You are on page 1of 49

CLASSIFICATION NOTES

No. 30.9

SLOSHING ANALYSIS OF LNG MEMBRANE TANKS


JUNE 2006

DET NORSKE VERITAS


Veritasveien 1, NO-1322 Hvik, Norway Tel.: +47 67 57 99 00 Fax: +47 67 57 99 11

FOREWORD
DET NORSKE VERITAS (DNV) is an autonomous and independent foundation with the objectives of safeguarding life, property and the environment, at sea and onshore. DNV undertakes classification, certification, and other verification and consultancy services relating to quality of ships, offshore units and installations, and onshore industries worldwide, and carries out research in relation to these functions. Classification Notes Classification Notes are publications that give practical information on classification of ships and other objects. Examples of design solutions, calculation methods, specifications of test procedures, as well as acceptable repair methods for some components are given as interpretations of the more general rule requirements. A list of Classification Notes is found in the latest edition of Pt.0 Ch.1 of the Rules for Classification of Ships and the Rules for Classification of High Speed, Light Craft and Naval Surface Craft. The list of Classification Notes is also included in the current Classification Services Publications issued by the Society, which is available on request. All publications may be ordered from the Societys Web site http://exchange.dnv.com.

INTRODUCTION This Classification Note was first issued in June 2006.

Comments may be sent by e-mail to rules@dnv.com For subscription orders or information about subscription terms, please use distribution@dnv.com Comprehensive information about DNV and the Society's services is found at the Web site http://www.dnv.com Det Norske Veritas Computer Typesetting (FM+SGML) by Det Norske Veritas Printed in Norway.

If any person suffers loss or damage which is proved to have been caused by any negligent act or omission of Det Norske Veritas, then Det Norske Veritas shall pay compensation to such person for his proved direct loss or damage. However, the compensation shall not exceed an amount equal to ten times the fee charged for the service in question, provided that the maximum compensation shall never exceed USD 2 million. In this provision "Det Norske Veritas" shall mean the Foundation Det Norske Veritas as well as all its subsidiaries, directors, officers, employees, agents and any other acting on behalf of Det Norske Veritas.

Classification Notes - No. 30.9 June 2006

CONTENTS
1. 1.1 1.2 1.3 1.4 1.5 1.6 1.7 2. 2.1 2.2 3. 3.1 3.2 3.3 3.4 4. 4.1 4.2 4.3 4.4 4.5 4.6 4.7 5. 5.1 5.2 GENERAL................................................................. 4 Introduction..................................................................4 Membrane type LNG tanks .........................................5 Purpose of this Classification Note..............................5 Documentation.............................................................5 DNV Classification of LNG carriers ...........................5 Design basis .................................................................7 Design principles .........................................................8 APPLICATION OF THE PRINCIPLES ............. 11 Limit state applications..............................................11 Strength assessment methodology.............................14 SLOSHING IMPACT DESIGN LOADS ............. 15 Environmental modelling ..........................................15 Ship motion calculation for sloshing tests .................16 Sloshing model experiments......................................16 Sloshing design loads ................................................19 PUMP TOWER DESIGN LOADS ....................... 22 Identification of relevant loads ..................................22 Sloshing loads............................................................22 Inertia and gravity loads ............................................25 Thermal loads ............................................................25 Hull girder loads ........................................................25 Internal tank pressure and external sea pressure........25 Combination of loads.................................................26 STRUCTURAL RESPONSE ANALYSIS OF CONTAINMENT SYSTEMS................................ 26 General.......................................................................26 Mark III system..........................................................26 5.3 5.4 5.5 5.6 6. 6.1 6.2 6.3 6.4 6.5 6.6 7. 7.1 7.2 7.3 7.4 8. 8.1 8.2 8.3 8.4 8.5 8.6 8.7 9. NO96 system ............................................................. 31 Material stiffness parameters .................................... 33 Mastic ........................................................................ 34 Simplified assessment of dynamic response ............. 34 ULTIMATE STRENGTH OF CONTAINMENT SYSTEMS ............................................................... 35 General ...................................................................... 35 Capacity assessment .................................................. 35 Mark III system ......................................................... 35 NO96 system ............................................................. 37 Plywood strength data ............................................... 42 RPUF strength data ................................................... 42 STIFFNESS AND STRENGTH OF HULL STRUCTURE ......................................................... 42 General ...................................................................... 42 Comparative basis (reference case)........................... 43 Stiffness of hull plating ............................................. 43 Strength of hull structure........................................... 45 RESPONSE AND STRENGTH OF PUMP TOWER AND SUPPORTS .................................. 46 General ...................................................................... 46 Response analysis of main structure ......................... 46 Response analysis of base support ............................ 47 Response analysis of liquid dome area ..................... 48 ULS assessment......................................................... 48 FLS assessment ......................................................... 48 Vibration check ......................................................... 49 REFERENCES ....................................................... 49

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

1. General
1.1 Introduction
The sea-borne LNG transportation is increasing dramatically. Not only is the industry increasing in size, it develops new trade routes, new ship designs and new operations. Traditionally LNG carriers operate in a fully loaded condition or in a ballast condition to their next port-of-call. Presently key technical developments are the large increase in vessel size (from ~140 000 m3 to ~250 000 m3 and above) and the discharge of cargo in offshore waters. These developments imply a change in cargo dimensions and the operation with partially filled tanks in a wave environment. For the structural integrity of the cargo insulation, the hull support structure and the pump tower and supports the loads induced by the motions of the liquid cargo inside the tanks need to be evaluated. These motions of the liquid cargo inside the tanks is called sloshing. Sloshing can induce various types of loads. Slowly varying motions cause a smooth change of pressure distributions inside the tanks and a static pressure model can be applied. Larger motions and/or more rapidly varying motions, causing higher accelerations, induce dynamic effects. The pressure fields inside the tanks can still be described by smoothly varying pressure distribution functions and the structural response can be calculated in a quasi-static manor. In case of more severe motions the fluid behaviour becomes violent, causing breaking waves and high velocities of the fluid surface. In this case the fluid can cause impact loads on the containment system. These loads can be characterised by a high load with a short duration acting on a limited area. Violent sloshing can be characterised by various fluid flow phenomena illustrated in Figure 1-1 - Figure 1-5. In the high filling range (>90% H) the impacts typically occur on the tank roof at the connection with the transverse bulkheads. Typically a flat fluid surface hits at high velocity the roof causing the impact.
Impact location

CL Tank roof

Chamfer

Impact location

Keel

Hopper

Figure 1-2 Typically high-filling (~60-70% H) impact due to a run-up against the longitudinal and or transverse bulkhead

CL Tank roof Impact location

Chamfer

Keel

Figure 1-3 Typically high-filling (~70-80% H) impact due to a run-up along the longitudinal bulkhead and chamfer

For fillings in the range of ~20% to ~40% the largest impacts occur at the longitudinal and transverse bulkheads due to breaking waves.
Tank roof

Tank roof

Chamfer knuckle
Chamfer knuckle Transverse BH Upper hopper knuckle

Transverse BH Impact location

Hopper knuckle

Keel

Keel

Figure 1-1 Typical high-filling (>90% H) impact in near head sea conditions

Figure 1-4 Typical low-filling impact in near head sea conditions

For fillings in the range of ~60% to ~80% the largest impacts occur in the corners and knuckles of the chamfer. These impacts can be caused by run-ups against the longitudinal or transverse bulkheads or by a flat fluid surface impact.

A characteristic phenomenon, which can occur at lower fillings is the so-called hydraulic jump or bore. This wave phenomena is characterised by a jump in the free surface level, which travels at high speed, which can cause a large impact.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

CL Tank roof

Chamfer

Impact location

specify the DNV requirements for approval of membrane type LNG carriers and floating structures exposed to liquid sloshing in their cargo tanks provide the LNG industry with the necessary methodology to assess sloshing impact loads, and to evaluate its impact on the design of membrane type containment systems for LNG, the supporting hull structure, and the pump tower structure provide the LNG industry with guidance on how to use this methodology to comply with the DNV class requirements. The Classification Note is mainly intended to be used for the approval of LNG carrier designs and operations that deviate from the designs and operations for which operational experience is available. This operational reference is defined in more detail in 1.6.2. The methodology presented in this Classification Note can also be applied in the sloshing load and strength assessment of other marine structures where sloshing is an important design parameter, such as floating LNG terminals and production units.

Keel

Figure 1-5 Schematic illustration of an hydraulic jump or hydraulic bore

Sloshing model experiments are required in order to assess the violent sloshing causing impact loads. Section 3 gives a detailed description on how to conduct sloshing experiments and how to determine sloshing design load predictions. The sloshing loads vary in size, duration and load area. In addition, the containment system and hull structure have different failure modes. Consequently, a careful analysis of the structural response and strength needs to be conducted for the various loads to assess the structural integrity. A detailed discussion on the response predictions, the failure modes and the strength are given in Section 5 to Section 8. In traditional direct wave load analysis of ships the load and strength assessments are conducted in an absolute approach. Due to uncertainties in the sloshing impact load assessment a comparative approach is used for the containment system and hull strength, which is described in detail in Section 2. For the pump tower, an absolute approach may be used. The basis for a comparative approach lies in the application of the equivalent safety principle as outlined in the DNV Rules for the Classification of Ships. In the comparative approach the sloshing load and strength of a new LNG carrier design or a new operation of an LNG carrier is compared with the sloshing load and strength of the existing fleet of membrane type LNG carriers that have traded in a safe and damage free operation. The former is referred to as the target vessel, whereas the latter is referred to as the reference case. The remaining of this section outlines in more detail the purpose and background of this Classification Note as well as the design basis and the design principles, which are used for the sloshing load, response and strength assessment of membrane LNG tanks.

1.4 Documentation
The following documentation is to be provided and submitted to the Classification Society as part of the design review and approval: the design basis used for the design of the vessel, i.e. the target case the sloshing experimental programme, including the ship motion predictions, the test set-up, the test scope, the postprocessing and the results the analysis of the sloshing experimental programme load and/or strength comparative analysis for containment system and hull strength, comparing the target case with the reference case load and strength calculations for the pump tower and supports drawings of containment system, pump tower, and supports.

1.5 DNV Classification of LNG carriers


1.5.1 Applicable Rules and Classification Notes The classification of LNG carriers is governed by the following rules and regulations in addition to the requirements to Main Class: DNV Rules for Classification of Ships Pt.5 Ch.5 Liquefied Gas Carriers /2/. International Maritime Organisation: International Code for the Construction and Equipment of Ships Carrying Liquefied Gases in Bulk (IGC Code) /1/. In addition to these documents, the Classification Society enforce limits on the allowable tank filling levels and limits on the allowable cargo tanks dimensions. These limits are subject to continuous considerations, and the applicable limits can be obtained upon request to the Society. 1.5.2 Approval procedure for non-standard ships or ship operations The main purpose of this Classification Note is to provide guidance to assess sloshing in increased size LNG carriers and/ or in partially filled conditions. Both require at present a direct sloshing assessment for approval if the applications are not within the standard approved tank dimensions, ship size and/or filling restrictions. A vessel with normal size and regasification facilities is an example of a non-standard ship operation, since it will involve partial filling during the regasification process.

1.2 Membrane type LNG tanks


This Classification Note considers Liquefied Natural Gas (LNG) membrane tanks as used in LNG carriers and LNG floating production and/or storage units. Membrane type LNG tanks are defined according to the International Code for the Construction and Equipment of Ships Carrying Liquefied Gases in Bulk, see the IGC Code /1/, and the DNV Rules for Classification of Ships Pt.5 Ch.5 Liquefied Gas Carriers, /2/. They are non-self-supporting tanks which consists of a thin layer (membrane) supported through insulation by the adjacent hull structure. The membrane is designed in such a way that thermal and other expansion and contraction is compensated for without undue stressing of the membrane.

1.3 Purpose of this Classification Note


The purpose of this Classification Note can be summarised as follows:

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

The present Classification Note describes how a sloshing load and strength assessment can be conducted. The basic approach as outlined in this Classification Note is illustrated in Figure 1-6. A short summary is given of the 2nd column describing the sloshing load and strength assessment. The first step starts by describing the design basis and the limit states for the vessel. The result of this is a detailed list describing the specifications, which are used to develop a scope of work for the sloshing experiments. It is strongly recommended that this first phase is discussed with DNV Classification.
General procedure

Sloshing experiments are conducted resulting in a set of sloshing loads, which are used in the comparative sloshing load and strength approach. For the comparative approach sloshing loads for the reference vessel and the target vessel are needed. The structural strength of the containment system(s) is needed in case of a strength comparative approach. For the known GTT systems, i.e. NO96 and Mark III this is established and described in this Classification Note. If a new containment system is under consideration a strength assessment is to be conducted.
Review by Classification References to Classification Note
1.6 Design basis 1.7 Design Principles

Sloshing load & strength assessment


Design basis and limit state analysis (design vessel and operational specifications)

Prismatic membrane tank? yes Design & operational conditions (sloshing test scope)

Intermediate review by Classification

2.1 Limit state applications 3.1 Environmental modelling

Need sloshing assessment yes

Sloshing Analysis

3.2 Ship motion calculation 3.3 Sloshing model experiments Intermediate review by Classification

Sloshing loads Apply Classification Note 30.9 Known containment system?

3.4 Sloshing design loads 4 Pump tower design loads

no

Approval request

yes

Strength Assessment

Sloshing strength comparative assessment Approval? no yes

5 Structural reponse analysis 6 Ultimate strength of cont. systems 2.2 Strength assessment methodology 7 Stiffness and strength of hull structure 8 Response and strength of pump tower and supports Final review by Classification 1.4 Documentation

Analysis report

Figure 1-6 Schematic description of the application of Classification Note 30.9 to assess sloshing in membrane LNG tanks for the containment system load&strength assessment

1.5.3 Guidance for qualification assessments 1.5.3.1 Increased size LNG carrier The size of LNG carriers has gradually increased in the past period of time from ~120 000 m3 to ~150 000 m3, with 138 000 m3 often designated as the standard or conventional size. Sizes of 210 000 m3 and 216 000 m3 are being built (2005/2006) and vessels in the size of around ~250 000 m3 are being ordered. Regarding the building costs a minimum number of tanks is desired. However an increasing tank size implies increasing sloshing loads, which need to be assessed. In addition the ship sizes are significantly larger, which imply different ship motions. Consequently, an optimum ship design needs to be developed balancing the number of tanks and their dimensions with acceptable sloshing loads. 1.5.3.2 Offshore loading/unloading By constructing LNG terminals offshore, where off-loading can take place far from densely populated areas and busy ports and estuaries, the risk to the environment and the people at large is reduced. However, discharging offshore implies that LNG carriers and LNG floating receiving terminals will experience all filling levels inside their tanks while operating in a wave environment. A liquefied discharge offshore implies a

rather short period of time having to operate with partially filled tanks, the required weather window is thus more easily predicted. An on-board re-gasification and a direct send-out to the onshore gas grid implies that the vessel will be on-site for several days. This poses different requirements to the weather window, while the availability should be high as there is no storage but direct grid supply. 1.5.3.3 Low-partially filled tank on a particular trade route or for restricted operation For some particular trades, cargo maybe discharged at more than one terminal during one vessel voyage. Hence the carrier will be sailing with a partially filled tank(s) after having delivered part of its cargo at the first terminal. Most obvious would be to re-distribute the cargo to arrive at a situation such that the filling levels in all tanks are within the approved filling range. Re-distribution takes time and is an extra cost factor. It is therefore of interest to investigate whether approval can be obtained for a specific filling case on a specific route. If partial filling is an operational requirement for the owner, but this operation is not specified as a site-specific offshore discharge operation nor a trade on a particular route, the partial filling conditions may be assessed according to this Classification Note in order to determine an environmental envelope for safe operation of the carrier.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

1.5.4 Guidance for sloshing load assessments The main focus of the Classification Note is to provide guidance to assess sloshing for: increased size LNG carriers offshore loading/unloading partially filled LNG tanks on a particular trade route. However, this Classification Note provides detailed information on the specification, execution and analysis of sloshing experiments. Consequently, the Classification Note may be used to assess other applications than the specific applications listed above. It is strongly recommended to inform the Class Society at an early stage and discuss and agree on the following key issues: design basis design principles application of the principles sloshing analysis scope of work structural response and strength assessment procedure documentation.

