged between 23% and 76%, at a displacement of 200 tonnes and VCG 4.347 metres at the criteria 2 and 10 Timits, heeled 6.2", variation of geometry 1D20 Deck intact with no superstructure and heeled to windward in beam seas. As the model rolled the heel gradually increased and it capsized to windward in 0.4 metre waves, Category B, ID21_ Deck damaged with superstructure and heeled to windward in beam seas, Same as 1D13, capsized in 0.6 metre waves. Category E. WOLFSON UNIT 16.3.11 Monohull 2 with superstructure, intact at a displacement of 1260 tonnes and VCG 9.277 metres at the criteria 5 and 9 limits, heeled 10° 1D22_ With the stern to windward, the heel gradually increased and the model capsized in 2.74 metre waves. Category A. 16.3.12Monohull 2 with superstructure damaged between 0% -21%, at a displacement of 1260 tonnes, and VCG 10.11 metres with a margin over the Code criteria, variation in initial list 123 Upright. Did not capsize. 1D24 _Heeled 3.8° to windward in beam seas, damage to windward. Gradually increasing heel as more ‘water remained in superstructure, and capsized in 1.48 metre waves. Category E. 16,3.13 Monohull 2 with superstructure, at a displacement of 1260 tonnes, at the criteria 2 and 10 limits, variation in damage compartments 1D25 VG 10.331 metres and heeled 4.6° to windward in quartering seas, damage 0% - 21% to windward, Same as 1D24, capsized in | metre waves. Category E. 1126 VCG 9,568 metres and heeled 7.5° to windward in bow seas, damage 21% - 45% to windward. Same as 1D24, capsized in 2.3 metre waves. Category E. 127 VCG 10.78 metres and heeled 5.1° to leeward in bow seas, damage 45% - 86% to leeward. Same as 1124, capsized in 1.7 metre waves. Category E. 16,3.14Catamaran 1 at a displacement of 93 tonnes, variation in damage compartments 1D28 CG 6.932 metres intact at the criterion 9 limit and heeled 10° to leeward in beam seas. Synchronous rolling, gradually increasing and capsize to leeward in 0.98 metre waves. Catgeory D. 1D29__ VCG 4.075 metres damaged 0% - 31% at the criterion 8 limit and heeled 15.1° to Leeward in bow seas. Synchronous rolling, one slightly larger roll and on the next wave capsized to leeward in 1,15 metre waves. Catgeory D, 130 VCG 3.887 metres damaged 31% - 82% at the criterion 6 limit and heeled 10°, Did not capsize. 16.3.15Catamaran 1 heeled 20° by ballast, the same test representing 100% raking damage, or intact ata higher displacement ID31_VCG 5.76 metres at a displacement of 93 tonnes, 100% raking damage, fails criterion 7 and heeled to leeward in beam seas, Synchronous rolling near to capsize, veered to quartering seas, tripped over leeward bow and capsized to leeward in 1.2 metre waves. Category D. ID31a_VCG 3.69 metres at a displacement of 170 tonnes, intact, fails criteria 9 and 11 and heeled to leeward in beam seas, Same as [D3 1, capsized in 1.2 metre waves, Category D. 16.3.16Catamaran 2 intact and upright at a displacement of 93 tonnes with a margin over the Code criteria, variation of VCG 1032 VCG 8.99 metres in beam seas. Windward and leeward hulls out ofthe water alternately until, on cone wave, the windward hull came out and roll continued with capsize to leeward in 4 metre waves. Category D. 1D33.VCG 11 metres in beam seas. Leeward hull out of the water on each roll until it continued its roll and capsized to windward in 2.8 metre waves. Category D. WOLFSON UNIT 16.3.17Catamaran 2 damaged 0% - 31%, at a displacement of 93 tonnes, variation of VCG or initial list 1D34VCG 4.698 metres, with a margin over the Code criteria, heeled 10° to leeward in bow seas. ‘Windward bow gradually thrown higher and higher and with consequent sinkage atthe stern as it pitched, and capsized to leeward in 1.9 metre waves. Category C. ID35_VCG 5.9 metres, heeled 10.7° to leeward, at the criterion 11 limit, in bow seas, Same as ID34, capsized in 0.9 metre waves. Category C. ID36VCG 4.7 metres heeled 13° to leeward, fails criterion 11, in bow seas. Same as ID34, capsized in 0.95 metre waves. Category C. 16,3.18Catamaran 2 damaged 8% - 31%, ata displacement of 93 tonnes, heeled 12, variation of VCG 1D37-_VCG 4.7 metres heeled to leeward in bow seas with a margin over the Code erteria. Whole of ‘windward hull clear of the waves as it rolled, and consequent sinking by the stern. After @ couple of rol: capsized in 1.6 metre waves. Category C. 1D38_VCG 4.968 metres heeled to leeward in bow seas with a margin over the Code criteria. Same as 1D34, capsized in 1.3 metre waves, Category C. 1D39__VCG 5.236 metres heeled to leeward in bow seas at the criterion 6 limit, Same as ID34, capsized in 1.2 metre waves. Category C. ID40— VCG 6.04 metres heeled to leeward in bow seas failed criterion 6. Same as ID34, capsized in 0.9 metre waves, Category C. 16.3.19Catamaran 2 damaged 24% - 58%, at a displacement of 93 tonnes, heeled 9.3° with a margin over the Code criteria ID41 — VCG 6.096 metres, heeled to leeward in beam seas. Rolling with windward hull coming clear of the water and as it rolled, pitching down by the bow. Capsized to leeward in 1.9 metre waves. Category C. 163.20Catamaran 3 damaged 0% - 39%, at a displacement of 1100 tonnes, heeled 13.8° with a margin over the Code criteria ID42_VCG 10.53 metres. Did not capsize. 163.21Catamaran 3 damaged 14% -72%, at a displacement of 1100 tonnes failed criterion 8, variation of initial list 1D43__VCG 10.53 metres, heeled 17.6° to leeward in bow seas. Rolling with the windward hull coming clear of the water. On one roll, rather than rolling back very much, it settled lower in the water, When the ‘windward hull hit the next wave the model capsized to leeward in 2.3 metre waves. Photographs of this, configuration are presented in Figure 41. Category D. ID44_VCG 10.53 metres, heeled 20° to leeward in beam seas. Rolling with gradually inereasing roll angles until it capsized to leeward in 1.9 metre waves. Category D. ID45 VG 10.5 metres with a superstructure, heeled 21° to leeward in beam seas, Rolling gradually increased with the superstructure awash and the model capsized in 3 metre waves. Category D. ID46 VCG 10.52 metres, heeled 21.8° to leeward in beam seas. Rolling gradually increasing and capsize to leeward in 1.3 metre waves. Category D. 163.22Trimaran 1 Intaet ID47_ VCG 5.273 metres at displacement of 175 tonnes, with a margin over the Code criteria using the second stability peak, and heeled 8.5°. Did not capsize. Figure 19 presents the stability curve for this model 20 WOLFSON UNIT and it can be seen that there are two angles of stable equilibrium, one at 8° and the other at 32°, For the frst few waves the model tended to roll about the 8° equilibrium point, A slightly larger roll caused it to adopt the 32" equilibrium position and thereafter the model rolled about this point. Owing to the large range, it did not capsize, but the resultant heel angles were extreme, ID48—VCG 4.7 metres at a displacement of 300 tonnes at the criterion 4 limit and upright, Did not capsize. 163.23 Trimaran | at a displacement of 175 tonnes, sponson damaged 61% - 100% with a VCG of 4.439 metres and a margin over the Code criteria, variation of initial list 1D49 Upright, representing passengers on the non-damaged side, Did not capsize, IDSO Heeled 3.6°, representing passengers evenly distributed. Did not capsize. Same motions as 1D47, rolling about 30° or so. IDS] Heeled at 7, representing passengers on the damaged side with a margin over the Code criteria using, the second peak. Did not capsize, Same motions as 1D47, rolling about 30° or so. 163.24 Trimaran 1 at a displacement of 300 tonnes, sponson damaged 61% - 100%, with a VCG of 4.195 metres fails criteria 11 when heeled 10° 1DS2_ Heeled to leeward in beam seas. Initially with irregular rolling about the frst stable eq) point and subsequently rolling about the second stable equilibrium point of 28°. Heeled slightly more with ‘each successive wave until it eapsized to leeward in 1.6 metre waves. Category D. 163.25Trimaran 1 heeled 8.9°, the same test representing 100% raking to sponson or intact at a higher displacement 153 VCG of 4.7 metres at a displacement of 300 tonnes, with 100% sponson raking damage, and with a margin over the Code criteria, Did not capsize 1DS3a_ VCG of 4.63 metres at a displacement of 307 tonnes, intact and with a margin over the Code criteria, Did not capsize 17 RELATIONSHIPS BETWEEN STABILITY PARAMETERS Ithas become an accepted convention that stability is assessed against criteria based on the parameters selected by Rahola many years ago. The 2000 HSC Code is no exception, incorporating requirements for GM, GZ, area under the GZ. curve, and the angle at which the maximum GZ. occurs. For upright, intact ‘vessels, these parameters are straightforward, For asymmetric damage cases, or when heeling moments are applied, the vessel takes up an equilibrium heel angle and criteria may be applied to the residual stability curve. Inthis project, many of the tests were conducted with initial heel because of asymmetric flooding or an applied moment. In order to make comparisons between those models that were initially upright and those that were heeled, the residual stability curves have been used, For vessels with irregular hull forms, or with superstructures included, the GZ curve may be of irregular shape. When a vessel has little residual stability however, its GZ. curve tends to be of parabolic form. Because of this, it is inevitable that there are relationships between the stability parameters. Figure 43 presents a comparison of selected stability parameters forthe tested configurations, in terms of their variation with GM. In this presentation, GZarea is the total area under the residual stability curve and GZmax is the peak residual GZ. value. All of the graphs reveal considerable scatter, and there is not a clear separation between the values for those models that capsized and those that did not. ‘There appear to be trends of GZmax and GZ. area with GM, but some data points lie well outside the main envelope of data, These are the values for intact, upright catamarans. The GZ. curve for a catamaran tends 10 21 WOLFSON UNIT be virtually linear over a range of angles, until one hull begins to emerge from the water. Applying, or increasing, a heeling moment therefore has the effect of reducing the maximum GZ. value and area under the curve, without affecting the residual GM value. With different heeling moments appl data points could be moved up or down vertically on these two graphs. GM frequent! stability, for example with a roll test, or GM monitoring system. For a given vessel, an increase in the VCG will result in roughly proportional reductions in GM, GZmax and GZarea, and so GM is useful in that respect, These data demonstrate, however, that GM is not necessarily a good indication of the GZmax or GZarea. ‘The graph of range of stability against GM reveals considerable scatter. For the monohull models, the range of stability varied considerably despite very little variation in GM, and no trend was apparent. This graph illustrates the fact that a catamaran requires a much greater GM to obtain a range of stability equivalent to that of a monohull. ‘The upper graph of Figure 44 presents the relationship between GZarea and GZmax. There is a strong trend, ‘as would be expected, but the data indicate different relationships for monohulls, catamarans and trimarans, ‘GZ curves for catamarans have a more pronounced peak, so that the area tends to be lower for @ given ‘GZmax. In general, the data indicate that those models that did not capsize had a relatively high GZarea for a given GZmax. ‘The lower graph of Figure 44 reveals a very strong relationship is between GZarea and the product of the range of positive stability and GZmax. The data collapse into a very narrow envelope. Monohulls and ‘catamarans display the same relationship, with GZarea consistently about 60% of the product of range and ‘GZmax. This represents the area ratio of the typical parabolic form of the GZ. curve to the enclosing rectangle defined by the range and GZmax. Some trimarans had a relatively low GZarea because thei GZ, ‘curves had pronounced double maxima, See Figure 19. ‘There isa similar relationship between the angle of maximum GZ and the range of stability, though this is not illustrated, with a maximum value at about half the range. ‘One may infer from these relationships that it may not be necessary to regulate GM, GZmax, GZarea, angle ‘of GZmax and range, as some parameters are related. 18 RELATIONSHIPS BETWEEN STABILITY PARAMETERS AND CAPSIZE. 