It is assumed that the vessel trades on a world wide basis. A design life of 20 years is assumed. The tank fillings are between 90% and 98.5% of the tank height or below 10% of the tank length. 1.6.3 Containment systems For the purpose of this Classification Note, the target LNG carrier utilises a membrane type containment system applied in prismatic chamfered tank configurations. The structural strength assessment methods described in this Classification Note is applicable to the Gaztransport & Technigaz Mark III and NO96 tank insulation systems, and moderate evolutions of these systems in terms of modified plate thickness, modified support spacings, and modified configurations of internal load bearing structure. 1.6.4 Cargo tank environment The strength of the insulation system should be evaluated on the basis of a cargo temperature of -163C, and a temperature of 20C at the level of the steel structure in the cargo tank compartment. In the comparative approach equal densities of the cargo are to be assumed. A density of 500 kg/m3 is recommended to be used in the analyses. 1.6.5 Normal operation of LNG carriers With the exception of vessels trading as specified in 1.6.8, and the loading/unloading operational phase of vessels described in 1.6.7, it is assumed in this Classification Note that LNG carriers mainly trade with tank filling levels above 90% H, and only occasionally at filling levels between 90% H and the minimum acceptable upper range. With this as a basis, it is not correct to directly compare the frequently occurring reference case (95% H) and a less frequently occurring target case at fillings below 90% H at the same exceedance probability level. Corrections shall be made for the fact that the relationship between the short term and the long term (lifetime) expected extreme sloshing load is expected to be larger for the 95% H reference case than for a target case with filling levels below 90% H, see 1.7.8.2. 1.6.6 Increased size LNG carrier An increased size LNG carrier is an LNG carrier with a total cargo capacity larger than the currently approved standard size, see 1.5.3.1. The number of tanks may be specified by the designer. The design speed is to be specified by the designer. It is assumed that the vessel trades on a world wide basis, but the vessel may be sailing on a dedicated route, which may be more severe than a general trade on a world wide basis. A minimum design life of 25 years is assumed. The design life is used to define the sea state contour in order to identify the critical sea states for ship motion calculation and sloshing model experiments. The design life is also used for the fatigue calculations of the pump tower. The tank filling restrictions are to be specified by the designer. 1.6.7 Offshore loading/unloading The operation of the vessel while trading to and from a discharge location is considered to be a normal trading pattern, for which the design basis is covered by 1.6.2, concerning a conventionally sized LNG carrier, or 1.6.6, if the target vessel concerns an increased size LNG carrier intended for offshore discharge. At the discharge site the vessel will approach the mooring site by positional DP and/or tug assistance. Different mooring configurations may be used, like spread mooring or moored by a bow-turret. The vessel may be moored along

Any alternative procedure to assess sloshing and its effects is to be agreed upon with the Class Society before commencement of the studies. 1.5.5 Guidance for novel containment system strength assessments The comparative strength assessment methodology used throughout this Classification Note is based on operational experience with membrane type LNG tank systems, and this limits its application to the Gaztransport & Technigaz licensed Mark III and NO96 membrane type LNG insulation systems. It is, however, believed that the methodology in some cases may be extended to cover novel systems based on a similarity consideration with the existing systems. This possibility must be discussed and agreed with the Class Society based on a review of the system. The strength acceptance criteria documented in the Classification Note are specific to the Mark III and NO96 membrane type tank systems. The structure and formulation of the criteria may be useful in the formulation of the strength criteria for novel systems.

1.6 Design basis


1.6.1 Materials The methodology for the assessment of the hull structure and the pump tower presented in this Classification Note is based on the presumption that materials are selected in accordance with the requirements of the DNV Rules Pt.5 Ch.5. The methodology for the strength assessment of the containment systems is based on the presumption that the materials satisfy the requirements specified by the license owner, Gaztransport & Technigaz S.A.s. 1.6.2 Comparative basis (reference case) The basis for the safety equivalence assessment of new LNG carrier designs and operations must be consistent with the ship designs and operations that represent the safe and damage free operation experienced with the fleet of membrane type LNG carriers so far. The latter is referred to as the reference case. The reference vessel is a 4-tank LNG carrier with a total cargo capacity of 130 000 to 140 000 m3 utilising the Gaztransport & Technigaz Mark III or NO96 type containment system. The containment system should be of the same type for the reference case as for the target case. The selection of an appropriate reference vessel may be subject to discussion with DNV Classification. A design speed of 19.5 knots is assumed.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

side a fixed or floating structure or moored to a subsea buoy. The vessel may be moored on-site for hours or several days. Liquefied or re-gasified discharge may be conducted, through the midship manifold, flexible hoses or risers. The number of tanks may be specified by the designer. While on-site no forward speed is assumed. While on-site the LNG carrier might be disconnected at any stage of the discharge operation due to an emergency. The total design life of the vessel is divided into a period moored on-site and a trading period. For the on-site operation a minimum design life of 15 years is assumed when discharging is conducted by re-gasification or 1 year when liquefied discharging is conducted. While discharging, the vessel will experience all filling levels inside the cargo tanks. 1.6.8 Low-partially filled tank on a particular trade route or for restricted operation The operation of the vessel while trading with tank filling levels within the approved filling range is considered to be a normal trading pattern, for which the design basis is covered by 1.6.2, or 1.6.6 if the target vessel concerns an increased size LNG carrier. The number of tanks may be specified by the designer. The design speed is to be specified by the designer. It is assumed that the vessel trades on a world wide basis, but the vessel may be sailing on a dedicated route, which may be more or less severe than a general trade on a world wide basis. A minimum design life of 25 years is assumed. The tank fillings are to be specified by the designer.

The overall functional requirements that define the limit states for the structures governed by this Classification Note are: integrity of the primary and the secondary membranes the insulation system should be able to withstand the sloshing impact loads without suffering damages or excessive deformations compromising the support of the primary and secondary membranes the inner hull structure should be able to withstand the combined effect of global hull stresses and sloshing impact loads without compromising the support of the tank insulation system the pump tower including its supports should be able to withstand the combined effect of gravity, temperature loads, inertia loads, and drag forces induced by liquid sloshing without suffering damage. 1.7.2 Safety equivalent approach 1.7.2.1 General The methodology presented in this Classification Note is based on the principles of safety equivalence as outlined in the DNV Rules for Classification of Ships Pt.1 Ch.1 Sec.1 B305 and B306. For the sloshing exposed insulation system and hull structure components of membrane type LNG carriers, the equivalent safety principle will be satisfied on a comparative basis, where the requirements to the sloshing exposed structures of the investigated ship design or operating profile (target case) is determined based on a comparison with similar structures installed in conventional LNG carriers (reference case). The details of the comparative basis, the reference case, are specified in 1.6.2. The choice of a comparative approach is motivated, in particular, by the uncertainty identified in the determination of the sloshing impact loads, but also by certain aspects of the structural response and capacity assessment. It has the advantage that it allows for a simplified treatment of design parameters that exhibit similar qualitative behaviour in both the reference case and the target case. Design parameters that exhibit different qualitative behaviour in the reference and the target case must be included with its correct qualitative behaviour, or the comparative assessment must be designed in such a way that potential differences in the qualitative behaviour of the design parameters can be disregarded. 1.7.2.2 Containment system and hull structure The safety equivalence assessment can be carried out on two levels of sophistication, depending on the similarity in impact characteristics and structural arrangement between the reference and the target cases, as well as the relationship between the sloshing impact load level in the two cases, as follows: Load comparative approach: The assessment of the target case is carried out based on a comparison of the sloshing impact load level between the reference and the target case. It is required that the structural arrangement is identical between the reference case and the target case, and that it can be demonstrated that the sloshing impact events can be considered identical in terms of load area, impact pressure time histories and statistical characteristics. Application of the load comparative approach is limited to cases where the sloshing impact load level for the target case is lower than for the reference case. Strength comparative approach: The assessment of the target case is in general based on an assessment of the structural strength margin for the target, assuming that the strength is fully utilised for the reference case. More details about the two methods, and the conditions under which they can be applied, are given in 2.2. The containment system and its supporting hull structure

1.7 Design principles


1.7.1 Limit state design principles The sloshing assessment is to be based on the principles of limit state design. A limit state is defined as a condition beyond which the structure, or part of the structure, no longer satisfies its functional requirements. The limit states concern the safety of life, property (ship+cargo) and the environment. The following limit states should in general be considered. 1) ULS Ultimate Limit State The ULS concerns the ability of the structure to resist the action of the maximum expected loads or load effects during the design life of the ship. The limit state corresponds to the maximum load-carrying capacity (or strain or deformation) under intact conditions. 2) ALS Accidental Limit State The ALS concerns the ability of the structure to resist accident situations. The limit state concerns the safety of life, property and the environment in a) Intact conditions under the action of abnormal loads. b) Damaged condition under the action of normal loads. 3) FLS Fatigue Limit State The FLS concerns the ability of the structure to resist timevarying (cyclic or repeated) loading. No abnormal environmental conditions are considered for the ALS, e.g. freak waves, tsunamis, 10 000 years storms. The possible severe consequences of loss of containment of the membrane system, and the severe economical consequences of damages to the system implicates that the functional requirements defining the Accidental Limit State will be the same as the ones defining the Ultimate Limit State and the Fatigue Limit State.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

should be treated consistently by utilising the same load data for the two structures. The load level should be derived from the load capacity of the containment systems installed in the reference LNG carriers. The supporting inner hull structure should be dimensioned to provide at least the same support for the containment systems in the target LNG carrier as in the reference LNG carrier. This is to ensure that potential interaction effects between the hull and the containment system are equal or less severe in the target case as in the reference case. 1.7.2.3 Pump tower and supports For the pump tower main structure and supports, an absolute strength approach is feasible. The sloshing loads on the pump tower are due to sloshing motion rather than sloshing impact, and the loads can therefore be assessed with a larger degree of accuracy than for the containment system and the hull structure. The response assessment is also expected to be more accurate, since the response of the pump tower is expected to be quasi-static and linear elastic. The sloshing assessment of the pump tower may therefore be carried out using either the comparative approach as described above, or by a direct strength assessment. 1.7.3 Sloshing design loads The determination of the sloshing loads for assessment of containment system and hull strength should be based on irregular sloshing model experiments. For the pump tower, the sloshing loads may be determined from analysis. The ship motions as used for the sloshing experiments are to be determined using a verified ship motion program. The sloshing loads used in the comparative assessment are based on the worst short-term loads as encountered by the vessel during operation at sea. The sloshing loads considered for the assessment of the containment system are impact loads acting on small areas, typical box or panel sizes and smaller. Larger areas may be relevant for assessment of the hull structure. Further details of the execution of sloshing experiments is given in Section 3. No load combinations are considered for the sloshing impact loads when assessing the containment system structural integrity. 1.7.4 Structural response analyses To be consistent with the principles of the comparative assessment, the structural response analysis methodology needs to be capable of accurately predicting the structural response in the response range up to where damages are likely to occur in the structure. Depending on the response characteristics of the considered structure, it may be required to consider non-linear structural response. The structural response assessment needs to consider and include all physical effects that affects the relative structural response between locations relevant for the assessment of the individual failure modes, and effects that potentially give different results for the reference and the target cases. The most important effects that must be included are: effects of the temperature variations through the thickness of the insulation system dynamic response effects. Details on how to handle these effects are given in Section 4. 1.7.5 Ultimate capacity models The ultimate capacity models should be able to predict the limit states associated with damages to the containment system. All physical effects that significantly affects the relative struc-

tural strength between the individual failure modes, and effects that may give different results for the reference and the target cases should be included in the capacity models. The most important effects that must be included are: effects of the temperature variations through the thickness of the insulation system strength differences caused by differences in the dynamics of the response. Details on how to handle these effects are given in Section 6 for the containment systems. 1.7.6 Fatigue capacity models The fatigue assessment of the containment system must be carried out using a semi-direct calculation. For consistency with the comparative ultimate strength assessment of the containment system the long term response spectrum for the target case should be scaled so that the maximum response in the spectrum corresponds to the design load used in the ultimate strength assessment. Acceptance is, however, based on direct evaluation of the calculated damage for the target case without consideration of the reference case. The cumulative damage (D) from repeated sloshing impacts should be calculated according to the Miner-Palmgren theory in combination with S-N curves, a characteristic stress range, and a long term response distribution curve, as follows:
n D= i Ni i

where the summation is over a number of i stress intervals, ni is the number of load cycles within each stress interval, and Ni is the number of cycles to failure for the constant stress range of that interval. 1.7.7 Design format 1.7.7.1 Load comparative approach The following acceptance criterion should be satisfied in cases where the load comparative safety equivalent approach is applicable.

p tar F

p ref

where: ptar is the sloshing impact pressure for the target LNG carrier. pref is the sloshing impact pressure for the reference LNG carrier. F is the partial load factor, defined in 1.7.8.2. M is the partial resistance factor, defined in 1.7.8.3. The criterion should be satisfied for the entire range of load areas relevant for the unit dimensions of the containment system. 1.7.7.2 Strength comparative approach The criteria for the assessment of the equivalent safety of the cargo containment system and hull structure between the reference and the target case are formulated on a partial safety factor format, as follows:

S ( p DAF F )

Rc

where: S is the structural response, in general a non-linear function of the dynamic load. p is the sloshing impact pressure scaled according to the com-

DET NORSKE VERITAS

10

Classification Notes - No. 30.9 June 2006

parative procedure described in 2.2. DAF is the dynamic load factor. Rc is the capacity in terms of the considered response parameter.

F is the partial load factor, defined in 1.7.8.2. M is the partial resistance factor, defined in 1.7.8.3.
1.7.7.3 Pump tower assessment The strength assessment of the pump tower and supports may be carried out using either a comparative approach or an absolute approach. The absolute approach will usually be most convenient, since the load and strength need to be calculated for the target case LNG carrier only. The strength is satisfactory if:

S F
where: S is the structural response

Rc

inherent uncertainties in the determination of design loads from statistical analysis of sloshing impact load time histories uncertainties in the determination of (long-term representative) design loads due to a limited test possibilities. (Physical limitations simply prohibit that all ship and sea state conditions can be assessed) potential differences in scaling from model to full scale in the reference and the target case (model scaling law, hydro-elastic effects) differences in statistical distribution of sloshing impact loads between the reference and the target case uncertaincies in the relative response assessment between the reference and the target cases. Applies only if the dimensioning impact area differs significantly between the two cases. differences in acceptance level between the failure modes in the procedure. Applies if the governing failure mode is changed between the reference case and the target case, and is caused by either: differences in confidence level for the governing structural capacity differences in the accuracy of the governing structural response. 1.7.8.2 Load factors for strength assessment of containment system The load factor to be used in the strength assessment of the containment system for the target case should be determined as:

Rc is the capacity in terms of the considered response parameter

F is the partial load factor, defined in 1.7.8.2. M is the partial resistance factor, defined in 1.7.8.3.
In a direct strength assessment, the absolute magnitude of the loads is of major importance, and a thorough investigation of the loads is necessary. Larger load factors need to be used in the absolute approach than in the comparative approach, in order to account for the uncertainties related to the load level. If the comparative approach is followed, as recommended for the containment system and the hull strength, the load and strength of the pump tower in the reference case are compared with the load and strength of the pump tower in the target case. The utilization for the target case, multiplied with a safety factor, should be lower than for the reference case. The strength is satisfactory if:

F = F 0 F1 F 2
where: F0 reflects the inherent statistical uncertainty in the experimental load assessment. F1 reflects the potential difference in the scaling law between impact areas of different sizes. (Scaling factor is expected to be higher for small impact areas than for large. The effect is therefore relevant only if the design area for the target case is larger than for the reference case.) F2 reflects the possible difference in the sloshing load probability distributions between the reference and the target case, and should in general be determined based on processing of the load data, see 3.4, as well as a consideration of the operational profile. Two main reasons for the possible difference are: 1) The sloshing loads as derived from a short-term assessment are not considered identically representative for the long-term sloshing loads. 2) The short-term sloshing load exceedance probability functions for the reference and the target case show significant different trends, such that the reliability of expected extreme values is not the same. 3) Combination of 1) and 2). The following paragraph explains these issues in more detail. In standard wave load analysis a wave scatter diagram is used to describe the long-term wave environment. Using linear response transfer functions a complete wave scatter-diagram can be assessed and the long-term response amplitude distribution can be calculated. From this distribution, the lifetime expected extremes can be determined, e.g. with a 25 years return period or at 10-8 probability level. Another approach is to determine the most critical sea state from the scatter-diagram and assess only the short-term statistics for that sea state. For linear responses this typically leads to a difference of 10 to 20 percent. Or in other words this shortterm expected extreme is some 10% to 20% lower than the long-term expected extreme. However, for nonlinear respons-

S R c
where:

S compare R tar c

ref

S is the structural response Rc is the capacity in terms of the considered response parameter

compare is a load factor that reflects the statistical uncertainty in the comparative load assessment
In the comparative approach, the uncertainty related to load level is reduced, since the main concern is the load increase from the reference case to the target case, rather than the absolute load level. 1.7.8 Partial safety factors 1.7.8.1 Uncertainties in the comparative approach The load and resistance factors shall be applied in the strength assessment of the target case to reflect the uncertainties in the comparative procedure and hence make sure that the safety level in the reference case is maintained in the target case. The main uncertainties involved in the comparative strength assessment of the tank systems of the membrane type LNG carriers are: uncertainties in the relative impact load assessment between the reference case and target case

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

11

es the response behaviour may be characterised by response amplitude distributions having more flat tails (i.e. Weibull fitted function with low values for the slope factor). This is further discussed in 3.4. 1.7.8.3 Resistance factors for strength assessment of containment system The resistance factor to be used in the strength assessment of the containment system for the target case should be determined as:

supports are given in Table 1-3.


Table 1-3 Applicable values of the load and resistance factors for pump tower and supports ULS FLS F, sloshing loads 1.3 1.0 F, wave induced loads (inertia, hull girder, pressure) 1.15 1.0 F, other loads (gravity, thermal) 1.0 1.0 M 1.15 1.0 compare 1.1 1.0

M = M 0 M1 M 2
where:

M0 reflects the general uncertainty in the response assessment


procedure. impact areas of different sizes. It is expected that the insulation systems are more redundant (have larger potential for redistribution of loads) for impacts on small areas of the structure than for impacts on large areas. The effect is therefore relevant only if the design load area for the target case is larger than for the reference case.

For FLS assessment, a design fatigue factor shall be used. The calculated damage, multiplied with the design fatigue factor (DFF), should be less than or equal to 1.0:

M1 reflects the difference in structural redundancy between

D DFF 1.0
The design fatigue factor should be taken as /11/: DFF = 1.0 for parts accessible for inspection DFF = 3.0 for parts not accessible for inspection, but not substantial consequence of failure DFF = 10.0 for parts not accessible for inspection, and substantial consequence of failure

in the acceptance criteria between the reference and the target case. The factor is relevant only if the critical failure mode changes between the reference and the target case. 1.7.8.4 Design values of load and resistance factors for containment system and hull structure The values of the basic load and resistance factors F0 and M0 applicable to the strength assessment of the target case are summarised in Table 1-1.
Table 1-1 Applicable values of the basic load and resistance factors for containment system ULS ALS FLS F0 1.1 1.0 1.0 M0 1.0 1.0 1.0

M2 reflects potential difference in statistical confidence level

2. Application of the Principles


This section describes the application of the design principles as outlined in 1.7 using the design basis as given 1.6.

2.1 Limit state applications


This section describes the application of the limit states, as described in 1.7.1, to specific LNG carrier applications. Consideration of fatigue limit states (FLS) will normally not be required for the membrane insulation systems. This conclusion is based on technical investigations carried out by DNV that can be generalised as a consequence of both the characteristics of the sloshing phenomenon as well as properties of the comparative strength assessment methodology. The following characteristics of the sloshing phenomenon is important: 1) Significant liquid sloshing motion requires a certain level of ship motions combined with motion periods in the same range as the natural period for the sloshing motion in the tank (resonance). The number of sloshing impacts during the lifetime of a vessel is therefore low compared to what is common for fatigue exposed structures. 2) Extreme sloshing impact loads do not only require the presence of extreme sloshing motion in the tank, but also requires a certain unfavourable fluid surface geometry at impact to avoid gas cushioning. The few high impact loads of this character occurring during the lifetime of the vessel are significantly higher than the other impact loads. The long term sloshing load distribution will therefore show a rapid decrease in load level with increasing number of cumulative impacts. 3) Variations in wave heading and filling levels will lead to variations in the sloshing exposed locations in the tank. This will further limit the number sloshing impacts per impact location in the LNG tank. A typical shape of the long term sloshing load distribution is shown in Figure 2-1.

The remaining load and resistance factors are not specified in advance and need to be determined from the analyses. 1.7.8.5 Load and resistance factors for stiffness and strength assessment of hull structure The load factor to be used in the stiffness and strength assessment of the hull structure is to be taken as:

F = DAF
where:

DAF is a partial factor accounting for potential difference in

dynamic response amplification between the target and the reference case.

Applicable load and resistance factors for the hull structure are given in Table 1-2.
Table 1-2 Applicable values of the load and resistance factors for the hull structure ULS DAF, hull stiffness 1.15 DAF, hull strength 1.3 M 1.0

1.7.8.6 Load and resistance factors for pump tower and supports Applicable load and resistance factors for the pump tower and

DET NORSKE VERITAS

12

Classification Notes - No. 30.9 June 2006

Scaled 20 years vs 40 years response spectra

20 years operation 40 years operation

Impact pressure

LOG(number of cycles)

1.0E+00 1.0E+01 1.0E+02 1.0E+03 1.0E+04 1.0E+05 1.0E+06

Number of cycles

Figure 2-1 Typical long term distribution of sloshing impact loads

Figure 2-2 Comparison of scaled long term response distributions for 20 years and 40 years North Atlantic operation.

For load distribution like this, representative for 20 years operation, DNV investigations have shown that: the main contribution to the cumulative fatigue damage comes from the upper range of the response spectrum (the highest loads) fatigue is not a relevant failure mechanism when the hot spot stress level is governed by the ultimate strength requirements as is ensured by the comparative strength assessment. This covers most of the relevant hot-spots for the considered insulation systems the fatigue damage at hot spots not governed by the ultimate strength requirements is acceptable for load levels up to what is considered the limits for strengthening of the insulation systems. These conclusions can be extended to govern longer operational times for the target case, e.g. 40 years operation by the following argument. A key component of the comparative strength assessment is the scaling of the sloshing impact load distribution to match the ultimate capacity of the containment system for the reference case, followed by a strengthening of the containment system to sustain the higher load in the target case as is expected when the service life is increased from 20 to 40 years. Consequently, the extreme response in the long term response distributions for the target case will either be equal to or smaller than the extreme response in the reference case (considering the same governing failure mode). In fact, the difference in partial safety factors between the ULS and the FLS will ensure that the target case extreme response is smaller than the reference case extreme response. It follows from this that the long term response spectrum governing the fatigue of the target case will be almost equal to or lower than the long term response spectrum for the reference case in the governing upper part of the distribution. This is illustrated in Figure 2-2, which shows a comparison of scaled long term response distributions for 20 and 40 years operation. The effect of increased service life on the fatigue performance of the insulation systems is hence be small. These conclusions apply to vessels and operations where the long term distribution of structural response for the target case is governed for normal trade operation of an LNGC. This is specified in more detail in the following sections. Vessels involved in offshore loading and discharge operations as well as vessels occasionally trading with reduced fillings are considered to fall into this category. Long term response distributions for the reduced filling operations are considered governed by the less severe short term distributions.

2.1.1 Comparative design reference case Only an ULS condition is considered for this case. No FLS and ALS conditions are considered. Drifting in beam seas due to an engine black-out is considered to occur in a less severe sea state that the ULS requirements for beam seas. 2.1.1.1 ULS condition The maximum ULS condition to be considered is 20 years world-wide operation. DNV Classification Note 30.5 /3/ provides a wave scatter diagram, which can be used to represent the long-term world-wide wave climate. The short-term sea states are to be modelled by the two-parameter Pierson-Moskowitz spectrum, see reference /3/. Waves are to be modelled as long-crested waves. Wave headings from stern quartering (45 degrees) to head waves (180 degrees) are to be considered. A minimum stepsize of 15 degrees is recommended. For beam sea conditions (60-120 degrees) the 20-years contour of significant wave heights may be reduced by 20% to reflect voluntary heading changes in severe beam seas. For quartering sea conditions (30-60 and 120-150 degrees) the 20-years contour of significant wave heights may be reduced with a linear varying reduction starting from 0% at 30 and 150 degrees to a reduction of 20% at 60 and 120 degrees wave hading. A ballast loading condition and a loaded condition are to be considered. For the latter a fully loaded or partly loaded condition maybe used. The loading condition giving the severest ship motions is to be used to determine the ship motions for the sloshing analyses. The speed of the vessel is to be reduced to design speed. A more refined speed-sea-state curve may be used, as for example depicted in Figure 2-3. Such alternative speed-sea-state curves are to be discussed with the Class Society. There are two aspects which are of importance to consider when applying a stepwise speed-sea-state curve in a comparative analysis. Firstly, the speed can have an important effect on the sloshing impact loads. In standard wave load analysis a large ship speed gives conservative design motions and loads in an absolute strength assessment. In the present guideline a comparative approach is followed, not an absolute sloshing load and strength assessment. This implies that the choice of a conservatively high ship speed for the reference ship implies that large sloshing loads are allowed for the target case. Secondly, the sloshing procedure, as recommended in Section 3, identifies the worst sea state on a contour of significant wave heights and periods. Applying this procedure for the reference and the target case

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

13

may lead to a situation where the ship speed for reference case is higher than for the target case, which for some cases is unreasonable or unacceptable.
Ship speed
Design speed

Waves are to be modelled as long-crested waves. For a bow-turret moored LNG carrier wave headings from beam seas (90 degrees) to head waves (180 degrees) are to be considered. A minimum stepsize of 15 degrees is recommended. A bow-turret moored LNG carrier, freely able to weathervane, is likely to experience predominantly a head waves conditions. A low-frequency drift study may be used to determine the probability distribution for headings from beam to head waves. Such a study needs careful modelling or assessment of at least the following aspects: draft and trim variations joint modelling of wave, current and wind mooring characteristics shallow water effects (if present) (wave and ship dynamics).

Manoeuvring speed

Significant wave height

Figure 2-3 Example speed-sea-state curve

2.1.2 Increased size LNG carrier Only an ULS condition is considered for this case, no FLS and ALS conditions are considered. 2.1.2.1 ULS condition The minimum ULS condition to be considered is 25 years North-Atlantic operation. IACS Rec. No. 34, see /4/, or DNV Classification Note 30.5, see /3/, provide wave scatter diagrams for the North-Atlantic ocean, which can be used to represent the long-term North-Atlantic wave climate. The short-term sea states are to be modelled by the two-parameter Pierson-Moskowitz spectrum, see reference /3/. Waves are to be modelled as long-crested waves. Wave headings from stern quartering (45 degr) to head waves (180 degrees) are to be considered. A minimum stepsize of 15 degrees is recommended. For beam sea conditions (60-120 degrees) the 25-years contour of significant wave heights may be reduced by 20% to reflect voluntary heading changes in severe beam seas. For quartering sea conditions (30-60 & 120-150 degrees) the 25years contour of significant wave heights may be reduced with a linear varying reduction starting from 0% at 30 and 150 degrees to a reduction of 20% at 60 and 120 degrees wave hading. A ballast loading condition and a loaded condition are to be considered. For the latter a fully loaded or partly loaded condition maybe used. The loading condition giving the severest ship motions is to be used to determine the ship motions for the sloshing analyses. The speed of the vessel is to be reduced to half design speed. A more refined speed-sea-state curve may be used as discussed in 2.1.1.1. 2.1.3 Offshore loading/unloading Only ULS and ALS conditions are considered for this case, no FLS condition is considered. 2.1.3.1 ULS condition The minimum ULS condition to be considered is 20 years sitespecific operation. The site specific wave environment is to be used for the modelling of the long-term wave climate. This data should be derived from at least a 20 years wave measurement period. The short-term sea states are to be modelled by a wave spectrum, which is representative for the offloading site. If no specific wave spectral information is available the JONSWAP wave spectrum may be used with a peakness parameter of 3.3, see /3/.

For an LNG carrier, which is moored in such a way that she is not able to weathervane, wave headings from following (0 degrees) to head waves (180 degrees) are to be considered. A minimum stepsize of 15 degrees is recommended. For an LNG carrier that maintains a near head waves condition by using DP assisted mooring it is recommended to contact the Classification Society in order to discuss a relevant set of wave headings that need to be considered in the sloshing experiments. A ballast loading condition, fully loaded and intermediate loading conditions are to be considered. The loading condition giving the severest ship motions is to be used to determine the ship motions for the sloshing analyses. While on-site no forward speed is to be considered. 2.1.3.2 ALS condition drifting in beam seas An emergency situation may force an emergency disconnection of the vessel from its mooring, leading to a situation where the vessel is drifting in beam seas without propulsion and manoeuvring capabilities. The minimum ALS condition to be considered is 1 year sitespecific operation. The site specific wave environment is to be used for the modelling of the long-term wave climate. The short-term sea states are to be modelled by a wave spectrum, which is representative for the offloading site. If no specific wave spectral information is available the JONSWAP wave spectrum may be used with a peakness parameter of 3.3, see /3/. Waves are to be modelled as long-crested waves. The vessel is considered in a beam waves condition without forward speed. A ballast loading condition, fully loaded and intermediate loading conditions are to be considered. The loading condition giving the severest ship motions, especially roll and sway, is to be used to determine the ship motions for the sloshing analyses. 2.1.3.3 ALS condition maintaining a head sea condition at manoeuvring speed An emergency may give an emergency disconnection of the vessel from its mooring leading to a situation where it will, as soon as possible, maintain a head sea condition with a low forward speed to maintain manoeuvrability. The minimum ALS condition to be considered is 1 years sitespecific operation. The site specific wave environment is to be used for the modelling of the long-term wave climate. The short-term sea states are to be modelled by a wave spectrum, which is representative for the offloading site. If no specific wave spectral information is available the JONSWAP wave spectrum may be used with a peakness parameter of 3.3, see /3/.