18.1 Data for the Vessels Modelled Figure 45 presents values of the minimum wave height to result in capsize for each model configuration, ‘against the residual GZmax and the total area under the residual GZ curve. There are no clear trends, although there appears to be a boundary, below which no capsizes occurred for a given value of each stability parameter. These ‘safe’ boundaries, for monohulls and catamarans, are included on one of the ‘graphs as an illustration, One can argue that the slopes of these boundaries provide evidence that greater stability provides greater safety, and that one can select a minimum value ofthe stability parameter, in this case GZarea, to provide safety against capsize in a chosen wave height. Thus, for an intact catamaran to be ssafe from capsize in waves up to 2 metres, apparently requires a GZarea of at east 0.3 metre radians, while a monohull would require a GZarea of only 0.05 meire radians to achieve the same level of safety. Itis clear that the boundary tends to be defined by a single model test and some other model, different in ‘some way to those tested, might capsize in a lower wave and render the boundary invalid. Grouping of data points along the lower boundary of an envelope therefore leads to confidence in that boundary. Similar grouping of the data points appears on the graphs of wave height against GZmax and GZarea, as ‘would be expected in view of their close relationship discussed in Section 17. An interesting resutt is the lack of any clear distinction between the intact and damaged configurations. 22 WOLFSON UNIT Plots of wave height against GM and Range of stability did not provide any trends. 18.2. Effects of Sealing Whilst the models were based on vessels of a specific size, they could equally represent some other size of ‘vessel ata different scale. Thus the model, monohull 2, representing a vessel of 96 metres ata scale of 1:40, ‘could equally be assumed to represent a vessel of similar size to monohull 1, ata scale of 1:16. The effects of re-scaling the model test data are that the wave heights required to capsize are scaled accordingly, as are the GZ properties. Scaling of GZ values may seem contrary to the assumption implied by most regulations, that is, GZ remains constant regardless of ship size and all ships must comply with the same minimum GZ characteristics, but is inevitable if Froude scaling is applied. With increasing ship size, GZ. values tend to remain constant because of practical constraints on the design, as well as to comply with regulatory minima. It isnot the case that GZ. values are non-dimensional. ‘The effect of re-scaling all of the test results shown on Figure 45 is that each point can be used to derive a line generated by scaling the model to a range of ship sizes. Because wave height, GM, GZmax and GZarea all scale in proportion to vessel size, each point on the graphs generates a line through the origin. The result is a group of radial lines, one of which defines the safe boundary. For this reason, the safe boundaries must be linear. ‘The implications of scaling on the graph of wave height against GZmax are that a monohull of 20 metres, required to operate safely in waves of | metre, requires a GZmax of 0.1 metres. A geosim of that vessel, 40 ‘metres long, might be required to operate in waves of 2 metres and would require a GZmax of 0.2 metres. If, however, the larger vessel were only required to operate in the same waves as the smaller vessel, it would only require the same GZmax, that is 0.1 metres. Since all other dimensions would be scaled in proportion, the lower GZ requirement would enable the vessel to be loaded to a relatively higher VG. Since the current regulatory system requires the same GZ minima for all sizes of vessel, the implication of this graph is that they are all assumed to operate in the same seastate, For different size vessels to achieve the ‘same GZ minima, as dictated by current regulations, requires them to have different VCG values relative to their depth. It is, therefore, more difficult for a small vessel to comply with the requirements than a large one. Itis logical that larger vessels are less vulnerable than small vessels and, in general, one would expect the data points defined by large vessels to lie above the safe boundary. This plot is perhaps rather misleading because it does not facilitate comparison of similar vessels. ‘The scaling mechanism enables each of the model test results to be used to represent a vessel with a specific value of a particular stability parameter. The results of such an exercise are presented in Figure 46, where all ‘model test configurations have been scaled such that the residual GZmax = 0.1 metre, In effect, this is a non- dimensional plot, because the scaling requires both the wave height and length to be multiplied by the factor 0.1/GZmax. ‘The values therefore have been normalised with respect to GZmax. This indicates very clearly the increasing level of safety with length of vessel. The vessels that ate indicated to be the least safe are the ‘multihulls because, fora given GZmax, they may be capsized by the smallest waves. In order to have such low GZ values, however, they must be very small vessels. As for Figure 45, this plot may be misleading, and itis not reasonable to design a 100 metre catamaran with a GZmax of 0.1 metres, ‘One might argue that, because the regulations require the same GZmax for all vessels, one should plot non- dimensional wave height against each stability parameter. If, for example, the data were plotted on a graph of ‘wave height/length against GZmax, re-scaling a data point would move it along the GZmax axis, but would not alter its position relative to the height/length axis. One could, therefore move all ofthe data points to any position along the GZmax axis. A larger vessel would have a greater GZmax, but would capsize in the same ‘wave height/length as a smaller vessel. This would indicate that the value of GZmax has no effect on the ‘wave height/length. Such an argument provides little support for the conventional use of common GZ 23 WOLFSON UNIT requirements for all sizes of vessel, except that the larger vessel could survive in larger waves than the smaller vessel. ‘The stability parameter that does not scale in this way is the range of stability. Re-scaling a model to representa larger size of vessel increases all of the GZ. values by the scale factor but does not affect the range. Rescaling the data points on the graph of wave height against range therefore would result in each t generating a vertical line. This explains why no trend was found on such a plot, the distribution of points resulting largely from the size to which the model data were scaled. ‘To avoid these scaling issues itis essential to plot data on axes that are dimensionally consistent or, preferably, non-dimensional. The plots of wave height against GZmax and GZarea are co fimensions, but a plot of wave height against range would not be, because range is non: ‘wave height scales with length, 19 THE IMPORTANCE OF RESIDUAL STABILITY 19.1 Evidence from the Tests General observations of the behaviour of the models led the experimenters to the hypothesis that it is the residual stability curve, after the application of any passenger crowding, wind heeling, or other quasi static ‘moments, that governs the response of a vessel to waves. In the lower graph of Figure 46 the upright and heeled configurations have been identified with different symbols, The upright and heeled data both span the width of the envelope, and show no difference in ‘vulnerability. The heeled monohull data are labelled to indicate their initial equilibrium heel angle, and confirm that there is no relationship between the initial heel angle and the vulnerability, fora given level of stability. The same was found to be true of the heeled multihull data, although the labels have been omitted for clarity. In their case the initial heel angles were due to a combination of asymmetric flooding and applied heeling moments. The tests suggest, therefore, that there is no increase in vulnerability with increasing initial angle of heel, yy remains adequate, This leads to the conclusion that, capsize in waves, there is nothing unique about the upright case. One may consider all vessels, regardless of the asymmetry of flooding or the applied heeling moments, as floating bodies characterised by their residual stability, 19.2 Methods of Assessment Ifheeling moments are anticipated, the stability in the absence of those moments is not a comprehensive assessment of the safety of the vessel. There are two options for stability assessment, One must either set minimum levels of stability when upright, with suitable margins based on the assumption that all vessels will be subject to similar heeling levers, or set minimum levels of residual stability, having estimated the anticipated heeling levers for the vessel in question, ‘The 2000 HSC Code adopts different approaches depending on the type of vessel, and whether itis intaet or damaged. The summary of criteria presented in Appendix 1 shows the vari approaches. For intact monohulls, the residual stability is assessed in two ways, wind heeling by the use of the weather criterion, ‘and passenger crowding by a maximum permissible heel angle. The combined effects of wind heeling and passenger crowding are not assessed, For intact multihulls, the same heel angle limit applies to passenger ‘crowding, and the combined effects of crowding and wind heeling are assessed using a minimum GZarea. ‘There is a potential problem with the method used for monohull. Consider the GZ: curves presented in Figure 47. In the absence of heeling moments, the curve has the required GZ and area, with a maximum GZ ‘occurring at the minimum angle permitted, 15 degrees. If passenger crowding, were to heel the vessel to the maximum of 10 degrees, however, the residual stability would be negligible but is not assessed. 24 WOLFSON UNIT In the same situation, a multihull would be required to maintain a residual GZarea of 0.028 metre radians. To achieve such an area might, for example, require a residual range of stability of 25 degrees and a residual GZmax of 0.1 metre. A monohull model configuration that had a GZarea of 0.028 metre radians was ID 5, and its GZ curve, presented in Figure 13, is very different to the residual curve in Figure 47, 20 COMPARISON OF MONOHULL AND MULTIHULL CRITERIA 20.1. Types of Criteria ‘Appendix I reveals several differences between the approaches for these different vessel types. The criteria ‘may be divided into three distinct groups: 1. ‘Those that address the stability in the absence of heeling moments. 2. Those that address the residual stability with heeling moments applied. 3. Those that provide for the safety of passengers rather than the survival of the vessel. Itis assumed that the maximum inclination angle for damaged craft, and the maximum heel angles with wind heeling or passenger crowding, are intended to provide a level of safety for passengers moving about, or evacuating from, the craft. This may also be the intention of the requirement for a minimum GM for intact monohull, but this is a standard IMO criterion and the philosophy behind its application in the HSC Code is not known, It is assumed that the other criteria are intended to provide for the safety of the vessel. 20,2 Intact Vessels ‘The high GZ values typical of multihull vessels appear to be recognized as, unlike monohulls, there are no requirements for minimum GZ values. ‘The GZarea requirements differ for all eases where the maximum GZ occurs at an angle below 30 degrees, which will be the case for most multihulls. Both incorporate an inereased GZarea requirement, but the minimum values are calculated by different formulae. The minimum requirement for multihulls is greater than that for monohulls. In addition, monohulls require a minimum area between 30 and 40 degrees, or the downflooding angle. In cases where the maximum GZ, occurs at less than 30 degrees, the area between that angle and 30 degrees is not regulated. ‘The minimum angle of maximum GZ, is 10 degrees for multihulls, and 15 degrees for monohulls Both monohulls and multihulls are limited to 10 degrees under the action of passenger crowding. ‘To address wind heeling for monohulls the IMO weather criterion is applied. For rmultituls, there is a ‘maximum heel angle of 10 degrees in a gust, and a minimum residual GZarca with combined wind heeling and passenger crowding, ‘The requirements for intact multihulls therefore are somewhat simpler than those for monohulls, and more cffective in terms of the assessment of residual stability. 