DET NORSKE VERITAS

14

Classification Notes - No. 30.9 June 2006

Waves are to be modelled as long-crested waves. The vessel is considered in a head waves condition with a minimum forward speed of 5 knots. A ballast loading condition, fully loaded and intermediate loading conditions are to be considered. The loading condition giving the severest ship motions, especially roll and sway, is to be used to determine the ship motions for the sloshing analyses. 2.1.4 Low-partially filled tank on a particular trade route or for restricted operation Only ULS and ALS conditions are considered for this case, no FLS condition is considered. 2.1.4.1 ULS condition The minimum ULS condition to be considered is 25 years route-specific operation. The wave environment to be modelled is to represent the route-specific conditions. Reference / 3/ provide wave scatter-diagrams that can be used to develop a route-specific scatter diagram. If no specific route is specified a North-Atlantic operation is to be used. IACS rec.34, see /4/, or DNV Classification Note 30.5, see /3/, provide wave scatter diagrams for the North-Atlantic ocean, which can be used to represent the long-term North-Atlantic wave climate. The short-term sea states are to be modelled by the two-parameter Pierson-Moskowitz spectrum, see reference /3/. Waves are to be modelled as long-crested waves. Wave headings from stern quartering (45 degr) to head waves (180 degrees) are to be considered. A minimum stepsize of 15 degrees is recommended. A ballast loading condition and a loaded condition are to be considered. For the latter a fully loaded or partly loaded condition maybe used. The loading condition giving the severest ship motions is to be used to determine the ship motions for the sloshing analyses. The speed of the vessel is to be reduced to design speed. A more refined speed-sea-state curve may be used as discussed in 2.1.1.1. 2.1.4.2 ALS condition drifting in beam seas An emergency or damage may lead to a situation where the vessel is drifting in beam seas without propulsion and or manoeuvring capabilities. The minimum ALS condition to be considered is 1 year routespecific operation. The route-specific wave environment is to be used for the modelling of the long-term wave climate. If no specific route is specified a North-Atlantic operation is to be used as described in 2.1.4.1. The short-term sea states are to be modelled by a wave spectrum, which is representative for the specific route. If no specific wave spectral information is available the PiersonMoskowitz wave spectrum may be used for open ocean conditions and the JONSWAP wave spectrum may be used with a peakness parameter of 3.3, for sheltered and or near shore waters, see /3/. Waves are to be modelled as long-crested waves. The vessel is considered in a beam waves condition without forward speed. A ballast loading condition, fully loaded and intermediate loading conditions are to be considered. The loading condition giving the severest ship motions, especially roll and sway, is to be used to determine the ship motions for the sloshing analyses.

ing hull structure are summarised in the following. The comparative aspect of the methodology lies in the assessment of the reference case, which is used to harmonise the load and the capacity to be consistent with the operational experience of membrane type LNG carriers operated to date. In practice this is achieved by scaling the sloshing design loads, but the harmonisation also covers other systematic uncertainties in the procedure. The sloshing impact load is, however, considered to represent the largest uncertainty in the procedure. The methodology can be summarised as follows: Reference case (see Figure 2-4): 1) Establish a curve relating sloshing impact pressure and sloshing exposed area based on experimental results as described in more detail in Section 3. Load factors should be disregarded in this step. 2) Establish a curve relating the impact load capacity of the insulation system and the sloshing exposed surface area of the structure as described in Section 6. Resistance factors should be disregarded in this step. 3) Establish the ratio between the load and the capacity for the entire range of load areas, and identify the maximum ratio between load and capacity. Denote this ratio by comp. 4) Scale the load uniformly for all load areas using the maximum identified ratio between the load and the capacity. The scaled load will now for any load area size be lower than the ultimate capacity of the insulation panels. This step is motivated by the damage free operational experience with the membrane type LNG carriers. The resulting load curve should be the basis for the strength assessment of the insulation system and its supporting hull structure.
Load/ Capacity Measured load Scaled load Capacity standard

Scaling of load to ULS capacity

Load exposed area

Figure 2-4 Scaling of loads for reference case to the ULS capacity

Target case (see Figure 2-5): 1) Establish a curve relating sloshing impact pressure and sloshing exposed area based on experimental results as described in more detail in Section 3. 2) Scale the load using the maximum ratio, comp, between load and response determined from the reference case. 3) Carry out a strength assessment of the insulation system according to the procedure specified in Section 6. 4) Carry out the necessary reinforcement of the insulation system so that the load for any load area is lower than the ultimate capacity of the insulation panels. 5) Carry out a strength assessment of the supporting hull structure according to the procedure specified in Section 6.

2.2 Strength assessment methodology


The overall methodology for the comparative strength assessment of the membrane type insulation systems and its support-

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

15

going through the above mentioned parameter combination.


Load/ Capacity Measured load Scaled load Capacity standard Capacity reinforced

Reinforcement of insulation sysetm

Scaling of load according to reference case

Load exposed area

Figure 2-5 Scaling of loads for target case using the same scale factor as for the reference case. Strengthen containment system according to the scaled load curve

Table 3-1 Standard sea state contours for the North-Atlantic and for world-wide operations North Atlantic North Atlantic DNV CN30.5 IACS Rec.34, /4/ IACS Rec.34, /4/ World-Wide, /3/ 40 years 25 years 20 years Tz (s) Hs (m) Tz (s) Hs (m) Tz (s) Hs (m) 4.5 2.8 4.5 2.7 4.5 4.4 5.5 5.6 5.5 5.3 5.5 6.6 6.5 8.4 6.5 8.0 6.5 8.6 7.5 10.8 7.5 10.4 7.5 10.2 8.5 12.6 8.5 12.2 8.5 11.3 9.5 13.9 9.5 13.6 9.5 12.0 10.5 14.8 10.5 14.4 10.5 12.3 11.5 15.3 11.5 14.8 11.5 12.3 12.5 15.3 12.5 14.8 12.5 12.0 13.5 15.0 13.5 14.5 13.5 11.4 14.5 14.3 14.5 13.7 14.5 10.4 15.5 13.2 15.5 12.5 15.5 8.7 16.5 11.3 16.5 10.2 -

The comparative strength assessment should be carried out for all insulation structure elements that will experience sloshing impact loads in the cargo tank. In practice this means the dedicated transverse and longitudinal corner/knuckle structure and standard flat wall structure adjacent to the corner/knuckles. The specific locations are determined by the applicable tank filling limitations and the operation of the vessel. A single comparative load scaling factor, comp, representative for the weakest element of the considered insulation structures should be applied in the assessment of all relevant insulation structures of the target vessel. This means that potential strength margins determined for the reference case can be utilised in the target case.

If another return period contour is needed for the North-Atlantic or the world-wide scatter diagram the following Weibullconditional log-normal distributions may be used to establish new contours.

F (H s , Tz ) = F (H s ) F (Tz H s )
H s

F (H s ) = 1 e
f (Tz H s ) = 1

3. Sloshing Impact Design Loads


3.1 Environmental modelling
The wave environment is to be described by a wave scatter diagram. The scatter diagram gives the probability of occurrence of short-term sea states. The basis for the scatter diagram, i.e. measurements and/or observation period, is to represent the long-term wave climate. The number of occurrences in the scatter diagram should therefore be sufficiently large; for example 100 000 occurrences considering a duration of 3 hours per occurrence. This reveals a total duration of ~34 years. Contours with a fixed return period can be established from the wave scatter diagram, like a 25 or 40 years contour. If a return period is to be determined while the scatter-diagram is established based on a lesser period of time, an extrapolation of the wave statistical data may be applied. Standard contours are given in Table 3-1. One method for establishing the contour is the constant probability density approach.. This is outlined in the following. First, the joint environmental model of the sea state variables of interest is determined. Second, the extreme value for the governing variable for the prescribed return period and associated values for other variables are estimated, e.g. the 100-year value for Hs and the conditional median for Tz. Third, the contour line is estimated from the joint model or scatter diagram as the contour of constant probability density

Tz 2

1 ln Tz 2

H f (H s ) = s

1 H s

a = a0 + a1H s 2

= b0 + b1H sb2
Table 3-2 Standard sea state contours for the North-Atlantic and for world-wide operations North Atlantic IACS DNV CN30.5 Rec.34, /4/ World-Wide, /3/ 3.041 1.798 1.484 1.214 0.661 0.856 a0 0.70 -1.010 a1 1.27 2.847 a2 0.131 0.075 b0 0.1334 0.161 b1 0.0264 0.146 b2 - 0.1906 - 0.683

DET NORSKE VERITAS

16

Classification Notes - No. 30.9 June 2006

The short-term sea states are described by a wave spectrum, which is a function of the mean wave period and a significant wave height. Different wave spectra may be used depending on the condition that is investigated, see 2.1. Recommendations for wave spectral formulation are given in /3/.

3.2 Ship motion calculation for sloshing tests


The motions for a sloshing test are to be determined by a dedicated ship motion analysis using a verified and validated computer code recognized by the Class Society. The predictions are to provide linear ship motion transfer functions, which can be used to generate tank motions for specified short-term sea states. The ship motion calculations are to be conducted for the loading condition and speed as specified for the associated sloshing test. In head and near head waves the surge motion is an important motion or excitation mode for sloshing. The surge motion equation is to be modelled properly in the code. It is therefore recommended to use a 3-dimensional ship motion program. In quartering and beam seas the roll motion is an important motion or excitation mode for sloshing. Roll damping, additional to the potential damping, is to be accounted for in the roll motion prediction. The effects of bilge keels, eddies, fins, skegs, etc. and their possible speed dependence is to be included. This inclusion can be done using linearization techniques. This linearization may be done for varying sea state severities. The coupled motion effect due to partially filled tanks may be included in the motion predictions. In case of a vessel moored side-by-side with another vessel the hydrodynamic interaction is to be taken into account in the motion predictions. Irregular time series of the ship in a seaway are to be calculated by combining a specified wave spectrum with the motion transfer functions. The calculated motions are to be calculated for the motion reference point of the sloshing rig. The motions are to be Froude-scaled for input to the sloshing rig. Repetition of the irregular motions of the tank are to be avoided. This can easily be achieved by using an inverse Fourier approach using a very small frequency stepsize, i.e. a large number of frequency components. Alternatively, the irregular time series can be composed by a linear superposition of harmonic components with unequidistant frequency stepsizes.

form, loaded with its maximum payload, is to be equipped with a motion response unit to verify if the generated tank motions correspond with the input motion signals. The size of the motion platform is preferably as large as possible. At least a scale of 1:50 for the sloshing model test tank should be used. Preferably a six-degrees of freedom motion platform is to be used. For a proper comparative approach the sloshing tests need to be executed at the same model scale and model test set-up for the reference case and the target case. The sloshing model tank is to be made sufficiently stiff such that natural frequencies of the tank and its parts do not interfere with the sloshing impact pressures. It is recommended to manufacture the sloshing model test tank of transparent material such that the behaviour of the fluid motions in the tank can be observed. The sloshing model tank is to be prepared for the mounting of pressure sensors, single and cluster-wise, at various locations throughout the tank in order to be able to measure sloshing pressures at possible critical areas. The local structure of the tank, where pressure transducers can be positioned, are to be sufficiently stiff to avoid fluid-structure interaction. The sloshing model test tank should be equipped with filling and emptying taps, such that level adjustments are made easily and accurately. At high filling (~95% of tank height) significant differences in impact pressures can be measured for small filling differences. Consequently, the accuracy of the filling level is an important aspect. In case sloshing tests are to be conducted with varying ullage gas conditions the sloshing model tank structure should be designed for depressurisation. The sloshing model tank should be prepared with taps for gas filling and emptying. When using different gases or fluids to vary the ullage or fluid conditions inside the tank, preparations must be made to contain the gases or fluids after testing to conform with environmental regulations if applicable. 3.3.2 Scaling of sloshing pressures ullage gas and fluid conditions The application of a comparative approach as outlined in 2.2 for the assessment of the containment structural strength does not require a scaling of the sloshing model-test pressures to full-scale. Sloshing model-test studies investigating sloshing impact pressures for various ullage conditions, see e.g. /5/, indicate that the scaling may be a function of, among other aspects, the load area. In case the comparative approach results in different critical load areas for the reference and the target case the potential difference in scaling is to be addressed. 1.7.8.2 addresses this aspect with the load factor F1. In order to be consistent in a comparative approach the model scale for the reference case and the target case is to be equal as well as the fluid and ullage conditions. Sloshing impacts may show an oscillating behaviour in the time history of the sloshing impact pressure, caused by entrapped gas. It is recommended to apply a similar scaling for sloshing impacts with or without this oscillating behaviour, unless experiments and documentation may argument a change of scaling. 3.3.3 Instrumentation and data acquisition Pressure transducers are to be mounted to measure the fluid pressures inside the tank. The sensing area should be positioned flush with the inner tank wall. The shock resistance of pressure sensors should be sufficient not to interfere with the impacts and accelerations expected.

3.3 Sloshing model experiments


Sloshing model experiments are tests where a model of a ship tank is placed on a motion platform, which simulates the motions of a ship in a seaway. The model tank is filled with liquid and is equipped with pressure transducers to measure the fluid pressures inside the tank. Sloshing model experiments are to be carried out in a laboratory environment with the necessary safety facilities and regulations. 3.3.1 Sloshing experimental lay-out Sloshing experiments require a motion platform capable or simulating the motions of the tank on-board a ship in a seaway. The motion platform should be able to simulate both regular and irregular motions as specified by time series from a ship motion analysis. The maximum strokes and angles for the motion platform are to be designed such that the worst ship motions anticipated for the intended sloshing tank model size can be simulated. It is recommended not to dimension the rig capabilities only by regular maximum expected motion amplitudes. Especially in beam sea conditions, the combination of large heave, large sway and large roll angles is expected to set capability envelope for the rig. The motion control needs to be designed such that the specified ship motions can accurately be simulated. The motion plat-

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

17

The pressure sensors should be applicable in a wet and/or corrosive environment. The rated pressures and the maximum measurable pressure should be sufficiently above the expected pressures, which will be measured inside the tank. The frequency response of the pressure sensor and signal amplifier must be sufficiently high to capture the pressure fluctuations at least at the intended sampling frequency, i.e. the signal should not be conditioned. The pressure sensors should preferably be sufficiently insensitive to temperature fluctuations or this should be corrected for in the testing. The sensing area should be in direct contact with the medium inside the tank. No protective cover or similar caps are to be used, which can affect the dynamics of the measurements. Pressure transducers should be calibrated before use. In addition to a static calibration, a dynamic calibration is recommended. This can be accomplished by use of drop-tests, where the pressure sensors are mounted in a wedge-shaped section being dropped down onto a flat water surface. The temporal and spatial shapes of the pressure pulse are functions of the impact velocity and wedge deadrise angle. They can easily be accurately calculated for a small deadrise angle, and provide an excellent means of assessing the pressure sensor performance. The size of pressure transducers should be small to be able to measure local impact pressures and to be able to position several pressure transducers close together. At least 9 sensors should cover a full-scale square area of 1.5 m2. Common hotspots for impact loads are at intersections between the tank roof or chamfers and the tank walls. The sensors should be placed sufficiently close to these intersections so that the corner panels are covered. A 4x4 cluster of sensors (16 total) should be used in order to facilitate a load versus area estimate based on 9 sensors for both the perimeter zone and the internal area excluding corner boxes. The data acquisition system should be able to sample the pressure sensors at a sufficiently high sampling rate to capture short impact pressure time histories. Typically sampling rates of ~10 kHz to ~20 kHz are required for sloshing tests at scales of 1:15 to 1:50 respectively. The pressure time histories and the tank motion histories are to be stored as raw data for further post-processing. Video capturing of the sloshing tests is recommended in order to study the resulting fluid behaviour inside the tank for specific conditions, like for example the wave heading, the sea state, the filling level, etcetera.

3.3.4 Data analysis statistical post-processing The stored pressure time histories are to be statistically postprocessed to determine impact pressure statistics. The sloshing impact peaks in the time series are to be identified. A peak-over-threshold method is recommended to identify the impact peaks and separate them from noise in the signal. Maxima are to be identified as illustrated in Figure 3-1. A moving time window may be used to identify only these global maxima within the window settings. Alternatively a minimum required time-step between global impact peaks may be set. A variation of the moving window size and the threshold is recommended to identify parameters for these such that reliable results and converging results are obtained.
Pressure
Moving window

Identified maximum

Threshold

Time

Figure 3-1 Identification of local sloshing impact maxima

A histogram of the identified sloshing impact is to be determined. A large number of bins are defined from zero to the largest impact pressure measured. All the identified peaks are sorted in these bins in order to establish the histogram. When normalising the area under the histogram the discrete probability density function (pdf) is obtained. Integration of the PDF gives the cumulative distribution function (CDF). Given the occurrence of an impact peak, the CDF gives the probability that this peak is lower than a given pressure value. Since the main interest is in the extreme values the CDF is preferably presented in the form of the Exceedance Probability Function (EPF) on a logarithmic scale. Figure 3-2 illustrates the various probability functions. For further details about the theory of probability functions one is referred to handbooks on statistics and stochastics.

DET NORSKE VERITAS

18

Classification Notes - No. 30.9 June 2006

Number of impacts

Pressure PDF Normalising

mathematical functions has a fundamental relation to the physics dominating the randomness of the impacts. A mathematical fit should be plotted together with the discrete data. A mathematical fit may be easy to use for further mathematical processing. When comparing two different cases the response periods may differ and hence the EPF plots may be difficult to compare. It is then recommended to normalise the EPF by dividing the probabilities by the response period. The EPF then gives the probability that a peak exceeds a given pressure value per unit time. A return period level in this figure has the same probability level irrespective of the response period.
Expected extreme to occur in 1 h Expected extreme to occur in 3 h Expected extreme to occur in 30 hours 0.01 Sloshing impact pressure peaks

Pressure
Exceedance probability (1/s)

Partial integration CDF 1.0

0.001 1 hour return i d 0.0001 3 hours

0.00001

30 hours

Pressure Inverse Pressure

0.000001

Figure 3-3 Example EPF with 3 different return periods

0.1

0.01

0.001 EPF

Figure 3-2 Schematic procedure to determine sloshing impact peak pressure probability functions

From the peak identification process the total number of identified sloshing impact peaks is obtained. The total time duration in which these peaks occurred is known as well, hence the average time between successive sloshing impact peaks can be calculated. This average time is often called the response period. The response period is to be used to determine probability levels in the EPF plot for specified return period(s), since for large N the probability of exceedance is given as 1/N. Using the mean encounter wave period or the mean encounter period of a typical ship motion mode as response period can give erroneous results. For example in some conditions only the severest wave or motion sequences may give impact recordings, which leads to only a small number of identified sloshing impact peaks per time, which implies a much larger response period than for example the mean wave period. Expected extremes to occur in a specified return period may be estimated when defining return periods in the EPF plots. An example plot is given in Figure 3-3. Fitting of the discrete statistical data may be done using a mathematical function, like the Weibull or the Pareto distribution. It should be remembered however that none of those

When conducting a series of irregular sloshing experiments for an identical condition, the probability functions per test run may be obtained. These can be combined using statistical theory to determine the probability functions for the entire test duration, which is the sum of the individual runs. An accurate estimation of the expected extremes for large return periods require that sloshing tests are conducted for sufficiently long duration. A test duration of X hours is normally not sufficient to estimate the expected extreme to occur in X hours. But this is highly dependent on the uncertainty in the extreme value data. Figure 3-4 shows two example EPF plots for different cases. As seen the data is believed to be sufficiently accurate to estimate the expected extremes in 1 and 3 hours, but for a 30 hours period the slopes of the two curves are very much different and by that the uncertainty in the extreme value estimates.
0.01 Sloshing impact pressure peaks Reference case Exceedance probability (1/s) 0.001 Target case 1 hour return period 0.0001 3 hours

0.00001

30 hours

0.000001

Figure 3-4 Two example EPFs illustrating the variation in extreme value uncertainty

The statistical post-processing procedures as outlined in this

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

19

paragraph applies to the measurements of single pressure sensor but is likewise to be applied to the pressure histories resulting from a summation/integration of a cluster of pressure sensors. The average pressure history on an area may be determined by a (weighed) sum of the individual pressure sensor time histories positioned in the specified area. It is to be noted that the application of this pressure history in a structural response dynamic analysis may give different results than a direct application of the individual pressure sensor time histories. After conducting a statistical post-processing of single sensors and clusters of sensors the expected extreme impact pressures are to be plotted as a function of load area. Figure 3-5 gives an example.
Expected extreme impact pressure

Pressure
Peak value

Impulse=? p(t)dt

Threshold

Time
Rise time

Figure 3-6 Characterisation of sloshing impact peak


Reference case

Pressure
Target case

Peak value

0.2

0.4

0.6

0.8 1 Load area (m2)

1.2

1.4

1.6

Peak value

Figure 3-5 Example curves of expected extreme impact pressures as a function of load area

From the pressure sensor time histories the characteristics of the single impacts can be determined. Figure 3-6 shows the main characteristics, which are to be determined. The rise-time is the time from the start of the impact to the time of the maximum peak. For some impacts this definition is not straightforward enough, as the slope of the impact may vary significantly. In order to avoid such cases it may be more appropriate to define the rise time as twice the time from half the peak value to the peak value, see Figure 3-7. The impulse is defined as the pressure integration of the impact. It is recommended to plot the sloshing impact peak values and the associated rise times for both the target and reference case in one figure. See an example in Figure 3-8. Such a figure gives information on the possible differences in rise times between the reference and target case, which might have an effect on the structural responses. The calculated rise-time should be verified for the 10 largest impacts for the design cases by a visual inspection of the pressure time traces.

Time
rise time

Figure 3-7 Alternative definition of rise time

Figure 3-8 Comparison of sloshing impact peak values and associated rise times for a reference and a target case. The impact pressure peak values are normalized by the expected extreme impact pressure

3.4 Sloshing design loads


Based on the set of ship and environmental conditions a sloshing test experimental programme is composed. The testing is divided into a screening phase and a design phase. The screenDET NORSKE VERITAS

20

Classification Notes - No. 30.9 June 2006

ing phase is to evaluate all the possible combinations of filling levels, wave headings, etcetera to identify the sea state giving the worst sloshing impact loads. The screening process is not straightforward. The number of test cases needs to be limited, and the selection of critical cases is based on a slowly converging statistical estimate of the impact load indicator This indicator can be selected in different ways. One option is to look at the characteristic impact pressure from each sensor separately. This is consistent with how the design loads are found. The obvious drawback of this approach is that a single test with a duration of 5 hours full-scale gives a large uncertainty of the expected extreme single-sensor pressure, and thus is a weak basis for critical case selection. An alternative is to combine the single sensor measurements, i.e. the event maximum from each sensor cluster is used a basis for the statistical expected extreme estimate. Only the maximum peak pressure recorded among the sensors in the cluster is counted each time an impact occurs. This approach implicitly assumes that the position of the high pressure is of secondary importance. The benefit is a much larger number of recorded peak pressures which gives a smaller variability in the expected extreme pressure. A third alternative is to look at the average pressure over a larger area, for instance a 2 2 grid of sensors in a corner hot-spot. Ideally, both single-sensor and larger area based expected extreme pressures should be found during screening. If the trends vary with area, the critical case selection should reflect any previous experience regarding critical loading area size. Based on the results from the screening phase the design condition(s) are determined. These design conditions are tested for a long duration in order to obtain sufficient data for proper statistical analysis of the sloshing impact loads.
Significant wave height
Screening tests

Significant wave height

Design case Additional design cases Long-term contour of Hs and Tz

Wave period

Figure 3-10 Identified sloshing design cases

In 1.7.8.2 load factors are discussed, where F2 is introduced to reflect the difference in the statistical distribution between the reference and the target case. In Figure 3-4 the expected extremes for the reference and target case are nearly similar, however the uncertainty in both is significantly different, hence a load factor larger than 1.0 should be applied in this case. In order to determine a proper load factor, F2, a number of aspects needs to be considered: 1) It should be investigated if the likelihood that the reference vessel will encounter the design sea state is equal to the likelihood that the target case encounters its design sea state. Lets consider for example that the reference case in Figure 3-4 represents a 95% H filling case in a head waves situation, whereas the target case represents a 25% H filling case in beam sea condition for a weather-vaning offshore discharging vessel. It can then be argued that the likelihood for the target case meeting this particular case is lower than the likelihood for the reference case. 2) It should be investigated if the sloshing loads from the short-term design sea state assessment for the reference and target case are likewise long-term representative. It is recommended to determine as well the sloshing design impacts for identical ship and sea state conditions but with lower sea state severity, as illustrated in Figure 3-10. Alternatively a screening procedure is redone for a lower contour, resulting in possibly a different design sea state. This testing can give information on the expected extreme impact pressures as a function of sea states severity for the reference and the target case. Combining this information with the sea state probabilities it can be investigated if the ratio between the short-term extreme values can be assumed to be lifetime representative. How well the sloshing load estimate based on a short-term assessment represents the sought long-term value will be illustrated in the following. Nonlinear responses like sloshing induced impact pressures may be characterised by response amplitude distributions having flat tails (i.e. Weibull fitted function with low values for the slope factor). The figure below shows some example curves with varying slope parameters. For linear responses the amplitude distribution is given by the Rayleigh distribution.

Long-term contour of Hs and Tz

Wave period

Figure 3-9 Sloshing screening procedure to identify worst sea state

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

21

Amplitude 0 1.0E+00 1.0E-01 1.0E-02 1.0E-03 1.0E-04 1.0E-05 1.0E-06 1.0E-07 1.0E-08 Rayleigh (slope=2.0) Weibull (slope=1.4) Weibull (slope=0.8) 2 4 6 8 10 12 14 16 18

Exceedance probability

for lower significant wave height with identical slopes and with scale factors linearly scaled by the significant wave height. (Note: expected extremes are linearly related to the scale factor for Weibull distributed responses). These Weibull functions are modelled for the selected sea states of Table 3-3. The resulting long-term exceedance probability curve is then determined by the weighted sum of these individual Weibull functions according to:

Qlong term ( p a ) = w H s , j Q st p a H s , j p H s , j
j =1

) (

where:

Q( p a ) is the exceedance probability function of sloshing impact peak values wH s, j = T TH s , j

Figure 3-11 Example figure showing effect of the Weibull slope factor

When the response is characterised by more flat tails the difference in determining a short-term or a long-term extreme is larger than 10% to 20%. Typically the sloshing impact pressure amplitude distribution can be modelled by a Weibull function with a slope less than 1.0. Consequently, by testing only the worst sea state the short-term expected extreme is not representative for the lifetime expected extreme but significantly lower. The example below describes this character in more detail.
Note: Consider the North Atlantic wave scatter diagram as given by IACS Recommendation No. 34, see /4/. Assume that sloshing screening tests have been done along the 40 years contour of Hs and Tz. The worst sea state is found to be given by Tz = 8.5 s and Hs = 12.8 m (as an example). It is assumed that the expected short-term extreme is 15 bar and that the impact pressure amplitude exceedance probability curve can be modelled by a 2-parameter Weibull function Q(pa) with a scale factor = 0.566 and a slope factor = 0.6, which typically can be observed from sloshing experiments.

is the weighing factor to account for relative number of cycles (= 1 since all response periods are taken equal to 8.5 seconds) T is the average period between impacts (long-term) TH s , j = average period between impacts for sea state j p(H s,j) = short-term sea state probability Here it is assumed that the number of impacts per time is equal for all sea states and is taken equal to the wave period, i.e. 8.5 seconds. For the 40 years storm condition a return period of 3 hours results in an exceedance probability level of 7.87 E-4. The sum of observation from Table 3-3 is 3 165 for a 40 years period. This corresponds then to a return period of 13 months, which results in an exceedance probability level of 2.49 E-7. Figure 3-12 shows the short-term exceedance probability curve for the 40 years storm condition (Hs = 12.8 m) and the long-term exceedance probability curve as the weighted sum of the selection of sea states. When deriving the expected extremes it is seen that the long-term value is almost twice as large as the short-term value (1.93). From this example it is believed that the testing of only the worst 40 years storm condition and determining the expected 3 hours extreme is a significant underprediction of the expected lifetime extreme.
Pressure (bar) 0 1.0E+00
Short-term curve - 40 years sea state

Q( p a ) =

p a e

5 10 15

20

25

30

35

40

Exceedance probability

From the IACS Recommendation No. 34 scatter diagram a selection of sea states is taken, see Table 3-3. (NOTE: The number of observations is adjusted in order to match a 40 years duration, i.e. 100 000 observations corresponds to 34 years) Table 3-3 Sea state selection from IACS rec.34 Hs Number of observations during 40 years (Tz = 8.5 s) 5.5 1 872.2 6.5 806.3 7.5 315.5 8.5 114.3 9.5 38.8 10.5 12.5 11.5 3.9 12.5 1.2 SUM 3 164.6 From sloshing tests it is observed that expected short-term extremes vary more-or-less linearly with the significant wave height. Secondly, the slope parameter varies little for varying significant wave heights. Hence Weibull functions can be modelled

1.0E-01 1.0E-02 1.0E-03 1.0E-04 1.0E-05 1.0E-06 1.0E-07 1.0E-08

3 hours return period

15 bar

'Long-term' curve 13 months return period

29 bar

Figure 3-12 Short-term and long-term exceedance probability curves (Weibull slope = 0.6) For completeness an additional example is shown, where the Weibull function is modelled with a slope factor = 2.0, which corresponds to the Rayleigh distribution as usually applied for linear ship responses. Figure 3-13 shows the results from which it is seen that the long-term extreme is only slightly larger than the short-term extreme (1.04).

DET NORSKE VERITAS

22

Classification Notes - No. 30.9 June 2006

Pressure (bar) 0 1.0E+00


Short-term curve - 40 years sea state

10

15

20

25

30

1.0E-01 Exceedance probability 1.0E-02 1.0E-03 1.0E-04 1.0E-05 1.0E-06 1.0E-07 1.0E-08

3 hours return period 'Long-term' curve 13 months return period

full load condition ballast condition part load condition, with all tanks equally filled part load condition, with one tank partly filled and other tanks empty.

15 bar

15.6 b

Figure 3-13 Short-term and long-term exceedance probability curves (Weibull slope = 2.0)

---e-n-d---of---N-o-t-e---

4. Pump Tower Design Loads


4.1 Identification of relevant loads
The loads relevant for the dimensioning of the pump tower are due to the ship motion, the motion of the LNG in the cargo tanks, and the temperature of the cargo. Loads that may need to be included in the strength assessment are hence: sloshing loads gravity and inertia loads thermal loads hull girder loads (for liquid dome area and for base support) internal tank pressure and external sea pressure (for base support). A description of each load type is given in the following sections.