20.3. Side or Bottom Damage For monohulls there are minimum requirements for GZ, GZatea, equilibrium inclination angle and range of positive stability, For multihulls the requirement is limited to the equilibrium inclination angle, which is the ‘same as that for monohulls of 10 degrees. With heeling moments applied, the only requirement for monohulls is a minimum residual GZ, value. For multihulls there is limit on the wind heeling angle and the residual GZarea. This residual GZarea limit is the same as for intact multihulls, but is determined with a lower wind heeling moment appli 20.4 100% Raking Damage ‘The criteria are identical for monohulls and multihulls, and include an assessment of the GZarea, inclination angle and range of positive stability, 25 WOLFSON UNIT 20.5. Intermediate Stages of Flooding During intermediate stages of flooding, before the equilibrium damage condition has been reached, the criteria are identical for monohulls and multihulls. They include an assessment of the GZ, GZarea and range of positive stability 21 COMPARISON OF VESSEL TYPES USING EXISTING CRITERIA 21.1 Methods of Analysis Most of the criteria in the HSC Code do not assess non-dimensional parameters and, because of the effects of scaling introduced in Section 18.2, itis difficult to make comparisons between vessel types on the basis of existing criteria, (One might attempt to use the data presentation in Figure 45 using the GZarea, up to 30 degrees or the angle of maximum GZ. for example. The eriteria for intact monohull and multihulls are similar but, generally, ‘multihulls that only just comply will tend to be smaller than monohull that just comply, and so the multihulls will be more vulnerable because of their size. This method would have the advantage that the most vulnerable test cases would be used to make the general comparison, but the implication of the eraph, that monohulls are much safer, is invalid because the effects of size would be neglected. Alternatively, one might compare sample test cases of similar size, This is dangerous because the cases selected might not represent the most vulnerable configurations. One might therefore be comparing a particularly good multihull configuration witha particularly vulnerable monohull configuration. Any conclusions drawn could be highly misleading. ‘Comparisons therefore should enable some judgement regarding the spread of data in terms of the variation Of wave heights to capsize for a particular vessel type, and should include the effects of size. 21.2 Comparison Using Scaled Model Capsize Data Figure 48 presents the results for the models that capsized, scaled such that their stability just complied with ‘one or more of the applicable Code criteria, The data are presented in terms of the variation of minimum ‘wave height to capsize with length. In some cases, the models were tested with the stability critical at the nominal vessel scale, and so there are clusters of data points at certain lengths. Where the stability at the nominal scale had a margin, a smaller scale was used to eliminate the margin. Conversely, ifthe stability at the nominal scale failed to comply, the scale was increased. ‘The trend of increasing wave height with size is clear, a isthe scatter due to variations in level of safety, even in the case of individual models. ‘The upper graph includes only those cases where the critical criteria were those that address the safety of the vessel, and excludes eases where the critical criterion was solely one addressing passenger safety, that is a heel angle limit. The upper graph therefore excludes all test cases where criteria numbers 6, 8 or 9 in ‘Appendix I were critical, ‘The Lower graph inelucls the cases where passenger safety criteria were critical. Because scaling of the ‘model data does not affect the heel angle, these cases can be scaled to any size above which the GZ properties just comply with the other criteria. The cases have been added to the graph at a number of lengths, the smallest of which corresponds to the limit due to a vessel safety criterion. ‘The upper graph suggests that the intact vessel criteria provide a higher level of safety than the damage ia, and it may be argued that this is reasonable on the basis that a combination of damage and severe conditions has a low probability of occurrence. The lack of a distinct separation between the monohull and rmultihull data suggests that the criteria generally provide similar levels of safety. 26 WOLFSON UNIT As in the other plots of capsize data, these graphs present only the minimum wave heights to capsize, Greater ‘wave heights were required at wave periods at which the model was less vulnerable, and some models could not be capsized in the waves generated. The latter, safe vessels, cannot be plotted because the critical wave heights are not known, ‘The solid line on each graph bounds the upper extent of the permitted operating limits of existing vessels for which data were available, The line forms the upper boundary of the data presented in Figure 4. It is ‘worrying that all but three of the capsize results lie below, and most are well below, the maximum height in which vessels are permitted to operate, particularly as waves encountered may be much higher than the significant height. 1 for different vessel types, but none provide an adequate level of safety. 22 IMPLICATIONS OF THE RESULTS FOR HSC CODE REVISION ‘On the basis of Figure 48, it may be argued that the criteria should be increased by a substantial factor. It is recognised that the fleet generally has good margins of stability over the criteria and, as there is an absence of capsize incident data, the fleet is generally safe. This study indicates that this good safety record is not due to the Code stability requirements, and therefore they might be increased without a significant impact on the fleet. It might be inappropriate, however, to increase all of the HSC Code criteria by a large factor on the strength of this finding, since a design may have good margins of stability with regard to most criteria, but may have small margins over one criterion, Increasing all criteria, regardless of their importance in ensuring safety from capsizing, might unnecessarily influence some designs. ‘The philosophy of existing sets of criteria is that, because GZ, curves vary considerably in shape, one must regulate a number of characteristics. To do this effectively itis necessary to identify the relative minimum requirements for each parameter that provide equivalent levels of safety in all cases. ‘An alternative approach would be to develop a method of assessment that identifies those stability parameters that govern the level of safety, and assess them in combination. This is the approach that is followed in the following sections. 23 NON-DIMENSIONAL STABILITY PARAMETERS AND WAVE HEIGHT TO CAPSIZE 23.1 Comparison of Wave Height with GZ Values Minimum wave height to capsize was plotted against GM, GZmax, and GZarea in Figure 45. Non- dimensional versions of one of these parameters are presented in Figure 49, to show the effects of normalising with respect to length and beam, On these plots, scaling the model data has no effect on the location of the points. There remains a clear distinction between the monobhulls and multihulls, but the plots do not enable a simple comparison of the level of safety offered by a particular value of GZmax, as used in conventional criteria 23.2. Comparison of Wave Height with Range of Stability Because range of stability appeared to be an important parameter, and is non-dimensional, it was necessary ‘to compare it with the minimum wave height to capsize in non-dimensional form. A number of methods were considered, and a simple ratio of wave height to a length parameter was selected. Various options were considered for the length parameter, including the following: 1. Length. This is the simplest representation ofthe size of the vessel. 2, Beam. This was used because the majority of capsizes occurred in beam seas, and the beam might have an important influence on the capsize mechanism. (Length x Beam)’, This enabled their combined effects to be included, |. (Length x Beam"), This was used because it relates to the waterplane inertia, and might therefore have an important influence on the capsize mechanism. 20 WOLFSON UNIT | 5. Displacement "*, This was used because displacement is a factor in the righting moment. { ‘The variation of wave height, normalised with respect to three ofthese parameters, with range of stability is presented in Figure 50. All three suggest a trend of increasing wave height with range. As with the graphs of | wave height against GZmax and GZarea, one can argue that it is the lower limit of the data that defines the | safe boundary. With the wave height normalised with respect to length, there is a distint difference between the monobull ‘and multihull data. The monohulls appear to be more vulnerable with a given range of stability because the lower limit oftheir envelope of data is lover than that for multihulls, With the wave height normalised with | respect to beam, however, the data collapse into a common envelope, the lower limit of which is represented l by both monohulls and multihulls, Using the product of length and beam resulted in an intermediate result, as might be expected, with resonable collapse ofthe data but some separation remaining between monohulls ' ‘and multihulls, Normalizing with respect to (Length x Beam!) , and Displacement , gave a similar result to that of normalising with respect to beam, with a common envelope of data and a similar degree of scatter. ‘A.common feature of the three graphs is that the trend of increasing wave height with range of stability | ‘appears to be non-linear, with some models with larger ranges of stability requiring substantially greater ‘wave heights to capsize. The trends suggest that the minimum wave height required to capsize is proportional to the square, or some greater power, of the range. f 233. Discussion This exercise showed that the vulnerability to capsize varied with the stability parameters, but scatter in cach plot was believed to be due to the influence of a stability parameter not included, For example, the scatter in the plot of wave height/length against GZmax/length appeared to be due, predominantly, to variation in the range of stability. In attempting to derive a useful capsize boundary, therefore, relationships were sought between the minimum wave heights to capsize and combinations of those parameters that appeared to be most important 24 THE ROLE OF DISPLACEMENT In Section 7, the concept was introduced that one model configuration might represent both an intact, and a ‘damaged configuration, Where damage results in total flooding of a compartment, there is no free surface and the condition is the same as if the compartment were full of ballast water, which would be included in the loading condition. The displacements and centres of gravity therefore would be different, The righting moments of the vessel would be the same but, because of the convention that the damaged GZ values are derived from the righting moment divided by the intact displacement, the GZ. values derived for the damaged ‘configuration are greater than forthe intact vessel. “The definition of the vessel condition has no effect on the wave height to capsize, but a substantial effect on the ratios of wave height to GZmax and GZatea. Both configurations have the same range of stability. For the case of the catamaran considered in Section 9, the following data apply: Righting Wave Height ID Displacement VCG GM Momentmax Gzmax GZarea Range List to Capsize tonnes. = mm tonne,m m ——maradians degrees degrees m 31 93 576 628 9 2 023 0028 = 11220 12 Bla 1703.69 2.53 012 001s 1220 12 Itis clear from this comparison that, ifthe vessel is assessed on GZmax and GZarea alone, the damaged configuration, ID 31, is considered to have about twice the stability of the intact configuration, because its displacement is 55% of the latter. 28 WOLFSON UNIT In order that a valid comparison can be made between these two configurations, the assessment of stability ‘must include the displacement. Only by comparing their righting moments, or their range of stability, are they assessed, correctly, as being equivalent, This exampl waterline condition, if judged aj I result in a more opti st GZ, criteria alone. 25 ‘THE IMPORTANCE OF RANGE OF STABILITY Conventional criteria tend to regulate GZmax and GZarea, and exclude the range, but it is unclear whether these are the most appropriate parameters to assess, or whether they have been deemed largely successful merely because they provide a reasonable estimate of the range. The comparison of GZarea with Range x GZmax in Figure 44 demonstrated that, for most vessels willt low stability, the two characteris similar. The trimarans were the exception, and their data enable an that governs the vulnerability to capsize, or the combination of GZmax and range. Some example tests have ‘been selected for comparison in the following table. Non-dimensional values must be used to eliminate size ‘and provide valid comparisons based on stability alone. vestigation of whether it is an extreme one, but any arrangement that enables flooding of compartments below the ic assessment of the stability than for an equivalent intact floatation 1D Disp. GZmax fonnes metres metreradians degrees Example 1 33 300 0.340 38 93 0.608 43 1100 0.785 Example 2 50 175 0.486 41 30.760 Giarea 0.146 ois 0.196 0.231 0.187 Range GZarea/l, GZmax/L 65 17 22.0 639 2.6 0.0028 0.0035 0.0028 0.0044 0.0087 0.006 0.019 oon 0.009 0.023, Range x Za, 0.406 0327 0246 0,602 0,527 Max Righting Mom/L!* x10 0.013 0.049 0.034 oon 0.062 Wave Height/L. ‘No Capsize 0.040 0.032 ‘No Capsize 0.058 In Example 1, model ID 53, a trimaran that was not capsized, is compared with models ID 38 and ID 43, catamarans 2‘and 3, which capsized in wave heights of 4% and 3% of their length. The trimaran did not capsize despite having the same GZarealL. as [D 43, and only 80% that of ID 38, It had a greater range x GZmax/L, despite a much lower GZmax/L.. This indicates that it was the range of stal safety from capsize. ty that provided the Example 2 shows two other model configurations, again a trimaran and catamaran, with results that support the finding. In this example, the trimaran had 76% of the GZarealL and only 39% of the GZmax/L of the ‘model that capsized. The latter, however, had only 34% of the range of positive stability of the trimarai In Section 24 it was argued that displacement is important and, in these examples, comparison of the non- dimensional maximum righting moments is similar to the comparison of the maximum GZ values. ‘These examples suggest thatthe range of st GZarea provides an estimate for it 26 DEVELOPMENT OF A SAFE BOUNDARY Other researchers have derived capsize boundari height and vessel parameters. These reveal the dependency of the level of safety on certain design x is the important parameter, and that use of the parameter from non-dimensional plots of various ratios of wave parameters. One such example, taken from Ref.4, is H/P against GM.Cp.T/B? where Hs significant wave hi F is freeboard block coefficient mean draught 29 WOLFSON UNIT B is beamat the flooded waterline ‘This presentation does not lend itself to the data from these tests because, in some cases, the freeboard was negative, but variations on the theme were studied. In an attempt to derive effective safe boundaries the test results were plotted on various graphs, using many combinations of parameters, seeking a linear trend and good collapse of the data, The preceding analysis indicated the importance of size, righting moment and range of positive stability, and these parameters were given prominence in the exercise. One of the more successful plots is presented in Figure 51, The stability function used here is the product of the range of positive stability in degrees and the square root of maximum righting moment in tonne metres. It is normalised by the product of length and beam, and plotted against the minimum wave height to capsize, normalised with respect to length. The purist will argue that this function is not strictly dimensionless, and that it should include water density and a factor to convert degrees to radians. These adjustments would not affect the result, as they are constants, and the benefits of maintaining values in degrees and tonnes are that they are the values familiar to naval architects, and provide a numeral of more convenient magnitude, An important feature of this graph is that several types of model configuration are represented near the lower ‘boundary of the data envelope, suggesting that there may be a common safe boundary forall types of vessel, Using various combinations of these parameters, it was possible to collapse the data into a relatively narrow cor common envelope, such that the lower boundary of the envelope was linear, or such that the lower ‘boundary was represented by several types of model configuration, Unfortunately, an expression that achieved all of these goals proved elusive. ‘The distribution of data points may be adjusted by selection of different expressions for the axes of the graph. Raising the range to a higher power has a significant effect on the linearity, for example, while the choice of length, beam or the product of the two, in the denominator affects the relative distribution of the monohull and multihull data. Whilst such massaging of the relationship affects the relative distribution of the data points, it has less effect on the safe boundary. ‘The data are plotted again in the lower graph of Figure SI, categorised according to the mode of capsize. ‘There is little distinction between the modes, and so the relationship used in this presentation appears to be equally applicable to the various capsize mechanisms and to the types of vessel. This result is rather ising, considering the wide range of vessel proportions and extents of flooding, and the variety of capsize mechanisms. ‘This result is very attractive from the point of view of stability assessment, since it suggests a common approach for all types of vessel, and simplification of stability criteria, 27 DEVELOPMENT OF AN ALTERNATIVE METHOD OF ASSESSMENT 27.1. The Value of the Test Results ‘Most stability criteria have been developed using a limited number of real casualties as a basis for determining what might be used as a boundary below which vessels may be considered vulnerable, Fortunately, stability casualties are rare, and well documented examples are rarer still. Stability casual usually involve a number of parameters that are dificult to define, including the loading condition and environmental conditions at the time. This test programme has provided a database of “casualties”, with their ‘vulnerability measured in a controlled environment. The tests have resulted in the minimum wave height required to capsize each configuration, and thus differ again from real casualty data. 27.2. Proposed Formula ‘As noted in Section 26, adjustment ofthe relationship has only a subtle effect on the safe boundary. The relationship presented in Figure 51 is a simple one in comparison with some that were studied, but this 30 WOLFSON UNIT simplicity offers an advantage in a regulatory environment, and does not appear to detract from the collapse oftthe data, On the lower graph of Figure 51, a solid line has been added to demonstrate the potential of this analysis to provide a method of stability assessment, The line represents the formula: Range /RMimax 108 Where: Wave Height is the minimum wave height that might result in capsize Range is the residual range of positive stability in degrees RMmax is the maximum residual righting moment in tonne. metres Bis the overall beam of the vessel WaveHeight = Formula 1 ‘The length of the vessel has been eliminated from the formula because it was used to normalise the expressions defining both axes of the graph. This normalising process was required in order to plot the test data in non-dimensional form, but is not required when applying a formula for a safe boundary, which has been derived from the data. Formula 1 is a very simple expression, but does not form a lower boundary to all of the data points, An alternative line, shown dashed on Figure 51, would provide a more conservative method of assessment. It represents the formula: Range YRMimax 108 0 Formula 2 WaveHeight = ‘The following table presents dimensions and hypothetical values of GZmax for typical high speed monohulls, ‘and multihulls of three sizes. The final columns indicate the ranges of stability that would be required to comply with Formulae | or 2. The table illustrates the effects of size, hull form and operational wave height on the required range, given a constant value of GZmax. Note that the values of GZmax and range are the residual values after allowance for any wind heeling or passenger crowding moments. ‘Vessel Type Hull Long Beam Displacement Wave Height GZmax Range Required Range Required m m tonnes m m degrees degrees Formula 1 Formula 2 Monohull 20 3 40 1 02 18 21 Monohull 20 5 40 2 02 35 39 Monohull 50 9 250 2 02 25 32 Monohull 100 1s 1000 4 02 2 33 Catamaran 20 7 40 Il 1 u 1B Catamaran 20 7 40 2 1 2 By ‘Catamaran so 4 250 2 1 1B 2 Catamaran 100 26 1000. 4 L 3 41 27.3. Choice of Wave Height ‘The wave heights used in the table may be considered to represent typical values of significant wave height in which vessels of these sizes might operate. Since wave heights greater than the significant height may be ‘expected, it would be prudent to select a value greater than the significant height in which the vessel is permitted to operate. The value selected will introduce a probability factor into the assessment method. As described in Section 14.2, the greater the wave height, the lower will be the probability of encounter. Ifa wave height of twice the significant were selected for the purposes of assessment by this method, it ‘would represent a likelihood of an encounter once ina few hours. Increasing the selected wave height further by a factor of 25%, to 2.5 times the significant, would extend the likely period for one encounter by a factor 31 WOLFSON UNIT of ten. Thus, in approximate terms, an increase of 25% in the value determined from Formula 1 will reduce the probability of capsize to 10% of its previous value. This demonstrates the value of size and good stability characteristics, in thatthe level of safety increases in greater proportion than the value of Formula 1 Ifthe wave height in which a vessel is to be certificated as “safe from capsizing” is increased by a factor of 2, the required range is also increased by.a factor of 2, assuming that the GZmax remains the same, In ‘contrast, if the same increase in wave height were required while the range remained constant, GZmax would need to be increased by a factor of 4. In practice, one cannot increase the range of stability of a design without affecting one of the other parameters, and so such a simplistic comparison is largely academic, ‘Assuming that a capsize wave height equal to twice the significant is considered appropriate, the method of assessment would incorporate one of the following versions of the proposed formulae to define the maximum permitted operational seastate: Range /RMimax Significant WaveHeight 2B Formula 1a Ri IRM cantante = EVEN. Forman 2a For vessels operating in protected waters, a minimum wave height requirement could be specified to ensure ‘that the residual stability is sufficient for the vessel to survive encounters with wash generated by other vessels, 274. Other Data “The HSC fleet casualty statistics do not provide data against which this method may be judged. MCA Research Project 504 studied incidents to the international HSC fleet and identified those related to stability ‘and buoyancy. The study identified few capsizes, and no reliable data were available for vessels that capsized while stationary in a seaway, in an intact or damaged configuration, Whilst this method has been developed through tests on models of high speed craft, their diverse range of hull forms, and their stationary attitude, lead to the expectation that it may be applied equally to other types ‘of vessel. Model test results, or casualty data, on other vessels therefore were sought for comparison. Much work has been conducted by the Wolfson Unit on capsizing in breaking waves, but such a mechanism is not relevant to this study because the seastates required would be well beyond those in which the vessels ‘operate, Some work on self righting of capsized rescue craft is considered relevant however. Models were placed upside down in the towing tank, and exposed to sea spectra to study the relationship between the ‘wave height required to self right and their inverted stability. Since rescue craft are designed with very high ranges of stability, often to 180 degrees, the inverted stability ofthe tested models was low. The conclusion of that study was that self righting ability was related to the range of stability and wave height, For the purposes of this comparison therefore, a successful self righting represents a “capsize” of the inverted model. ‘An inverted rescue craft floats on its wheelhouse and cannot be considered to representa typical HSC, but these test results enable a check on the hypothesis that this method may be applied to any floating body. “There have been a number of well documented cases of capsizes in the UK fishing vessel fleet, and in some ‘cases the vessels capsized in a seaway as a result of low stability. Data were extracted from MAIB incident investigation reports on six vessels ranging from 6.5 to 26.3 metres in length. All of these model and full scale data have been derived from capsizes in irregular seastates. The wave height thet caused the eapsize is not known, but the significant height of the seastate was Known in the model tests, and estimated for the fishing vessel casualties. These cases are presented in Figure 52, a graph with the same axes as the previous figures. The envelope of test data, and the line defined by Formula 1, are included 32 WOLFSON UNIT for comparison. The data points all lie above the Formula | line, most of them higher than the envelope of test data from this project. ‘There are two important differences between these cases and the test data that might explain this, |. The fishing vessels were in irregular seastates, with wave heights estimated from other environmental conditions, or estimated by observers. Both of these sources frequently provide overestimates of seastates, but the wave that resulted in capsize might have been greater than the significant height. The data therefore may be over estimates but can only be regarded as approximate. 