4.2 Sloshing loads


Sloshing loads on the pump tower structure occur due to motion of the liquid inside the cargo tanks. Usually, tank no. 2 is considered to be critical for sloshing, i.e. having the maximum sloshing motion, due to size and distance from the ships center of rotation. However, both tank no. 1 and tank no. 2 need to be considered, to determine which one is most critical with respect to the combined effect of sloshing loads and inertia loads. The sloshing loads are to be determined using model testing or numerical analyses by computational fluid dynamics (CFD). Irregular motions are to be considered. 4.2.1 Load conditions and filling levels The sloshing loads should be considered for various filling levels according to the ships filling restrictions. For normal filling restrictions (less than 10% L, above 70% H), the following filling levels should be considered: 10% L 70% H, 80% H, 90% H and 95% H. For unrestricted filling, the following intermediate filling levels should also be considered: 20% H, 30% H, 40% H, 50% H, 60% H. The ship loading conditions that are considered to be critical for sloshing motion in the tanks must be determined. Loading conditions that may need to be considered are:

4.2.2 Environmental conditions The speed of the vessel is to be reduced to design speed. A more refined speed-sea-state curve may be used, as depicted in Figure 2-3. If a comparative approach is used, there are two aspects which are of importance to consider when applying a stepwise speed-sea-state curve. Firstly, the speed can have an important effect on the sloshing loads. In standard wave load analysis a large ship speed gives conservative design motions and loads in an absolute strength assessment. Secondly, the sloshing procedure identifies the worst sea state on a contour of significant wave heights and periods. Applying this procedure for the reference and the target case may lead to a situation where the ship speed for reference case is higher than for the target case, which for some cases is non-conservative. For calculation of FLS loads, an operational profile through the lifetime of the ship must be assumed. Critical wave headings must be determined. In principle, all headings 0 to 180 should be considered for all filling levels. The sea states that are critical for sloshing motion in the tanks must be determined. A scatter diagram is used to describe the probability of occurrence of short-term sea states. The North Atlantic scatter diagram should be used, as given in /3/. The short-term sea states are described by a wave spectrum. The Pierson-Moskowitz spectrum may be used, in combination with a cosine squared wave spreading function. For the relevant loading conditions, ship motion analyses are to be carried out. The analyses are to provide linear ship motion transfer functions, which can be used to generate ship motions for specified sea states. The tank motions are then calculated by accounting for the distance from the ships rotational center to the center of the tank. For the sea states assumed to be critical with respect to sloshing motion, irregular time series of motion are to be calculated using the ship motion transfer functions. The sloshing loads on the pump tower frame structure due to the irregular tank motions are then determined, using experiments or analyses. Expected extreme loads are to be determined for a duration of not less than 3 hours. 4.2.3 Load prediction Based on the set of ship and environmental conditions a sloshing test experimental or calculation programme is composed. The testing is divided into a screening phase and a design phase. The screening phase is to evaluate all the relevant combinations of loading condition, filling level and wave heading to identify the sea state giving the worst sloshing loads. Based on the results from the screening phase the design condition(s) are determined. These design conditions are tested or calculated for a long duration in order to obtain sufficient data for proper statistical analysis of the sloshing loads. The procedure described in 3.3.4 is recommended for statistical post-processing of the load time histories. The 25 years contour of significant wave height and period of this scatter diagram may be used to identify the worst conditions for the sloshing motion. For beam sea conditions, the 25 years contour may be reduced by 20% to reflect voluntary heading changes in severe beam seas. The sloshing load for the ULS assessment should be taken as the load occurring once during the lifetime of the ship, as determined from the long-term load distribution. The actual longterm distribution of sloshing loads should in principle be determined as a weighted sum over all the sea states given in the scatter diagram and all headings considered, as explained in 4.2.3.1. The procedure may also be found in the DNV Classi-

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

23

fication Note 30.7 /6/. If no evaluation of the long term distribution is made, the ULS load may be determined by increasing the load calculated for the critical short-term sea state by a factor of 1.15:

of the tower, i.e. along the z-axis indicated in Figure 4-1.


Discharge pipe y

p long term = 1.15 p short term


Filling pipe

The short term values are based on a 25 years contour. The factor may be used in case of normal filling operations. For evaluation of unrestricted filling, special considerations are to be made of the long term distribution.

Transverse bulkhead Float level gauge Emergency pipe

4.2.3.1 Establish long term distribution In order to establish the long-term distribution of sloshing loads, the cumulative distribution may be estimated by a weighted sum over the sea states used for estimating the sloshing load. The long-term stress range distribution is then calculated from,

Discharge pipe

Q ( ) =
where: pij rij=ij/0

seastates headings

r
i =1 j =1

ij

Q ij ( ) pij

Figure 4-1 Illustration of pump tower lay-out and coordinate system

is the probability of occurrence of a given sea state i combined with a heading j. is the ratio between the response crossing rates in a given sea state and the average crossing rate.

The fluid forces on the pump tower structure may be assumed to be drag dominated, so that the fluid acceleration is of secondary importance relative to the fluid velocity. The velocities and accelerations may be determined by one of the following methods: experimental tests analysis with computational fluid dynamics (CFD). The forces acting on the actual pump tower can then be calculated using Morisons equation, as explained in 4.2.4.3.

0 =

seastates headings i =1 j =1

ij

ij

is the average crossing rate.

1 ij = 2
Q()ij

m 2ij m0ij

is the response zero-crossing rate in sea state i and heading j.

4.2.4.1 Fluid velocities and accelerations based on experiments The experiments should facilitate velocity estimates at least ten points along the vertical axis.
A direct flow velocity measurement can be achieved by use of Particle Image Velocimetry (PIV). The fluid is seeded with tiny reflective particles with density close to that of the fluid. The particles are assumed to follow the flow, and they are illuminated by e.g. a laser sheet. One or more cameras are used to capture two frames within a short instant. With two cameras in a stereoscopic setup, particle displacements along all three axes can be found. The velocity field in the imaged part of the laser sheet is typically found based on particle displacements from a cross-correlation analysis and the time separation between the images. The challenge of light reflection from the free surface can be partly overcome by use of fluorescent particles. PIV can also be used to estimate fluid accelerations. An alternative to a direct velocity measurement method is given in the following example. The setup is illustrated in Figure 4-2. The bending moment in a pipe located in the tank is measured with strain gauges positioned to measure vertical strain. Force transducers should be fitted at top and bottom, in order to measure the total reaction forces on the pipe. By fitting for instance a spline function to the measured bending moments, the shear force is found as the derivative of the moment. By requiring force balance for a pipe segment, the difference in shear force at the end of the segment equals the external force on the segment. This external force contains gravity, inertia, buoyancy and hydrodynamic components. The inertia effect may be found by measuring the acceleration by an accelerometer. The gravity and buoyancy must also be accounted for. Then the fluid force on the segment is obtained. For estimation of the fluid velocities and accelerations, Morisons equation can be used. Again, drag may be assumed to dominate the fluid forces on the pump tower structure.

is the Weibull distribution from the experiment or CFD calculation in sea state i combined with heading j. For the case of unrestricted filling, special consideration of the long term value of the load should be made. The FLS assessment is to be carried out using the same longterm load distribution as applied for the ULS assessment.

4.2.3.2 Simplified long term distribution Alternatively, the long-term distribution for FLS may for simplicity be assumed as a Weibull distribution:

h Q ( ) = exp q
where Q() is the probability of exceedance of the stress , h is the shape parameter, and q is the scale parameter, defined as

q=

0
(ln n 0 ) 1 / h

where 0 is the reference stress value exceeded once out of n0 cycles. The shape factor should be taken as h = 1.0. 4.2.4 Sloshing experiments and analysis The sloshing loads on the pump tower resulting from the applied tank motion are a function of the liquid velocities and accelerations. The velocities and if possible accelerations should be determined at several vertical locations along the midpoint

DET NORSKE VERITAS

24

Classification Notes - No. 30.9 June 2006

The post-processing described in 3.3.4 should be carried out on each of these time series, in order to determine design bending moment and reaction forces. From these, the critical time instants for top and bottom support and main structure may be determined. The velocity fields at each of these time instants should be selected for the structural analysis, and the sloshing loads acting on all structural members at these instances should be applied to the structural model.
Pumptower location Tank roof Chamfer

Figure 4-2 Example of test setup for pump tower loads

Monitoring locations for velocity and acceleration

4.2.4.2 Fluid velocities and accelerations from CFD If analyses are carried out using CFD, the following requirements should be fulfilled:
The adequacy of the program used is to be documented, especially for low filling heights where breaking waves are expected. The adequacy is strongly related to the treatment of the free surface. To ensure that the software is capable of describing the physical sloshing phenomenon, the CFD software should be verified with model tests. The program should be capable of handling irregular tank motions and long simulation length. Requirements related to modelling, mesh, and time step are to be given careful consideration. The requirements are important for numerical stability and an adequate discretisation of the problem which can provide a physical solution. For calculations with an Eulerian grid, the discretisation is recommended to be at least 40 elements in lengthwise direction, 40 elements along the tank breadth and 30 elements in the vertical direction. Velocities and accelerations in the liquid at the location of the pump tower is to be determined. The velocities and accelerations may be calculated for a vertical axis located on the mean distance between the discharge pipes and the emergency pipe. The loads will vary in the vertical direction, but for each vertical position the velocities and accelerations may be taken as equal for all the pipes. The number of positions should be sufficient to describe the flow field along the pump tower length accurately. Typically more than 10 positions should be used.

Keel

Figure 4-3 Monitoring locations for velocities and accelerations

The loads on the pump tower segment are then calculated using Morisons equation, as described in /3/. The force F acting on a part of a member is

F = L (1 + C m )Va +

1 LCDv | v | A 2

where L is liquid density, Cm is added mass coefficient, V is volume of liquid displaced by the part, a is particle acceleration normal to the member axis, CD is drag coefficient, v is liquid particle velocity normal to the member axis, and A is area of the part projected onto a plane normal to force direction. Values of CD and Cm should be determined for each structural member according to recommendations given in /3/. For a cylindrical tube Cm is to be taken as 1.0. It should be noted that for oscillatory motions, CD is a function of the KC-number,

KC = U M T / D
where: D = diameter of pipe T = sloshing wave period UM = maximum sloshing velocity The drag coefficient may be taken from Figure 4-4, /3/.

4.2.4.3 Load calculation from fluid velocity and acceleration Based on the time series of the velocity field, the time instant of maximum bending moment and upper and lower reaction force are determined. Maximum bending moment and reaction force should be considered both for the longitudinal direction and the transverse direction.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

25

spectrum. The Pierson-Moskowitz spectrum may be used. A cosine squared wave spreading function may be assumed. Critical wave headings must be determined. All headings between 0 and 180 should be considered for all relevant loading conditions. The long-term distribution of inertia and gravity loads should be determined as a weighted sum over all sea states given in the scatter diagram and all headings, as explained in DNV Classification Note 30.7 /6/. The loads to be used for the ULS assessment are to be taken as the load occurring once during the lifetime of the ship. The FLS assessment is to be carried out using the actual long-term load distribution. For the relevant loading conditions, ship motion analyses are to be carried out. The analyses are to provide linear ship motion transfer functions, which can be used to generate ship motions for specified sea states. The inertia and gravity loads may be calculated for a vertical axis located on the mean distance between the discharge pipes and the emergency pipe. The loads will vary in the vertical direction, but for each vertical position the accelerations may be taken as equal for all pipes.

Figure 4-4 Drag coefficient CD as function of KC for cylinders in waves, Re > 5.0 105

4.4 Thermal loads


Thermal loads are due to thermal shrinkage of the pump tower material in the low temperature condition, relative to the room temperature condition. The temperature effect is most important for the upper and lower supports of the pump tower. For each filling level considered, the temperature distribution over the height of the pump tower is to be determined. For filling levels above 70%H, the temperature can be taken as constant and equal to -163C. For filling levels below 10% L, the temperature of the submerged part of the pump tower can be taken as -163C, while the temperature of the non-submerged part can be assumed to vary linearly from -163C at the liquid surface to -30C at the top of the pump tower. The initial temperature of the steel can be taken as 20C. Thermal expansion coefficients relevant for the pump tower material should be applied. Stainless steel is usually applied. Material properties for the stainless steel 304L are given in 8.2. The temperature field should be applied to the FE model used for the response analysis, in order to determine the stress field in the structure resulting from the thermal shrinkage. Low-cycle fatigue due to cyclic variation of the thermal stresses between the empty and loaded condition may need to be included in the fatigue life calculations. Thermal stresses acting in the base support due to the thermal gradient from the level of the primary membrane to the level of the inner bottom plating should also be calculated.

For cylinders that are close together, such as the emergency pipe and the filling pipe, group effects may be taken into account when determining the drag coefficients. If no documentation of the group effect is available, the drag coefficients for the individual cylinders should be used. When the load on each segment is determined, a beam model may be used to determine top and bottom support forces and bending moments. The boundary condition of this model should reflect the actual boundary conditions of the pump tower.

4.3 Inertia and gravity loads


Inertia loads on the pump tower are due to the accelerations of the vessel. Translational tank accelerations due to the pitch and roll accelerations of the ship is to be accounted for. Weight of liquid inside the pipes should be included in the calculation of inertia loads. Gravity loads are due to the self-weight of the pump tower, and due to the roll and pitch motion of the ship. When the gravity loads are calculated, the buoyancy of the pipes should be deducted. For inertia and gravity loads, the pump tower in tank no. 1 is usually the most critical, due to its distance from the ship center of motion. However, both tank no. 1 and tank no. 2 need to be considered, in order to determine which one is most critical with respect to the combined effect of sloshing loads and inertia loads. The combined effect of sloshing and inertia load is to be summed as a total load. For both inertia load and gravity load calculation, the weight of additional elements (structural members and equipment) that are not included in the finite element model used for the response analysis should be included as lumped or distributed masses. The ship loading conditions that are considered to be critical with respect to ship accelerations and motions must be determined. The loading conditions that need to be considered are given in 4.2.1. For calculation of FLS loads, an operational profile through the lifetime of the ship must be assumed. The sea states that are considered to be critical for ship accelerations and motions must be determined. A scatterdiagram is used to describe the probability of occurrence of short-term sea states. The North Atlantic scatter diagram should be used, as given in /3/. The short-term sea states are described by a wave

4.5 Hull girder loads


The global bending moment acting on the ship hull girder causes a stress field in the liquid dome area. The stress field resulting from the still water and wave bending moment should be determined, and applied to the FE model used for analysis of the liquid dome area. The maximum bending moment acting along the ship length should be considered. The global stress should include the stress concentration due to the liquid dome opening. Similarly, the longitudinal stress acting in the bottom due to the global bending moment should be determined, and applied to the FE model used for analysis of the base support.

4.6 Internal tank pressure and external sea pressure


The double bottom stress resulting from the internal tank pressure and external sea pressure acting on the double bottom should be determined, and applied to the FE model used for analysis of the base support.

DET NORSKE VERITAS

26

Classification Notes - No. 30.9 June 2006

4.7 Combination of loads


For ULS assessment, the worst possible combination of sloshing load and inertia load on the pump tower should be determined. In principle, this can be done by investigating a time series of irregular tank motion, and monitoring the combined effect of sloshing load and inertia load. By post-processing the combined load effect, the maximum load may be determined. Alternatively, a conservative approach is to add together the maximum values of sloshing loads and inertia loads:

Ftower = Fslosh + Finertia


When the total pump tower loads are combined with hull girder loads and pressure loads for analysis of the pump tower supports, quadratic summation may be applied for the force components:

A simplified approach including the use of a dynamic load factor along with quasi-static response analyses should be used to assess the dynamic responses of the membrane type insulation systems. The requirements and guidelines for the quasi-static response analyses are described in 5.2 for the Mark III system and 5.3 for the NO96 system. The dynamic factor is defined in 5.6. Analysis should be carried out for transverse corner/longitudinal knuckle insulation structures and for the flat wall insulation structure adjacent to corners and knuckles. Note that the response analysis methodology specified for the Mark III system in 5.2 is also applicable to the CS 1 system, but that the comparative strength assessment methodology can not be applied to this system.

Fx ,dyn = Fx2,tower + Fx2,hull + Fx2, press


The static loads due to gravity and thermal gradients should be added:

5.2 Mark III system


5.2.1 General A schematic representation of the quasi-static load-displacement response of a Mark III insulation panel is shown in Figure 5-1. The elastic response limit of the structure is reached at Point 1 on the curve, where the elastic capacity of the reinforced polyurethane foam (RPUF) is reached at the mastic support locations as illustrated in the upper right hand sketch on Figure 5-1. This is not considered to represent the capacity of the structure. Loading past Point 1 on the response curve requires that the incremental load is transferred to the mastic supports as shear force in the plate, and given sufficient strength of the bottom plywood plate, the stress in the foam will reach a distribution as shown in the lower right hand sketch of Figure 5-1. Loading beyond this limit will lead to excessive deformation of the panel, and it consequently represents an upper bound on the capacity of the insulation panel.