2. The tests conducted for this study involved the equivalent of an optimisation process. The models ‘were tested at all headings in a range of wave periods to find the minimum wave height that could capsize the model. The other data do not represent the minimum wave heights to capsize, Bearing in mind the second of these in particular, the data appear to support the model test findings. 28 CONSIDERATION OF DOWNFLOODING Investigation of downflooding was not included within the objectives or suggested work programme of the project, and the model tests were not designed to adress the issue directly. It was raised as an item of | interest during early project progress meetings, and so was included as an item for consideration during general observations of the tests and their implications, ‘On some early model configurations, downflooding pipes were fitted. These allowed flooding of an intact, compartment following immersion of the opening, or intermittent flooding by wave action. Open vertical pipes of adjustable height were fitted through the model deck. When the attitude and motion of the model resulted in water entering the pipe, the resulting reduction in freeboard increased the likelihood of subsequent downflooding, as would be the case for a full scale vessel. It was apparent that a simple marker could be used in place of an open pipe, with careful viewing of the video record enabling the level ofits immersion to be monitored. ‘The local wave elevation is dependent on the flooded waterline and heeling moment, the wave height and period, and the attitude of the vessel to the waves. Observations of the models showed that local wave action also was highly dependent on the local structural arrangement. The distance of the opening from the side of the vessel, the height above deck, whether horizontal or vertical, the degree of protection by bulwarks or adjacent superstructures, and, if vertical, the in which the opening faced, were all factors that might influence the vulnerability to intermittent downflooding by wave action. In this project, aimed at capsize assessment on simplified and generalised models; such a large range of variables could not be incorporated. For the same reasons, itis not considered worthwhile for a safety standard such as the HSC Code to attempt detailed assessment of downflooding by wave action. Generalised requirements for permanent openings, of those used for evacuation, to be a specified height above the damaged waterline, with sill heights at a specified height above the deck, are simple to apply and likely to be effective. As for stability assessment, the damaged waterline used must be that resulting after the application of any anticipated heeling moments, including passenger crowding and wind heeling, Inthe 2000 HSC Code, section 2.6.11, the height of openings above the waterline is related to the anticipated operational wave height, and this is considered to be appropriate Standard sill heights have been proven on many types of vessel operating with low freeboard, and are ‘accepted as providing adequate protection from water on deck. Evacuation of passengers with varying agility places conflicting demands on openings, requiring sill height requirements to be relaxed on passenger ‘vessels. The appropriate compromise is not within the scope of this project. 3 WOLFSON UNIT Itis difficult to generalise on the preferred location or arrangement of openings. Those that are protected by other structures will be less likely to suffer wave impacts that might lead to downflooding or injury, but a more complex arrangement might impede evacuation, As described in Section 15, the implications of these tests are that the combination of vessel configuration, stability and wave parameters govern the angle to which the vessel will be rolled, and the statical stability determines whether the vessel subsequently rights or capsizes. If serious downflooding would oceur at an angle lower than the residual range of stability, the stability curve should be truncated at that angle, as is the case in the HSC Code methods of assessment. 29 CONSIDERATION OF WIND HEELING The findings of Research Projects 503 and 537, studies of the wind heeling characteristics of multihulls, were that the method of predicting the moment used in the 2000 HSC Code would benefit from some adjustment. Revised formulae were proposed with which to estimate the effective wind pressure, the variation of heeling moment with heel angle, and the gust factor. ‘The principal differences resulting from the revised formulae are: 1. The estimated wind heeling moment is increased for vessels of high bearrvheight ratio. 2. The heeling moment reduces with heel angle for vessels with low beanvheight ratio, and increases with heel angle for vessels of high beam/eight ratio. 3. A gust fuctor of 2 on wind pressure is assumed, rather than 1.5 as in the 2000 HSC Code, Itis recommended that the proposed formulae be used to estimate the effects of wind heeling, in order to enable a realistic, or a conservative, estimate of the residual stability ofa vessel. ‘The dynamic effects of gusts were found to be insufficient to warrant consideration by the HSC Code, Generally, multihulls respond in a quasi-static way to the wind pressure in the gust, and the gust wind pressure should be used when estimating the heeling moment and residual stability. 30 CONSIDERATIONS OF PASSENGER SAFETY ‘The 2000 HSC Code intact stability criteria include a requirement that the maximum angle under passenger ‘crowding should not exceed 10 degrees. The damage criteria include a similar requirement for the maximum inclination angle. With 100% raking damage, the inclination angle requirement is relaxed to 20 degrees. ‘These requirements do not provide any particular level of safety for the vessel, but ensure safety for the passengers in terms of their security and movement about, or evacuation fom, the vessel. They should not be confused with the other stability criteria that are intended to address vessel safety or survival. The proposals made in this report do not address passenger safety and should not affect these requirements. ‘Tests with the trimaran highlighted a potential problem with criteria in assessing the stability when a GZ. ‘curve has two maxima and a pronounced minimum value below the angle of downflooding. The curve for ID 53 is an example, See Figure 19. It might occur in a damage configuration, with wind heeling and passenger crowding giving a heel angle of 9 degrees in calm water. The curve indicates a very good range of stability and value of maximum GZ, because of the large second peak. If stability is progressively reduced, the curve ‘ends towards that of ID 52. At some point, it becomes tangential to the x-axis, after which the range and maximum GZ. values are dramatically reduced to those given by the fist peak. This is the point at which the maximum allowable KG would be defined, In practice, a vessel that has a very small GZ. value at the local minimum will comply with the criteria with ease, When in a seaway, however, it is likely to heel quickly to ‘the angle at which the second peak provides a substantial righting moment, inthis case about 30 degrees. Such an attitude would not ensure safety for the passengers, and a vessel with these characteristios should be given careful consideration by the regulatory authority. 34 WOLFSON UNIT 31 IMPLICATIONS FOR THE HSC FLEET ‘Whilst there have been few capsizes within the world’s HSC fleet, the statistics should not lead to complacency with regard to stability. Capsizes usually result in very high mortality rates, and capsize of a passenger vessel would be a major incident. Stability booklets for some existing HSC were inspected to obtain the parameters required for the application ‘of the proposed method for deriving a maximum operational seastate on the grounds of stability. These included two monohulls, two catamarans, a trimaran and a SWATH and, in some cases, damage stability data were available, In some cases, estimates of residual stability were required because the appropriate information on heeling moments or downflooding angles was not available, but the examples serve as a reasonable indication of the implications. The wave height was calculated in accordance with Formula 2a, ‘and the data are presented in Figure 53. The figure includes the available data for the maximum permitted ‘operativial seastates of a number of HSC for comparison. These are as presented in Figure 4, The solid line ‘on the graph bounds the upper extent of the permitted operating limits of existing vessels for which data were available. ‘The data for intact vessels generally lie above the line indicating the highest current operational limits, This implies that each vessel assessed by this method has adequate intact stability for the seastate in which a vessel of that size might be expected to operate, and that use ofthis method would not affect the current operation of these vessels. A line drawn through the lowest data points is labelled as the “implied limit for intact vessels”. The shape of this line, suggesting that 40 metre vessels are more vulnerable than 20 metre vessels, is an unlikely result that demonstrates an inadequate number of vessels in the sample, ‘The damage stability for the vessels studied is lower than their intact stability, and so the derived wave height to capsize is lower. In most cases, this is because of a relatively low downflooding angle resulting in a low effective residual range of stability. Most of the data points lie within the envelope of operational seastate limits. This implies that the vessel will be vulnerable if damaged, but only ifthe following conditions are all met It encounters the highest seastates in which a vessel of that size typically is permitted to operate. It encounters waves of twice the significant height. ‘These waves have a period to which the vessel is most vulnerable to capsize. ‘The vessel’s heading to the waves is that at which the vessel is most vulnerable to capsize. ‘The vessel is subjected to the maximum anticipated wind heeling moment, in terms of wind speed and heading, ‘The vessel is subjected to the maximum anticipated passenger crowding, or other, moment, ‘The vessel configuration is as vulnerable as the worst examples of those tested. ‘The probability ofall of these conditions being met simultaneously is very low, and perhaps this suggests that the assessment method might be relaxed for damaged cases, as is customary with conventional criter For example, iit were considered acceptable for the maximum permitted operational seastate to be the same as the predicted minimum wave height to capsize, from Formula | or 1a, this would increase the values of maximum permitted seastate by a factor of 2. The data for these sample vessels suggest that such a ratio of 2/1 for the intact/damage stability would bring them roughly into alignment in terms of the mai permitted seastate, This exercise indicates that the proposed method has the potential to provide an acceptable means of regulating stability in relation to the size of vessel and antici however, that the sample of vessels used here is very small with some data estimated, and a much wider trial ‘ill be required in order to ensure that there are no apparent anomalies for particular vessel configurations. Ifsuch a tral indicates that a large percentage of vessels fail to comply with the proposed criterion, it is likely that the level of the criterion should be adjusted, because there is no indication that the fleet in general 35 WOLFSON UNIT operating at an unsafe level of sta icates that a small number of vessel fail t0 comply, it may be because such vessels are relatively unsafe but have not yet suffered casualties, or because some characteristic not identified by this study provides additional safety. If their design differs substantially from the models tested, there may be grounds for additional mode! tests. Figure 53 includes predictions of the maximum permitted significant wave height forthe fishing vessel ‘casualties described in Section 27.4. The predicted values are lower than the seastates in which the vessels ‘were reported to have capsized, but the method suggests that they would all have been vulnerable in waves of significant height less than half a metre. t should be noted that the stability data are for the casualty condition that, because of flooding or offSet loading, may have been somewhat worse than a standard condition in the stability booklet. 32. SUMMARY OF FINDINGS Almost 800 test runs were conducted on 6 models, in 53 different configurations of damage and loading condition, Modelling of damage stability configurations is problematic because model hull thickness and structures do not represent those of full scale vessels, even if constructed to minimum practical dimensions. If mode! tests are conducted to investigate a particular vessel stability condition, the stability of the proposed model configuration should be ealculated, and checked by measurement afloat, to ensure that it represents ‘accurately the full scale condition. In general, models will have greater stability characteristics than the full scale vessels they are intended to represent. It may not be possible to model correctly the displacement, stability and inertia properties ofa particular vessel, and predictions of capsize vulnerability based on model tests are likely to be optimistic. ‘The capsize mechanisms were highly variable, with the most vulnerable heading and capsize mechanism being unpredictable prior to testing each configuration. The capsize mechanisms and most vulnerable headings varied with wave frequency, The capsizes were categorised by five principal capsize modes. Some configurations were found to be more vulnerable in the transition phase ofa test, when waves were ‘encountered prior to the natural drift rate becoming established. This suggests that an encounter with a discrete group of waves, in otherwise calm conditions, might be more onerous than sustained exposure to an equivalent seastate. This finding may have implications for stationary small craft encountering the wash of larger vessels “The minimum wave height required to capsize a vessel varies with the wave frequency. In this study, capsize boundaries were defined to identify this variation and, from these boundaries, the minimum wave height required to capsize each model was determined. For most configurations with stability that just complied with the Code, the minimum wave heights to capsize were found to be within the range of typical operational seastates. The existing Code criteria, therefore, do not provide immunity against capsizing under normal operational conditions. ‘The wave height to capsize is dependent on the stability and the size of a vessel. An increase in ether of these will result in an increase in the wave height required to capsize, Scaling of all parameters of a vessel results in an increase in the size and stability, and consequently with the same increase in the wave height required to capsize. In practice, large vessels tend to have similar GZ values to small vessels, and therefore have the benefit of size but with less stability per unit length, Conventional regulations set constant GZ. criteria regardless of the size of vessel or the anticipated ‘operational seastate. In effect, this relies on a balance between size, stability and seastate that is assumed rather than assessed, 36 WOLFSON UNIT ‘when upright, in the absence of anticipated heeling moments, do not pr 8 valid assessment of safety from capsizing. Maximum anticipated heeling moments due to wind, passenger crowding, and any other significant weight movements, should be estimated and applied in order to predict the minimum likely residual stability. The residual stability curve should be assessed against the selected criteria, In the HSC Code, the requirements for intact monohulls are an example of such a lack of assessment of the residual stability, but this principle applies to regulations for all types of vessel. ‘The stability parameters that govern the wave height required to capsize are the maximum righting moment and the range of positive stability, The results of these tests indicate the range to be the more important of these two parameters. ‘The criteria in the 2000 HSC Code, and indeed most conventional criteria, are based on GM, GZarea, GZmax and the angle at which GZmax occurs. These criteria have been in use for many years and have gained widespread acceptance, This study suggests that they provide an approximation to the level of ‘vulnerability by approximating the essential parameters. GM is an indicator of GZmax, but the relationship is not reliable. GZarea is closely related to the product of range and GZmax for most vessels but there are anomalies, particularly vessels with unorthodox hull forms, such as trimarans. The angle of maximum GZ is an indicator ofthe range of stability for most vessels but there are anomalies, as for GZarea Assessment based on GZ. values is convenient because naval architects are familiar with the criteria values, ‘which apply to ships of all sizes. The method enables all vessels to be presented on a common graphical scale, ‘The 2000 HSC Code criteria for multihulls differ from those for monohull, in terms of not only the values of specific stability parameters, but in the parameters assessed, and the consideration of heeling moments. ‘Comparison of the levels of safety afforded to monohulls and multihulls by the existing criteria is not possible because of the effects of scale, and the different approaches to the assessment of stability. One might compare specific vessels but not generic groups of vessels. Multihull vessels tend to have large margins of stability over the Code minimum requirements, and generally are less vulnerable than monohulls to capsize by wave action. This characteristic of actual vessels should not bbe confused with the levels of safety provided by the Code, but may influence any revision of the minimum criteria. [tis likely that the minimum requirements for multihulls may be increased, to provide improved minimum levels of safety, without significant impact on the compliance of contemporary vessels, {A formula has been developed that relates the minimum wave height to capsize to the beam of the vessel, its range of stability and maximum righting moment, regardless of the type of vessel or mode of capsize. Whilst this formula does not predict the minimum wave height precisely in all cases, it errs on the conservative side in most cases, and therefore offers potential as a means of safety assessment. The formula recognises that the range of stability is relatively important in comparison with the maximum GZ, of righting moment. An alternative formula that would provide a more conservative estimate of the minimum wave height to capsize is suggested for consideration, ‘The range of stability should be defined as the residual range with all anticipated moments applied, with the stability curve truncated by the downflooding angle if appropriate Conventional assessment of damage cases uses the intact displacement to determine the GZ. value from the calculated righting moment. This results in greater GZ. values than for equivalent intact conditions, where bottom damage results in compartments being pressed full of floodwater, The proposed method avoids this inconsistency. ‘The 2000 HSC Code stability criteria include heel angle requirements to ensure safety for the passengers rather than the vessel. The proposals made in this report do not address passenger safety and should not 37 WOLFSON UNIT affect these requirements. There may be cases however, particularly with trimarans, where a safe heel angle under wave action is not assured. 33. RESPONSE TO THE OBJECTIVES ‘The project objectives were listed in Section 2, and the following paragraphs are numbered to correspond with the numbering of the objectives. 1. Critical capsize modes were found to be dependent on a number of parameters, including the wave frequency. Various capsize mechanisms were observed and grouped into five categories. The tests did not provide a method by which the critical capsize mode can be predicted for a particular vessel configuration, but the critical wave height does not appear to be dependent on the capsize mode so this is not considered to be a problem in terms of assessment, 2, ‘The existing requirements for intact monohulls are flawed because they do not address residual stability after passenger crowding. For a vessel with a low angle of maximum GZ the residual stability after passenger crowding may be negligible. Such a vessel may be capsized by waves that are small in relation to typical operational seastates, ‘The damage stability requirements for monohulls require a minimum residual GZ, and therefore provide a better assessment of safety than the intact requirements. It is not possible, however, to establish the critical environmental conditions for which the requirements provide sufficient stability because they are dependent on the range of stability and size of the vessel, neither of wihich is regulated. ‘The requirements for intact and damaged multihulls include an assessment of the area under the residual GZ. curve, and therefore provide a better method of assessment than either the intact or damaged monohull requirements. This study has shown that the GZarea provides an estimate of the product of GZmax and range of stability, both of which are important for safety. Its not possible, however, to establish the critical environmental conditions for which the requirements provide sufficient stability because they are dependent on the size of the vessel, which is not regulated. 3. A method of assessment has been proposed that will provide equivalent levels of safety for ‘monohull and multihulls. A formula has been suggested that includes the righting moment, range of stability, size of the vessel, and seastate to which the vessel may be exposed, and which may be adjusted to provide the required level of safety. 4, Stability data were available for a small number of HSC of various types, and these were used to determine the likely implications of the proposed method on the operational limitations of the flee. Because the method is somewhat unorthodox, it warrants a more thorough study with a greater ‘number of vessels. ‘The method is simple, but may be adjusted to provide the desired level of safety. Precise evaluation ‘of the impact on the fleet therefore will depend on the level selected by the IMO or regulatory authority 5. Papers are to be drafted for the IMO, but are beyond the scope of this report. 38 WOLFSON UNIT 34 CONCLUSIONS L 10. 