Fstatic = Fgravity + Fthermal


The total load is found by adding the static loads to the total dynamic load:

Ftotal = Fdyn + Fstatic


The summation of forces should be carried out componentwise, i.e.

Fx ,total = Fx ,dyn + Fx , static Fy ,total = Fy ,dyn + Fy , static


Fz ,total = Fz ,dyn + Fz , static
It should be noted that both positive and negative values of the dynamic loads must be considered when the total load is calculated, i.e. Finertia, Fhull, and Fpress. For the yield check, the maximum stress is of interest, while for the buckling check the maximum compression stress is of interest. For FLS assessment, four long-term load distributions should be determined; one for the full load condition, one for the ballast condition, one for the part load condition with all tanks equally filled, and one for the part load condition with one tank partly filled and the other tanks empty. The fatigue damage should be calculated for each condition, and the total damage is found by addition of the four contributions. An operational profile for the ship must be assumed. In principle, the long-term load distributions should be determined for the total load, consisting of all load effects. For simplicity, however, the sloshing loads and inertia loads may be combined by assuming that the number of occurrences are approximately the same for both contributions. The long-term stress distribution may then be determined for the dominating load, i.e. either the inertia or sloshing load, and the reference stress level is found by adding the stress distribution due to the sloshing load and the inertia load. For ULS assessment and FLS assessment, both tank no. 1 and tank no. 2 should be considered, in order to determine which one is most critical with respect to the combined effect of sloshing loads and inertia loads.

Load

Displacement

5. Structural Response Analysis of Containment Systems


5.1 General
The strength assessment of structures exposed to transient dynamic loads such as those generated by sloshing impacts generally requires the assessment of the dynamic response of the structure.
Figure 5-1 Schematic illustration of the load-displacement response and stress states in the polyurethane foam at the foam/plywood interface of a Mark III/ insulation panel

The actual capacity may be governed either by excessive de-

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

27

formation of the insulation panel at the load level identified at Point 2, or shear or bending failure of the bottom plywood plate at an intermediate load level between Point 1 and Point 2. It is clear that an assessment of the bottom plate capacity requires a determination of its structural response in the non-linear response range of the insulation panel. The non-linear response assessment should be carried out using linear elastic finite element response analysis to determine the elastic stress and deformation state at Point 1 (5.2.2), in combination with a simplified analytical model based on linear elastic beam theory to assess the incremental shear force and bending moment in the bottom plywood plate (5.2.3 and 5.2.4). 5.2.2 Finite element analyses The finite element response analyses should be carried out using a finite element model covering a sufficiently large portion of the insulation panel to make sure that the structural response is well confined within the interior of the model. The model need not reflect the full length and width of the insulation panel. Symmetry in geometry and/or structural response of the insulation panel may be exploited. For the corner/knuckle panels it is sufficient to model only one side of the corner/knuckle as illustrated in Figure 5-2. The model should include a sufficiently large portion of the adjacent flat insulation panel to avoid significant effects of boundary conditions for the considered load areas.

The model should be built up using a mix of continuum, shell, and membrane elements, as follows: The polyurethane foam, the mastic strips, and the hardwood key of the corner/knuckle panels should be modelled using eight node 3D continuum elements. Elements using a reduced integration scheme may be used. The upper and lower plywood plates as well as the secondary barrier level plywood plate of the corner/knuckle panel can be modelled using four-node shell elements. The primary corrugated steel membrane may be modelled using membrane elements with small in-plane stiffness, but with a representative mass per unit area to include its inertia effects on the dynamics of the panel. The secondary triplex membrane may be disregarded. The shell and the continuum components of the model should be constrained to enforce kinematic continuity between the components, taking into account the thickness of the plates. The membrane elements representing the primary membrane should be constrained to follow the deformation of the 12 mm top plywood plate. The finite element mesh requirements are given in the following. Mesh densities deviating from these requirements will be accepted provided that the adequacy of the mesh can be demonstrated with a mesh convergence test. In regions of the model where the structural response will be extracted: 3 elements should be used across the width of the mastic supports at least 7 elements should be used across the free span of the bottom plywood plate the element edge length in the direction of the mastic supports and in the through thickness direction of the panel should be determined so that the elements are practically square. In other parts of the model:

Figure 5-2 Required model extent for corner panel

An example finite element model comprising half the width of a flat wall Mark III panel (495 mm) and approximately a quarter length of a panel is shown in Figure 5-3.

2 elements should be used across the width of the mastic supports at least 5 elements should be used across the free span of the bottom plywood plate the element edge length in the direction of the mastic supports and in the through thickness direction of the panel can be rectangular, with a maximum aspect ratio of 3.0. In the case the three planes of symmetry in geometry and response of the panel is exploited, the boundary conditions for the panel should be taken as follows: restrained translations of the bottom nodes of the mastic support symmetry conditions on secondary foam, bottom plywood plate and centre mastic at panel symmetry planes (see Figure 5-4) symmetry conditions on primary foam and top plywood plate at cross-panel boundaries (see Figure 5-4). The far end plane of the model may be kept free. All patch loads should be applied away from this boundary, and even in the case of uniform load the results are taken at the opposite end of the model.

Figure 5-3 Example finite element model of a Mark III panel

DET NORSKE VERITAS

28

Classification Notes - No. 30.9 June 2006

1) For a reference patch load, establish the elastic through thickness stress on top of the mastic support located as far as possible directly below the centre of the loaded area (see Figure 5-6). If the load patch covers more than one mastic support, the average of the stress at all the mastic supports may be used. Let this be 0p . 2) Establish the elastic through thickness stress on top of the same mastic support(s) for the case of a uniform reference load with the same magnitude as above (see Figure 5-6). Let this be 0u . 3) Calculate the ratio r =

4) Calculate the average through thickness stress as = r p

0p 0u

Figure 5-4 Symmetry planes for the example Mark III insulation panel model

The plywood and the reinforced polyurethane foam materials may be modelled as homogeneous linear orthotropic elastic materials. In this case the bending stiffness moduli should be used to represent the plywood stiffness. Alternatively the plywood may be modelled as a 0-90 laminate of layers of orthotropic elastic materials. The mastic may be modelled as a homogeneous linear elastic material. Material properties are given in 5.4, and should be selected based on the material directions specified in Figure 5-5.

u 0

0p
Figure 5-6 Definition of the stress quantities used to define the average through thickness stress

2 1

Figure 5-5 Definition of RPUF and plywood material directions relative to the Mark III insulation panel

The procedure outlined above implies that the average compressive stress is defined based on the linear elastic through thickness load dispersion in the model. This is conservative, since inelastic deformations below the loaded region of the panel will result in a change in the relative stiffness between this area and the adjacent areas, and thus also increase the load distribution to adjacent areas. Another source of conservatism in this calculation is the selection of the peak elastic stress in the patch load case to enter the calculation of the ratio r. 5.2.4 Shear force and bending moment in bottom plate In the case of simple load situations where the load on the panel surface extends over an area spanning several mastic support spacings, the structural system for assessment of the incremental shear force at the supports is statically determinate. The additional force applied to the system beyond the elastic response limit (Point 1 in Figure 5-1) can not be transferred through the foam into the supports because the elastic limit of the foam has been reached at the mastic support locations. The entire force must hence be transferred to the supports as shear force in the plywood plate.

5.2.3 Average through thickness stress in foam The average through thickness stress may for a patch load case be interpreted as the equivalent uniform stress applied to the surface of the panel. It follows that for a uniform load situation the stress is directly defined by the applied load, p:

=p
The average through thickness stress for the secondary foam insulation at the interface with the bottom plywood plate should be assessed as follows:

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

29

el ,1 =

el , 2

F ref foam Mc = ref M1


Qc Q1ref

el ,3 =

el = min(el ,i )

Mel(x), Qel(x) Mel-pl(x), Qel-pl(x)

Mpl(x), Qpl(x)

Figure 5-7 Illustration of the concept of combining linear finite element response analyses for the linear elastic response range and simplified analytical models for the elasto-plastic load range of the insulation panel

An assessment of the bending moment in the plate is more complicated, and requires that an assumption is made with respect to the distribution of incremental vertical stress exerted by the foam on the plywood plate. The incremental bending moment will therefore be a more uncertain quantity than the incremental shear force. For patch load situations the assessment of the shear force will be approximate, since the procedure will not account for the redistribution of forces that will take place in the structure as soon as the stiffness locally beneath the load decrease as a result of inelastic material response. The approach outlined above is illustrated in Figure 5-7, where the selected parabolic additional foam stress function qpl(x) is also specified. The amplitude q 0 is a function of the magnipl

tude of the incremental load, as outlined in more detail in the following. The response assessment requires the execution of the following analysis steps: 1) Linear elastic structural response analysis using a reference load amplitude, pref. 2) Extract the critical response parameters from the model. This involves (see Figure 5-7): a) The vertical stress in the reinforced polyurethane foam at the center of the most highly loaded mastic support, ref . foam

DET NORSKE VERITAS

30

Classification Notes - No. 30.9 June 2006

b) The bending moment per unit plate width of the bottom plywood plate at mastic supports and mid span, M 1ref , M 2ref , M 3ref . c) The shear force per unit plate width of the bottom plywood plate at mastic supports, Q1ref , Q2ref . d) The rotations of the bottom plywood plate at the mastic supports, 1ref , 2ref . The structural response should be output at locations of maximum response centred below the applied load. When the bending moment and the shear force is obtained at the integration points of the shell elements adjacent to the mastic supports the corrections shown in Figure 5-8 and Figure 5-9 should be made to make the results representative for the plate cross-section adjacent to the mastic support. Lint is the distance from the integration point of the element to the edge of the mastic.

Foam Q QFE Lint


Q = Q FE + 0.9 Foam Lint

specified in Table 6-7, and Mc and Qc is the bending moment capacity the shear force capacity of the bottom plywood plate in the relevant cross-section direction, specified in Table 6-6. If min(el,i)= el,1 the next step is to calculate the incremental shear force. If not, the capacity is governed by plywood plate failure in the linear elastic response range of the structure, and can be estimated using either the bending moment or shear force strength criterion. The remaining steps are presented under the assumption that min(el,i)= el,1. The incremental shear force and bending moment are calculated under the assumption that an incremental pressure (-el)pref is applied to the panel, where is the load proportionality factor describing the ratio between the design sloshing impact pressure including dynamic effects ( p D ) and the reference pressure used in the finite element analyses. 4) As mentioned earlier in this section, the incremental load must be transferred to the mastic support as shear force in the plywood plate. It is assumed that the shear force is equally distributed to each end of the plate, and the incremental shear force at the end of the plate is therefore:

1 ( el ) p ref Lref 2 5) The total plate end shear force per unit plate width in the non-linear response range can be estimated as: Q pl =
1 Q1 = el Q1ref + ( el ) p ref Lref 2 6) For the assessment of the incremental bending moment per unit plate width, the incremental load specified above is assumed to be distributed according to the parabolic function shown in Figure 5-7, as follows:

Figure 5-8 Correction of FE calculated plate shear force to obtain a better estimate of the plate end value

q pl ( x) = 6( el ) p ref
Foam Q M MFE Lint
0 M pl , 2 =

Lref L3

( x 2 Lx )

The bending moments for the case of (-el) = 1 at end 1, mid span and end 2 (see Figure 5-7) are:
M = M FE + 0.9 Q Lint
0 M pl ,1 =

p ref Lref L2 k 2 (6 EI + k1 L) 90(3( EI ) 2 + EIL( k1 + k 2 ) + k1 k 2 L2 ) p ref Lref L2 k1 (6 EI + k 2 L) 90(3( EI ) 2 + EIL( k1 + k 2 ) + k1 k 2 L2 )


5 p ref Lref L 32 EI

0 M pl , 3 =

p ref Lref L2 k 2 (6 EI + k 1 L) 90(3( EI ) + EIL( k 1 + k 2 ) + k 1 k 2 L )


2 2

Figure 5-9 Correction of FE calculated plate bending moment to obtain a better estimate of the plate end value

3) Determine the minimum elastic pressure capacity of the insulation panel (Point 1 in Figure 5-7) in terms of a proportionality factor el on the applied reference load. This is the minimum of the load proportionality factors el,i satisfying the following equations: a) b) c)

The elastic modulus E should represent the bending modulus of the bottom plate in the direction transverse to the mastic supports, and the area moment of inertia I should be calculated for a unit width plate strip. The support bending stiffness k1 and k2 are calculated from the plate end bending moment and rotation obtained from the linear finite element analyses, as follows:

el ,1 ref = F foam
el , 2 M 1ref = M c

k1 =

M 1ref

1ref

, k2 =

M 2ref

2ref

el ,3 Q1ref = Qc

where F is the crushing strength of the polyurethane foam in the through thickness direction of the insulation panel

The sign conventions for the bending moment are indicated in the upper left hand sketch in Figure 5-7. The total plate end and mid point bending moments per unit plate width in the non-linear response range can be estimated as: 0 M i = el M iref + ( el ) M pl ,i , i = 1, 2, 3

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

31

5.2.5 Shear force and bending moment in the secondary barrier plywood plate in corners/knuckles The plate shear force and bending moments should be output at high stress locations at the edges of the hardwood key (transition between the hardwood key and the softer primary foam). When the bending moment and the shear force is obtained at the integration points of the shell elements adjacent to the hardwood key, similar corrections as specified for the bottom plate should be made (Figure 5-8 and Figure 5-9) to make the results representative for the plate cross-section adjacent to the hardwood key. Foam should in this case be taken as the differential stress between the foam on the primary and secondary insulation side of the plate.

The primary and secondary invar membranes may be disregarded.

5.3 NO96 system


5.3.1 Finite element analyses The primary and the secondary insulation boxes should be modelled as separate entities. Both the primary box and the secondary box may be modelled as continuous structures, meaning that full translational and rotational continuity may be assumed between the bulkheads and the top and bottom plates. The discontinuous stapled connection between the plates are hence disregarded. In the case of the reinforced primary box with double cover plates, the upper and lower plates should be modelled as separate entities. The plates should be connected using elastic springs at the positions of the staples, as indicated in Figure 5-10, and the interaction between the plates should otherwise be treated as contact in the normal direction and free sliding in the tangential direction. The stiffness of the connecting springs should be taken as 10 000 N/mm. The cover plates should be modelled as three individual parts in order to account for the discontinuity of the plating at the invar tongue slits. Analysis of the corner boxes should include the adjacent flat wall boxes and the plywood plate that covers the gap between the corner structure and the flat wall structure.
Figure 5-11 Cut through the 3D model

Node

Spring

Figure 5-12 Modelling of staples in the FE model

The NO96 insulation boxes should be modelled using 4 node shell elements. The mesh requirements are as follows: Cover plates: at least 7 elements between the vertical bulkheads of the primary box the element aspect ratio for the cover plate should not exceed 2.0, and should be determined based on the bulkhead plate mesh requirements given below. Primary box bulkheads: the element mesh should be congruent with the mesh in the cover plates at least 7 elements should be used across the height of the bulkhead plates the elements in the bulkheads should as far as possible be square. Secondary box: at least 9 elements should be used across the height of the secondary bulkheads the elements in the bulkhead plates should as far as possible be square. The following displacement restraints shall be imposed on the model: the secondary box are to be fixed in the vertical direction at the intersection lines between the bulkhead plates and

Figure 5-10 Location of staples in the FE model

The interaction between the primary and the secondary box should be modelled as a contact constraint. In addition the primary box should be connected to the secondary box in each corner using elastic springs to prevent undesired relative displacement of the boxes relative to each other. This is a simplification of the stud bolt arrangement that secures both boxes to the hull structure. The stiffness of the connection springs should be taken as 10 000 N/mm. The resin ropes may be disregarded, and the support boundary conditions may be directly imposed at the intersection lines between the bulkhead plates and the bottom plate of the secondary box.