12, 4, ‘The 2000 HSC Code minimum eriteria do not provide immunity from capsize in typical operational seastates. ‘Multihull vessels tend to have a higher level of safety because they tend to have greater margins of stability over the Code requirements than monohulls, Ths isa reflection on typical vessel parameters and not on the levels of safety provided by the Code. Critical capsize modes are dependent on a number of parameters, including the wave frequency. ‘They are highly variable and unpredictable. ‘The minimum wave height required to capsize is dependent on the size of the vessel relative to the waves, and its stability. ‘The vulnerability to capsize by wave action is governed by the residual stability in the presence of any heeling moments, such as those due to wind or to passenger crowding, Assessment of the stability in the absence of such moments does not provide an adequate measure of the level of safety ‘The 2000 HSC Code criteria for intact monohulls do not address residual stability after passenger crowding, This neglect may lead to very low levels of safety in some cases. ‘The vulnerability to capsize does not appear to be dependent on the form of the vessel, the number of hulls, or the existence or extent of damage. Al configurations may be considered as floating bodies characterised by their residual stability curves. ‘The vulnerability to capsize is dependent primarily on the range of positive residual stability and, toa lesser extent, on the maximum residual righting moment. Conventional criteria provide a level of safety that has become accepted, but they set minimum, requirements for parameters that provide approximate measures of the range and maximum righting moment, rather than by addressing them dircetly. Conventional criteria set minimum GZ, values regardless of the size of the vessel, and therefore incorporate an inherent assumption that larger vessels operate in higher seastates. Comparison of the levels of safety provided for monohulls and multihulls by the 2000 HSC Code is not possible because of the differing parameters assessed, the differing account taken of heeling ‘moments, and the neglect of the effects of size of the vessel relative to the seastate, ‘A method has been proposed to assess the level of safety on the basis of the size of the vessel relative to the operational seastate, the beam, the range of residual stability and the maximum residual righting moment. A single formula may be applied to all forms of HSC, intact and damaged. The method is believed to be equally applicable to other vessel types. ‘The proposed method will provide equivalent assessment of different forms of vessel, and the formula may be adjusted to have an appropriate impact on the existing fleet. Some criteria provide a level of safety for passengers, generally by limiting the residual heel angle under the influence of heeling moments, and should not be confused with those that provide a level of safety against capsizing, 39 WOLFSON UNIT 35 RECOMMENDATIONS It is suggested that the following recommendations be put before the IMO for consideration, 1. ‘The HSC criteria should be revised to provide equivalent levels of safety for all types of vessel. ‘The criteria should take into account the size of the vessel and the seastate in which it is expected to operate. 2. The HSC criteria should be revised to provide, in all cases, an assessment of the residual stability after the application of the anticipated heeling moments. 3. The proposed! method and formulae should be considered for adopt existing criteria, in the Code in place of the 4, ‘The impact assessment exercise undertaken within this project was limited by the small number of sample vessels available, and should be widened through industry involvement at an international level. ‘The appropriate values of the minimum requirements for intact, damage, and 100% raking damage configurations may be selected to provide acceptable levels of safety, and may be influenced by the impact assessment. ‘The criteria numbered 6, 8 and 9 in Appendix | provide safety for passengers and should be retained with no difference between the values for monohutls and multihulls. 7. ‘The method of estimating the wind heeling moments for multibull HSC was addressed in MCA Research projects 503 and 537. The method proposed by that project should be incorporated into the Code, Consideration should be given to the wider application of the proposed method to other types of, vessel. It is understood that other researchers are studying the IMO Weather Criterion, but model tests conducted on ‘monohull sailing vessels and multihull motor vessels indicate that dynamic responses to gusts are not significant. Such research should be extended to monohull motor vessels to determine the extent of their dynamic response to gusts. Ifthe dynamics of gust response can be neglected it may be possible to eliminate the requirement for assessment against this, relatively complicated, criterion, 36 REFERENCES 1. Capsize and Sta ‘August 1995 2. Global Wave Statistics, N. Hogben, N.M.C. Dacunha, G.P, Olliver, Compiled and edited by British Maritime Technology Ltd., 1986 3, Probabilistic Theory of Ship Dynamies, W.G.Price and R.E.D.Bishop, Published by Chapman & Hall Lid, 1974 4, Experiments with a Floodable Model of a Ro-Ro Passenger Ferry, I.W. Dand, Second Kummerman Foundation International Conference on Ro-Ro Safety and Vulnerability - The Way ‘Ahead, RINA, London, 1991 yy of Sailing Multihulls, MCA Research Project 427, Wolfson Unit Report 1738, oust ereaidde ame On ‘Bupmon xed + PUM LAIN 820°0 Bunny 10 Bupwor xed +4986 + puMn WM 820" perw eae Zo enpises wn ‘Buyoune} 120954 ‘Buppwon xed ‘gum Z9 Ferpse "UN ob ‘fue Bupmon 0b sabvassed xen | © \y2i0 06129 30} 07 "yen xed Joy SL 1876 U!.OL ei6ue Bureeu | UP UUW pauddy swowow SunoH HIM ale 75 peseanul UA 01 20'S tues unum | 2 02 az 0b ob oeue LvoReupU! *e WY exw Gupoopumop 2®.070,06 | $ eve Z0 UNL 200 s190 s190 siv0 ‘Suipoojumop 20 1z 915100 Z0 eu of6ue 0} dn (06 10 Supoopunon 10 20 ‘ew ofbue)ne x 550°0 Zo xeUl Jo 6j6ue 0} da (Zo xeup ape de)L000 + s80°0 peru "wave ZO UNLAIY ob Z st xew Jo 2/6Ue “UIA S00 +o wZ9 anu wwe wn sebes oun | umyoulR nano pyouow cd woueND mnyouoW ‘eveipewienu ‘ebeweg fuy | eBeweq Supreu x00 LINN NOSSIOM ‘aBewe wonoE 30 ePIS e12ay140 Jo Aremumng —2p0D ISH 0007 I xIPusddy WOLFSON UNIT Appendix 2 Summary of Test Configurations, Stability Parameters and Min, Wave Heights to Capsize ID Model Configuration & Damage Extent Mt Mi Mi Mt Mi Mi Mt Mt Mi 10 Mi 4 oMt 12 Mt 13 Mt 14 Mt 16 Mt 16 Mt 17 Mt 18 Mt 19 Mt 20 Mt 21 Mt 22 M2 23 M2 24 M2 25 M2 26 M2 27 M2 28 Ct 29 C1 30 ct 31 Ct Sta Ct 32 C2 33 C2 34 C2 35 C2 36 a7 38 39 2 ca. 2 2 40 02 41 C2 42 63 43° 03 44 63 45 63 48 C3 art 43 Tt 49 Tt 50 Ti 51 Tt 52 Tt 83 Tt 53a Tt {This aroumes thet int for passongor erat. For cargo craft the limits 20 Intact Intact Intact Intact Intact 23% to 76% 23% to 76% 29% to 76% 5% to 36%, superstructure 5% to 35%, superstructure 5% to 36%, deck intact 23% to 76% 23% to 76%, superstructure, deck intact 23% to 76% 23% to 76%, deck intact 23% to 76% 23% to 76%, superstructure 23% to 76%, deck intact 23% to 76%, superstructure, deck intact 29% to 76%, deck intact 28% to 76% superstructure intact, superstructure 0% to.21%, superstructure 0% to.21%, superstructure 0% t0.21%, superstructure 21% to 45%, superstructure 45% to 86%, superstructure Intact 0% t0.31% 31% to 82% 400% reking Intact Intact Intact 0% to. 31% 0% to 31% 0% to 31% 8% to 31% 8% to 31% 8% t0 31% BY to 31% 24% to 68% 0% to 39% 14% to 72% 14% to 72% 14% to 72%, superstructure 18% to 72% Intact Intact 61% to 100% 61% to 100% 61% to 100% 61% to 100% 100% sponson raking damage Intact Critical Disp. VCG Criterion List tonne —'m — No, deg 123 40044 5.0 123 4684 489 © 100 200 4.470 Margin 00 200 4.650 FllsS 49 200 4505 589 100 123 4.950 Margin 000 123 5012 Margin 100 123 5035 2810 130 200 3790 2 00 200 3070 Faile2&4 0.0 200 3799 2810 80 200 4.050 Margin 0.0 200 4.130 Margin 0.0 200 4130 Margin 0.0 200 43472 00 200 43472 00 200 43472 00 200 4130 Margin 55 200 4.130 Margin = 55 200 4347 2810 © 62 200 4347 2810 82 1260 9277 ‘5&9 100 126010110 Margin 0.0 126010110 Margin 38 126010331 2810 46 1260 9.568 2810 = 75 126010780 2810 Bt 93 6932 | 9 100 93 4075 8 154 93 3887 6 100 93 5.760 Fois7 200 170 3.690 Fails9811 20.0 93 8.990 Margi 00 931.000 Margin 0.0 93 4.698 Margin 10.0 93 5.900 11 107 93 4700 Fails11 130 93 4.700 Margin 120 93 4.968 Margin 120 93 52% 6 120 93 6.040 Faisé 120 93 6.098 Margin 93. 110010.530 Margin 13.8 110010530 Fals8# 176 110010530 Fais8# 200 110010.600 Fals8# 21.0 110010520 Fais8# 21.8 175 5273 Margin 85. 300 4700 4 00 475 4499 Margin = 0.0 175 4439 Margin = 36 175 4.439 Margin = 7.0 300 4.105 Falls11 10.0 300 4.700 Margin 89 307 4.630 Margin = 89 50 1D A4is csc 42 Min. Wave to GM Gzmax GZAreaRange Capsize mom mrad deg m 237 018 0.0588 335 © 210, 073 004 0.007 153 0.80 143 033 0.162 415 115 018 0061 303 330 105 0.09 0028 242 1.65 029 0.13 0031 263 © 1.10 027 G07 O00 142 © O77 057 0.04 0.008 98 050 O71 0.10 0.029 262 046 0.08 0.014 21.3 047 0.04 0.005 164 1.40 065 017 0.047 26.1 059 0.15 0099 247 © 1.60 089 015 0.038 247 O47 G10 0022 210 073 O47 0.10 0022 210 0.65 047 0.10 0022 210 0.88 089 0.08 O04 165 0.50 089 0.08 O04 165 1.15 062 0.04 0.005 117 0.40 062 0.04 0005 117 0.60 129 007 0014 168 274 102 014 0.035 217 084 0.08 OO 166 148 065 0.04 0.005 108 1.00 044 0.04 0006 126 230 060 0.04 0004 97 170 385 0.33 0043 122 0.98 327 023 0044 173 1.15, 555 181 0670 452 628 0.23 0.028 112 1.20 253 043 0015 112 1.20 1198 233 0637 292 4.00 1014 190 0428 237 © 2.80 716 049 0.113 210 1.90 516 023 0029 115 0.90 488 0.18 0.025 121 0.95, 11.48 063 0.122 188 1.60 1148 081 0115 177 1.30 1100 057 0100 164 1.20 1014 047 O71 137 90 1348 076 0.187 226 © 1.90 4010 0.43 0.066 136 1108 079 0.196 220 2.30 798 0.46 0.08 164 190 668 0.33 0083 141 3.00 642 0.26 0095 122 130 163 085 0.165 310 * 069 039 0209 638 160 055 0319 706 149 048 0.231 659 065 043 0166 385 * 042 040 021 200 * 160 O71 034 0146 635 069 0.33 0.143 635 * 2nd stably peak WOLFSON UNIT Figure 1 Distribution of Vessels in the World Fleet with Regard to Length Monohulls Catamarans 350 300 250 200 150 400 o888eees [a Word] UK No. in Operation ‘No. in Operation <20 2040 4060 >60 <20 20-40 4060 >60 Length - metres Longth - motros 4B WOLFSON UNIT Figure 2 Fleet Principal Characteristics 4 —e— Monohuils — + = Catamarans | 30) © Monehull Models. ee 3 ode po i (1 Catamaran Models ett é E Trimaran Model a8 oo = 204 | é s 8 | 10} | ° 0 20 40 60 e100 120s 1HO«18D Length Overall- metres — = - Catamarans 6] [Catamaran Models| Z A Trimaran Model = ° 0 20 40 60 e100 120 140160 Length Overall metres 100- —e— Monohulls — = - Catamarans 3 © Monchuil Models @ $1000) ©) Catamaran Models| ¥ ‘Trimaran Model E 8 & 500) fs o 0 20 40 60 go 00,120, 4DS«N8D Length Overall - metres 44 WOLFSON UNIT Figure 3 Variation of Passenger Capacity with Length ou ——— Morcha — = - Catamarans) adie ‘soo E & 10004 3 5 3 seo 20 4 60 80, 100 120 140 160 Length Overall - metres Figure 4 Maximum Permitted Operational Sea States ——°— Monohulls # 3 i i £4 g = eS Eo a 5 : Length Overall - metres 45 WOLFSON UNIT Figure § Monohull 1 Prineipal Dimensions Length Overall 38.80 metres Moulded Beam 8.96 metres Depth 3.66 metres Displacement 200 123 tonnes LCG (fwd of transom) 14.85 14.83 metres Draught 157 1.21 metres Model scale 1:25 Body Plan Profile Vi I 1 I 1 0% 5% 23% 35% 18% 100% Bulkhead positions 46 WOLFSON UNIT Figure 6 Monobull 2 Principal Dimensions Length Overall 96.00 metres Moulded Beam 17.00 metres Depth 10.66 metres Displacement 1260.0 tonnes LCG (fwd of transom) 36.00 metres Draught 2.51 metres Model scale 1:40 Body Plan | | | I 0% 21% 45% 86% 100% Bulkhead positions 47 Figure 7 Catamaran 1 WOLFSON UNIT Principal Dimensions Length Overall 32.70 Moulded Beam 8.32 Hull Beam 267 Hull Separation 5.67 Depth 3.39 Displacement 3 LCG (fwd of transom) 1231 Draught 14 Model seal 1:20 Body Plan metres metres metres metres ‘metres tonnes metres metres Profile Compartments ~ Port hull — 0% 24% 49% 58% ! 0% 8% Bulkhead positions Compartments ~ Starboard hull 31% 48 N 100% D fl 82% 100% WOLFSON UNIT Figure 8 Catamaran 2 Principal Dimensions Length Overall 32.70 metres Moulded Beam 11.07 metres, Hull Beam 2.67 metres, Hull Separation 8.40 metres Depth 3.39 metres Displacement 93 tonnes LCG (fwd of transom) 12.31 metres Draught 1.14 metres Model scale 1:20 Body Plan Profile Compartments ~ Port hull i= 2% 6 st 100% Compartments ~ Starboard hull | I I 0% 8% 31% Bulkhead positions 49 Figure 9 Catamaran 3 WOLFSON UNIT Principal Dimensions Length Overall 70.8 metres ‘Moulded Beam 26.1 metres Hull Beam 436 metres Hull Separation 21.67 metres Depth 6.61 metres Displacement 1100.0 tonnes LCG (fwd of transom) 26.92 metres Draught 3.11 metres Model scale 140 Body Plan a | 0% Bulkhead positions Profile Compartments ~ Port hull Ll 39% I | 1% 100% Compartments — Starboard hull Clee i Il 29% om 100% 50 WOLFSON UNIT Figure 10 Trimaran 1 Body Plan Principal Dimensions Length Overall 53.00 metres Moulded Beam 12.04 metres Contre Hull Beam 4.00 metres ‘Sponson Separation 10.00 metres Depth 5.60 metres Displacement 175.0 300 tonnes LCG (fwd of transom) 202 20.2 metres Draught 1.48 2.13, metres, Ratio Sponson Volume to 65 38 percent Displaced Volume Model seale 1:25 Body Plan Plan & Profile I 0% 30% 61% Bulkhead positions 31 WOLFSON UNIT Figure 11 Photographs of the Monohull Models Monohull 1 Monohull 2 WOLFSON UNIT Figure 12 Photographs of the Multihull Models Catamaran 2.(Catamaran 1 had the same hulls at closer spacing) 3 WOLFSON UNIT Figure 13 Monohull 1 GZ Curves for 1D 1-8 020 ——1 an —— «ny £0.10 nN 8 00s be ee Fe 0.00 ~ oC Heel Anie- deorees 03s cel Ds 030 a seems ---- ws 02s 8 020 N =e. gyo1s = = Ff * oto — / e ~SY\ 0.05 > * : jaeeeal | \ 0.00 rr Te ee Hest Ande-donroen 020 10s 0.15 Peal i ---- 0s £0.10 1] 8 a 0s — ae tk 0.00 for os 40 is 208 SDSS Heel Angle - degrees WOLFSON UNIT Figure 14 Monohull 1 GZ Curves for ID 9-21 020 De 015 —— > Pill i =--- on Zot 8 i 0.05 => aa ~ itl x 0.00 rT a a ee Heo! Angle - degrees 020 ——on 015 — — - IDs 13& 14 8 “XN ---- IDs 18819) 3 Y Ni E oto 8 a TN 005 ON 000 CETUS FEM Mo te eae tao ett IEE so 5 Heel Angle - degrees 0.20 ibs 15,168 17 ie — —-ps20821 £010 4 8 005 = o NX 0.00 eens Orta Tag ew sy Ur EEE aa Heel Angle - degrees t 55 WOLFSON UNIT Figure 15 Monohull 2 GZ. Curves for 1D 22-27 0.20 1D 22| a8 — ID 26 8 ---- 107 3 E010 8 | — 0.05 sat is ‘ N 0.00 ° 6 200250 8H AB Heel Angle - degrees. 0.20 ——_ 102 ous - ID 24 8 f oN -- 77 1025 0.410 y = 8 0.05 eo ea ~| * 0.00 0 5 10 18 20258088 40 45 Heel Angle - degrees WOLFSON UNIT Figure 16 Catamaran 1 GZ Curves for 1D 28-31 10 10.28 08 +1031 —e—iD3ta 18 —— 129 ID 30) 16 14 12 04 02 0.0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 Heel Angle - degrees 37 WOLFSON UNIT Figure 17 Catamaran 2.GZ Curves for 1D 32-41 24 a2! 22 - 133 = ai 20 v{\ pa 16 if 3g" 1H £12 H! Yio f 08 / I 6 |} n i oa i 0.2 H 00 0 6 10 45 00 os —— pm Bos = 1935 é ---~ 1036 oz 7 00 ~ 0 5 1 15 20 2 30 35 40 45 50 0 6 Heel Angle - degrees 08 —— ws 208 - 1038 i ---- 1939) E04 ~ —-— a0 \ Boo i AN 00 0 5 10 15 20 2 30 35 40 45 50 55 60 65 Heel Angle - degrees. 58 WOLFSON UNIT Figure 18 Catamaran 3 GZ Curves for ID 42 ~ 46 142 08 - 143 0 6 10 15 20 2 30 36 4 Heel Angle - degrees gs 8 19 Daa 08 ~ 1D45 3 =--- 146 boo e yt 02 > 1" NY 00 0 5 10 15 20 2 30 35 40 45 50 55 60 65 Heel Angle - degrees. 39 WOLFSON UNIT Figure 19 Trimaran 1 GZ Curves for ID 47 ~53 g TA — = op Xe PNY aa 7 Heel Angle - degrees os Die tee es g O4)— ID 53a) | g eet S =9- -¢ = | — Sh bb eb eb b ae + 02 a Heel Angle - degrees 60 WOLFSON UNIT Figure 20 Monohull 2 GZ, eurves on a Wave Effect of waves on the GZ curve of Monohull 2 at critical wave height and period calm water — e— - 8s, 3.2m, crest at stern ‘4 — 85, 3.2m, crest at bow — ~ — 85, 3.2m, crest amidships: 06 Heol Angle - degrees 61 WOLFSON UNIT Figure 21 Monohull Raking Study — Narrow Hull, Light Draught 18 55% Raking Fwd Intact — = = 50% Flooding Equiltxium = — ~~ 75% Flooding gt — — - 25% Flooding {391020204060 35% Raking Aft g10 Bos 00 ‘s9 10209040 '35% Reking Aft, no double bottom in ER. g10 3 é Bos 7020300 100% Raking eee 702030 40S 100% Raking, no double bottom in E.R. WOLFSON UNIT Figure 22 Monohull Raking Study — Narrow Hull, Deep Draught 15: gro 5 Sos 00 ‘55% Reking Fwd —— intact — ~ = 50% Flooding 75% Flooding 020 90 40 60 ‘35% Raking Aft, no double bottom in E.R, 0. to 2030 400 100% Raking, no double bottom in E.R. ‘Anglo - dogroos 6B WOLFSON UNIT Figure 23 Monohull Raking Study ~Mid Width Hull, Light Draught 1s 5 Raka Fd peg rg Equibium ~~. 75% Fong re = + 24 Foodng 7020 90400 ts Trang deen WER gto e Bos A ace tit tg 00 100% Faking eae 7020304078 100% Raking, no double bottom in E.R. WOLFSON UNIT Figure 24 Monobull Raking Study — Mid Width Hull, Deep Draught 1 55% Raking Fwd Intact — = = 50% Flooding Equltrim — — ~~» 76% Flooding pro a Peaeg i Bos oo 00 9595089 90a 6% Raking At gio i é Bos aH ee 90 1520 30 ao 000. 7060 BE Raking Aro double Dotom ER. ie i Bos = a - 00 9p go pad 80 670 700% Rang ye i Bos 00 9019 99 30 a 0). 097060 i Bos LS, SS, #5, oO 10, 20 30 40 50, 60 70 80 90 65 WOLFSON UNIT Figure 25 Monohull Raking Study — Wide Hull, Light Draught 15 gro E Bos a0, 15 7 i dos 15 a0 ‘Wide Hull, Light Draught ‘55% Raking Fwd —— intact 10 — 50% Flooding 78% Flooding —— Equitbrium 25% Flooding NON SS 20300 36% Raking At 10. 23904080 ‘3636 Raking Aft, no double bottom in ER. OF saan, tae 10 2030 40 OC 100% Raking en 2 3 4 «80 6 70 @ 0 Angle - degrees WOLFSON UNIT Figure 26 Monohull Raking Study ~ Wide Hull, Deep Draught Wide Hull, Deep Draught 18 155% Raking Fwd intact — = = 50% Flooding pik - = Ton rosie gro Ta rong e Bos oe aera mA gro j : Bos 0 Sega ee ep aaa en oneae Fe Bos 0 SS Se 100% Raking - | Bos 00 as! 4020 40. 6070, 100% Raking, no double bottom in E.R, g10 Bos 10 20 0 4 6 6 70 @0 0 Angle - degrees 00, ° or WOLFSON UNIT Figure 27 Monohull Raking Study — Comparison of 34% & 45% Flooding 15 ‘35% Raking Aft, no double bottom in E.R, Intact — ~ = 80% Flooding g 10 Equilibrium = — - - - 75% Flooding z — — - 25% Flooding Bos {s0_10 203040 45% Raking Aft, no double bottom in E.R, i 19 Bos pB§aaae SS 0 1 2 4% 4 60 6 70 0 90 Mia With Hull, Deep Draught 15 35% Raking Aft, no double bottom in E.R. B10 Bos s 1 oO Py ananeseey SEEDY PERT SORENTO TERETE TT CNTY EEE Period - seconds 1 WOLFSON UNIT Figure 32 Monohull 1 Capsize Boundaries for 1D 11-19 Wave Height- metres Wave Height - metres Wave Height- metres 6 — on 5 4 3 2 1 36 75 5 105 6 —— 5 - 1D 18) ---- p19 4 3 2 ; -4- 0 38 5 65 78 5 105 Period - seconds 6 ID 15 5 - 1016 ---- 7 4 2 Z 2 Lo 1 ° 25 35 45 55 65 75 85 95 10.8 Period - seconds n WOLFSON UNIT Figure 33 Monohull 2 Capsize Boundaries for ID 20 ~ 21 { 6 1020 5 = 1024 8 3 a4 | Be 3 Z 82 3 2 1 ° 25 35 4 5 65 75 5 105 Period - seconds Figure 34 Monohull 2 Capsize Boundaries for 1D 22-27 7 122 i 6 1026} ---- 27 7) Zz a Z Zi 0 25 35 75 108 5 ° ID24 3 4 ID 25) E 2 j ey 2 g | s1 0 25 36 45 55 65 75 85 95 105 Period - seconds B WOLFSON UNIT Figure 35 Catamaran { Capsize Boundaries for 1D 28-31 6 1D 25] = 1031 = z {1 asin] FRnoninadal (be>=racaamy pee ee | 0 25 45 5 75 9. 105 6 —— 2 5 g 4 £ g2 5 a Peers ° 25 35 45 55 65 75 85 95 10.5 Period - seconds 4 WOLFSON UNIT Figure 36 Catamaran 2 Capsize Boundaries for ID 32-41 6 132 5 - 1wss| £ [---- 0m g4 7 ea = és | | £ 4 ’ 1 ° 5 35 45 35 65 75 as 95 105 Period - seconds 6 D3 5 ° ID 35] # |=—-0% —- i Bs ? 2 2 g? To g = 1 o ° 5 35 45 35 75 85 95 105 Period - seconds 6 1037 s} 0 De £ |---- 0s 2 4]— —- 10] ° Zz ° 2 §? o = s a ; a - — 0 25 35 45 55 65 75 85 95 10.5, Period - seconds 5 Figure 37 Catamaran 3 Capsize Boundaries for ID 43 ~ 46 WOLFSON UNIT 6 —— 1043 5 Ba Ss 3 — g2 = 1 ° 25 35 5 5 6 —— D4 : 5 = 1048 goo ee a e 2 Bo 2 2 g? 7 g ee eae 1 ° 45 5 65 75 5 105 Period - seconds Figure 38 Trimaran 1 Capsize Boundaries for 1D 51 6 —— 0s 5 8 Ba Bs z g? — 3 1 0 25 35 45 55 65 75 85 95 105 Period - seconds 16 WOLFSON UNIT Figure 39 Wave Statisties ‘Contour chat shows 4s | sistibution of occurrence of ‘waves of varying height and [ fiequency in the North Sea, 45, | Caribbean and Western 4 | Mediterranean, The contours sad, | Beat 1% intervals AN oe ant, i = S ' WOLFSON UNIT Figure 40 Photographie Sequence of a Capsize of Configuration ID 4 Capsize mode A B Figure 41 Photographic Sequence of a Capsize of Configuration 1D 43 Capsize mode D Figure 42 Tested Wave Conditions For a model that Did Not Capsize WOLFSON UNIT v 42 Wave Height - metres 35 45 55 65 Period - seconds 80. 75 85 95 10.5 WOLFSON UNIT Figure 43 Variation of Stability Parameters with GM ‘© Monohuts Not Gapsized © Intact Monohuts * Damaged Monohuils © Catamarans Not Capsized| % Intact Catamarans Damaged Catamarans 20 ‘5 Trimarens Not Capsized 4 Damaged Trimaran . i : g 5 #10] 3 . ao os o =. 2 hoo . on ofo# 004 ~ 0 2 a é 3 7 12 14 GM- metres a a . . = . af a ou eee . 004 a 0 2 a 6 3 10 2 14 GM metres cs . bool ° : 3 ™ i yl age . . & |. a . fol 3 " . 2 a é 3 10 12 14 GM-meties 81 WOLFSON UNIT Figure 44 Relationship between GZmax, GZarea and Range of Stability 7 © MenohalsNo aps 7 © Catamarans Not Copazd . os] Tanaans Nt Capa < Ruaconhile ool Mat Clare a mera 5 Damaged Monohulls . Bos] © Damaged Catamarans B a. Damaged Trimaran : _ gos a2. 4 ” 0.4: a aol : o% 3 is 7 2 aenaleudar or vane q cama q ook Fammane 0. i 5 E iN gost 8 a 024 8 a 0.1- aed le a a o 7 we 7 i Renge x GZmax - metre.radians. 82 WOLFSON UNIT Figure 45 Variation of Minimum Wave Height to Capsize with Stability Parameters Intact Monohulls 5 Intact Catamarans Wave Height to Capsize - metres os to 15 20 25 GZmax - metres . © Damaged Monohulls = Damaged Catamarans Damaged Trimaran Weve Height to Capsize - metres ds 10 15 20 25 Gzmax - metres > Monohull ‘safe! bounday Catamaran ‘safe’ boundary © Intact Monohulls 2 Intect Catemarans| Wave Height to Capsize - metres oa 05 08 0 Gzarea- motes BE é ed i oe . 2 ° . 2 fs 7. © Damaged Monohuls gr -— . Damaged Catamarans| 3 Damaged Trimaran Eo 00 om 02 a3 oa os 08 07 GZarea- metres 3 WOLFSON UNIT Figure 46 Results Re-Scaled to Represent Vessels with GZmax = 0.1 metre 6 © Intact Monohulls © Damaged Monohulls| 2mex= 0.1 metres . os 2 Intact Multinulls a 3 = Damaged Multhuils Hee & 6 i Ss 2 5 £2. & 2 0 80 100 160 200 280 Length - metres 6 © Upright Monchuts a ®— Heeted Monohute| G2max= 0.1 metres 5 ee Point labels indicate the initial heel angle ©] © Upright Muithuls 8 | Heeted mutinuts 3 a + 4 10, * j % Ss 2 z 48, Zo, *% 10, 10, 38, g 6.2. 3 18, = 62" 3 30 G0 150 20 250 Length - metres 84 WOLFSON UNIT Figure 47 Stability Curve for a Monohull that Just Complies with the Code. ——— 62 Cune Satisfying the Requirements of the Code for an Upright Monohull — — - Heeling Moment Residual GZ when Heeled to 10 degrees 04 — a 03 = E02 = 8 oa 00 acer Eee ° 10 20 30 40 60 60 Heel Angle - degrees 85 WOLFSON UNIT Figure 48 Results Re-Scaled to Represent Stability Corresponding to Minimum Code Criteria 7 oO Intact monohulls ge Damaged ie 8 . ged monohull 2 Intact muitihutts 8 5 . Damaged multihulls B | critical heel angle cases excluded be 2 Bo g oO & 2 5 - 3 a 2 . gy . . . on . wa t ° ° 20 4% & 3 00 Length Overall- metres 7 ° Intact monohulls : © Damaged monohulls ° zg Intact muttinulis Current vessels’ maximum permitted E significant wave height BS][_§ Damaged mututs Wher canoe : waves often exceed & | Siteathest angle cases nduae Serna 3 . a factor of 2 or more 2 . z Oo 3 * A o 6 a . 82. © = 9 ° é er . . «a 4 q o- 0 20 2 0 30 100 Length Overall - metres 86 WOLFSON UNIT Figure 49 Variation of Non-Dimensional Wave Heights with GZmax © Intact Monohulls, © Damaged Monohulls Intact Mutinulls = Damaged Multihulls ‘4 100% Raking Damage Multinulls| a ° . Wave Height / Beam § 0.10 0.45 0.20 0.25 GZmax / Beam S § 0.10 S ° o $ 0.05 |}-—2 g " 5 0.00 0.02 0.04 0.08 0.08 Gzmax I Length 87 WOLFSON UNIT Figure 50 Variation of Non-Dimensional Wave Heights with Range of Stability ou 5 3 > moeo Intact Monohulls Damaged Monohulls Intact Muttihulls Damaged Mutthuils 100% Raking Damage Mulithulls Wave Height /Lenath B28 3 3 0.40" 3 10 15 20 25 30 3 Range of Positive GZ - degrees 0.35- 5 boss Boz| 5 0.104 0.004 0 5 10 15 20 25 30. 3 Range of Positive GZ - degrees 0.30: 28 8 3 Wave Height / (LB"3)"0.25 0.054 0.0 6 10 8 20 25 Ey 35 Range of Positive GZ - degrees 88 WOLFSON UNIT Figure $1 Variation of Non-Dimensional Wave Heights with Combined Stability Parameters 7 0.12{ © Intact Monohuls © Damaged Monohuis cro] 2 tae Catamarans s "Damaged Catamarans | g + 100% Raking Catamaran} | : } 0.081 4 Damaged Trimaran 2 0.064 a 7 3 ° : . s = 0.00 02 oa 08 o8 a 1 Range(RMmax)*0.5 iB 212) capsize Mode ae Formula 4 Sore o 8 ‘ormula 010 ee 4 aoc 4 4 3 vo : ©. Formula 2 0.08: x E a é e a 2 2 B08 a a 2 3 y 4 z Fe 2 § : 4 So. 7 eos : ov PU = 0 L ee oor} ,* yp can Oa ' 4 0.00. 0.0 o2 rv 06 o8 to 12 Range(RMmax)*0.5 | 8 89 WOLFSON UNIT Figure 52 Other Model and Full Scale Capsize Data Y Inverted Rescue Boat Model © Fishing Vessel Casualties 0.20 - e 7 a . Boss 4 3 . Z 0.104 ’ . i i eee go, eer Formula 1 < J er s | feo 0.00- 00 02 os 08 oe 10 12 Range(Ftaman)0.5 1B Figure 3 The Application of Formula 2 to Existing HSC © Current Permitted Operational Seastate - Monchulls, Current Permitted Operational Seastate - Catamarans Maximum Seastate by Formula 2a - Intact, Maximum Seastate by Formula 2a - Damaged FV Casualties - Maximum Seastate by Formula 2a ene acy . Current vessels’ operating limit Implied limit for intact vessels by Formula 2a a *., Implied limit for damaged vessels by Formula 2a ‘Maximum Permitted Sig. Wave Height- metres 20 40. 60 80 100 120 140 Length Overall - metres 90

Você também pode gostar