DET NORSKE VERITAS

32

Classification Notes - No. 30.9 June 2006

the bottom plate of the secondary box. In case the resin ropes are included in the model the same condition should be enforced at the bottom of the resin rope elements a minimum restraint against in-plane translation and rigid body rotation of the model should be imposed at two of the bottom corners of the secondary box. The plywood material may either be modelled as a homogeneous linear orthotropic elastic material, or as a 0-90 laminate of layers of orthotropic elastic materials. In the former case the cover plate stiffness should be represented by the bending stiffness moduli of the plate, whereas the membrane stiffness moduli should be used for all other components. Material properties are given in 5.4, and should be selected based on the material directions specified in Figure 5-13 unless other directions are specified on the drawings.

mum moment occur at mid span of the unsupported plate field, and the element mesh contains an element row centred on the plate field.
p Q QFE Lint
Q = Q FE + 0.5 p Lint

Figure 5-14 Correction of FE calculated plate shear force to obtain a better estimate of the plate end value

p Q M MFE Lint
M = M FE + Q Lint 0.25 p L2 int

Cover plates, primary

Cover plates, secondary

Bulkhead plates secondary, transverse end plates primary 2 3

Bulkhead plates primary, transverse end plates secondary 2 3

3 1 2 2 3 1

Figure 5-15 Correction of FE calculated plate bending moment to obtain a better estimate of the plate end value

Figure 5-13 Definition of material directions relative to the NO96 insulation panel

5.3.2 Shear force and bending moment in the cover plates The output parameters should be the bending moment and shear force per unit width of the plate cross-section at locations of maximum response centred below the applied load. In case the bending moment and the shear force are obtained at the integration points of the shell elements adjacent to the mastic supports, the corrections shown in Figure 5-14 and Figure 5-15 should be made to make the results representative for the plate cross-section adjacent to the primary box bulkheads. Lint is the distance from the integration point of the element to the edge of the mastic. Note that the corrections are specified under the assumptions of double cover plates, and that the shear force and bending moment corrections are relevant for each individual plate. No correction of the bending moment is required if the maxi-

5.3.3 Reference stress for buckling strength assessment The in-plane reference stress to be used in the buckling strength assessment of the vertical bulkheads of the primary and the secondary box should be selected at mid height of the bulkheads and in line with the intersection between the bulkheads in the primary and the secondary box as illustrated in Figure 5-16 for the primary box bulkhead and in Figure 5-17 for the secondary box bulkhead.

Reference stress output location

Figure 5-16 Selection of vertical in-plane membrane stress for buckling strength assessment of the primary bulkhead

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

33

5.4 Material stiffness parameters


5.4.1 Plywood The material stiffness properties for the plywood laminate and the individual wood layers of the laminate are given in Table 5-1 and Table 5-2, respectively. The properties are given in terms of the material coordinate systems shown in Figure 5-19. Subscript m denotes the in-plane (membrane) stiffness properties of the plates, and subscript b denotes the bending stiffness properties.
Reference stress output location

Figure 5-17 Selection of vertical in-plane membrane stress for buckling strength assessment of the secondary bulkhead

Surface grain direction

1 2

Grain direction

1 2

If the element mesh contains an even number of elements across the height of the bulkhead and thus does not allow for selection of an element at the centre of the bulkhead, the element farthest away from the stress concentration at the primary and secondary bulkhead intersections should be selected. This implies the element just above the centreline should be selected for the primary bulkhead and the element just below the centreline should be selected for the secondary bulkheads. The stress should be selected from the most highly loaded bulkhead for the considered load condition.
5.3.4 Reference stress for bulkhead intersection crushing assessment The in-plane reference stress to be used in the strength assessment of the bottom plate of the primary box, the cover plate of the secondary box, and the bulkhead plate edges should be obtained at mid height of the primary box bulkhead and in line with the intersection between the bulkheads in the primary and the secondary box as illustrated in Figure 5-17. This is the same stress as should be used in the buckling strength assessment of the primary box bulkheads. 5.3.5 Reference stress for bottom plate crushing assessment The reference stress for the crushing strength assessment of the bottom plate of the secondary box at the intersection with the vertical bulkheads should be selected at the bottom element row of the secondary bulkhead, as illustrated in Figure 5-18. This should be the vertical membrane stress component in the bulkhead.

Figure 5-19 Material coordinate systems for the plywood laminate(left) and the individual layers of wood

The relationship between the individual layer and the integrated laminate properties are:

E m ,i =
t/2

1 Ei ( x3 )dx3 , i = 1, 2 t t / 2

t/2

Gm ,ij =

1 Gij ( x3 )dx3 , i = 1, 2 j = 2, 3 t t / 2

Eb ,i =

12 Ei ( x3 ) x 32 dx3 , i = 1, 2 t 3 t / 2

t/2

The local Ei and Gij of the individual layers should be evaluated in the material coordinate system of the laminate.
Table 5-1 Integrated orthotropic material stiffness properties for the plywood plates Parameter 20C -163C Em,1 [MPa] 9 450 13 200 Em,2 [MPa] 8 000 11 200 Em3 [MPa] 820 1 800 Gm,12 [MPa] 790 2 900 325 700 Gm,13 [MPa] Gm,23 [MPa] 260 550 12 [-] 0.1 0.1 13 [-] 0.1 0.1 23 [-] 0.1 0.1 Eb,1 9 mm [MPa] 10 950 15 350 Eb,2 9 mm [MPa] 6 550 9 150 Eb,1 12 mm [MPa] 10 450 14 650 Eb,2 12 mm [MPa] 7 000 9 800 3] -10 Density [ton/mm 6.8 10 6.8 10-10 Table 5-2 Material properties for the individual wood layers Parameter 20C -163C E1 [MPa] 13 950 19 500 E2 [MPa] 3 550 4 950 E3 [MPa] 820 1 800 G12 [MPa] 790 2 900 G13 [MPa] 1 500 3 150 G23 [MPa] 160 340

Reference stress output location

Figure 5-18 Selection of vertical in-plane membrane stress for crushing strength assessment of the bottom plate of the secondary box

DET NORSKE VERITAS

34

Classification Notes - No. 30.9 June 2006

12 [-] 13 [-] 23 [-]

Density [ton/mm3]

0.1 0.1 0.1 6.8 10-10

0.1 0.1 0.1 6.8 10-10

f1

f2
DAF
(-)

5.4.2 Reinforced polyurethane foam The orthotropic linear elastic material properties for the reinforced polyurethane foam are summarised in Table 5-3.
Table 5-3 Material stiffness properties for the reinforced polyurethane foam Parameter 20C -163C E1 [MPa] 135 170 E2 [MPa] 180 215 E3 [MPa] 65 95 G12 [MPa] 7 11 G13 [MPa] 7 11 G23 [MPa] 7 11 12 [-] 0.4 0.4 13 [-] 0.2 0.20 23 [-] 0.2 0.20 Density [ton/mm3] 1.25 10-10 1.25 10-10

0.5

1.5

2.5

3.5

Rise time/Natural period

Figure 5-20 Illustration of the curve to be used in the simplified assessment of the dynamic amplification factor

The factors f1 and f2 relevant for the Mark III and the NO96 systems are given below.
5.6.2 Mark III system The factors f1 and f2 are given as a function of the vertical position of the considered response variable: At the bottom plywood plate (furthest away from the load) the following values apply: f1 = 1.65 f2 = 1.25 At the upper plywood plate (closest to the load) the following values apply: f1 = 1.25 f2 = 1.1 Linear interpolation may be applied when response variables in between the top and bottom are considered. For moderate design modifications of the system including changes of the mastic support spacing and the bottom plate thickness the natural period to be used in the assessment of the dynamic load factor can be taken as 2.0 milliseconds. More radical design modifications requires that the natural period is established by dynamic finite element analyses. 5.6.3 NO96 system The factors f1 and f2 are functions of the vertical position of the considered response variable, as given in the following. At the mid part of the secondary box bulkhead: f1 = 1.6 f2 = 1.2 At the mid part of the primary box bulkhead: f1 = 1.2 f2 = 1.05 At the cover plywood plates (closest to the load): f1 = 1.4 f2 = 1.1 Linear interpolation may be applied when response variables in between the top and bottom are considered. For moderate design modifications of the system including changes of the internal bulkhead plate thickness the natural period to be used in the assessment of the dynamic load factor can be taken as 1.3 milliseconds. More radical design modifications requires that the natural period is established by dynamic finite element analyses.

5.5 Mastic
The material stiffness properties for mastic is summarised in Table 5-4.
Table 5-4 Material stiffness properties for mastic Parameter 20C E [MPa] 2 900 [-] 0.3 Density [ton/mm3] 1.6 10-9

5.6 Simplified assessment of dynamic response


5.6.1 General The dynamic response is determined from the quasi-static response described above through the application of a dynamic amplification factor, denoted DAF. The factor is a function of the ratio between the pressure pulse rise time tr defined in Figure 3-6 in 3.3.4 and the natural period Tn of the insulation system.

Natural periods to be used in the selection of the dynamic load factor are specified in 5.6.2 and 5.6.3 for the Mark III and the NO96 systems, respectively. Note that the natural period in general varies with the loaded area. However, the consequences of the expected variations is considered to be within the uncertainty limits inherent in the present approach to determine the dynamic load factor. A piece-wise linear curve is used because of the uncertainties involved in determining the rise time of the pressure pulse. Due to this uncertainty, the design curve is specified to envelope the maximum-points on the actual calculated DAFcurves. The design DAF curve is illustrated in Figure 5-20, and is defined as follows: Tr < 0.5 : DAF = f1 Tn

T Tr (0.5,1.0) : DAF = f 1 2 ( r 0.5)( f1 f 2 ) Tn Tn T Tr > 1.0 : DAF = max(1.0, f 2 0.33 ( r 1.0)( f 2 1.0)) Tn Tn

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

35

6. Ultimate Strength of Containment Systems


6.1 General
Acceptance criteria for the ultimate strength assessment of the Mark III and the NO96 insulation systems are given in 6.3 and 6.4, respectively. Note that the acceptance criteria specified for the Mark III system in 6.3 is also applicable to the CS 1 system, but that the comparative strength assessment methodology can not be applied to this system. 6.2 gives guidance on how to use the acceptance criteria to calculate the load capacity of the insulation system as needed to determine the load level for the reference case in the comparative assessment (see 2.2).

If the statistical processing of the load data does not distinguish between foot print orientations 2 and 4, 3 and 5, and 7 and 8, it is sufficient to consider only one from each pair for the flat wall panels.
Table 6-2 Definition of load foot prints for the corner panels Foot Print Width (mm) Length (mm) Area (mm2) FP 1 210 340 71 400 FP 2 210 340 71 400 FP 3 210 680 142 800 FP 4 210 680 142 800 FP 5 420 340 142 800 FP 6 420 680 285 600 FP 7 >600 >680 >408 000
140 mm mastic spacing FP1 150x150mm FP2 150x400mm FP3 150x800mm

6.2 Capacity assessment


The load capacity of the insulation systems in terms of peak impact pressure for a given impact area size should be calculated based on the strength acceptance criteria given in this section, as follows: 1) Calculate the design capacity of each relevant response parameter including dynamic effects by solving the limit state equations for the maximum allowable dynamic response parameter. 2) Invert the quasi-static response analysis to determine the surface pressures corresponding to the critical response. This will be the maximum effective dynamic pressure the insulation system can sustain. 3) Calculate the peak pressure capacity for each considered impact area by dividing the effective dynamic pressure capacity by the relevant dynamic factor.

1
FP4 400x150mm

2
FP5 800x150mm

3
FP6 400x400mm

5 4
FP7 400x800mm FP8 800x400mm

6
FP 9 150x150mm

6.3 Mark III system


6.3.1 Impact areas The strength of the Mark III insulation system should be checked for the impact load foot prints identified in Figure 6-1 and Figure 6-2, and defined in Table 6-1.
Table 6-1 Definition of load foot prints for the flat wall panels Foot Print Width (mm) Length (mm) Area (mm2) FP 1 150 150 22 500 FP 2 400 150 60 000 FP 3 800 150 120 000 FP 4 150 400 60 000 FP 5 800 150 120 000 FP 6 400 400 160 000 FP 7 800 400 320 000 FP 8 400 800 320 000 FP 9 150 150 22 500 FP 10 510 855.5 436 305
7
FP 10 Uniform Load

10

Figure 6-1 Load footprints for strength assessment of the flat wall Mark III insulation panel

DET NORSKE VERITAS

36

Classification Notes - No. 30.9 June 2006

100 mm mastic spacing FP1 210x340mm FP2 210x340mm FP3 210x680mm

6) Bending failure of the secondary barrier level plywood plate (corner/knuckle) at the edge of the hardwood key.

6 5

2 3
FP4 210x680mm FP5 420x340mm FP6 420x680mm

Figure 6-4 Visual identification of locations of critical failure modes for the corner/knuckle Mark III insulation panel

6 4 5

FP 7 Uniform Load

6.3.3 Crushing of primary foam The strength criterion for this failure mode is a stress control criterion for the average through thickness compressive stress in the foam on the form: p DAF F F M

Figure 6-2 Load footprints for strength assessment of the corner/knuckle Mark III insulation panels

where: p is the peak impact pressure acting on the panel. DAF is the dynamic amplification factor as defined in 5.6.1. F is a partial load factor, if applicable. F is the crushing strength of the reinforced polyurethane for the relevant temperature and load rate conditions, as defined in 6.6. M is a partial resistance factor, if applicable.
6.3.4 Crushing of secondary foam The strength criterion for this failure mode is a stress control criterion for the average through thickness compressive stress in the foam on the form:

6.3.2 Failure modes The ultimate strength assessment of the Mark III insulation panel should consider the following set of critical failure modes (see Figure 6-3 and Figure 6-4):

DAF F

F M

where: is the average through thickness stress in the panel. The foam crushing criterion will ensure that the loads do not reach levels where the structure will experience total collapse or excessive deformations. A main issue with this kind of strength criteria is that it is not unique in terms of the local state of stress and strain in the structure for different mastic support conditions or other relevant design modifications. This is not believed to be a big issue for the ultimate strength, since the foam appears to be robust and able to sustain rather large ranges of inelastic deformations without any observable differences in strength and damage.
Figure 6-3 Visual identification of locations of critical failure modes for the flat wall Mark III insulation panel

6.3.5 Shear failure of bottom plywood plate The maximum shear force in the bottom plywood plate in the most highly loaded cross section adjacent to the mastic support should fulfil the following requirement:

1) 2) 3) 4) 5)

Crushing of primary foam Crushing of secondary foam at plywood interface Shear failure of plywood plate at support Bending failure of plywood plate. Shear failure of the secondary barrier level plywood plate (corner/knuckle) at the edge of the hardwood key.

Q ( p DAF F )

Qc

where: Q is the calculated dynamic shear force per unit plate width at the most highly loaded cross-section of the bottom plywood plate for a quasi-static pressure p DAF . The calculation of this quantity is described in more detail in 5.2.4.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

37

Qc is the shear force capacity per unit plate width of the plywood plate for the considered material orientation, as defined in 6.6. Unless a specific orientation of the plywood plate with respect to the mastic support can be guaranteed the value should be representative for the weakest cross-sectional direction of the laminate.
6.3.6 Bending failure of bottom plywood plate The maximum bending moment in the bottom plywood plate in the most highly loaded cross-section should fulfil the following requirement:

considered cross-section of the plate can be guaranteed, the value should be representative for the weakest cross-sectional direction of the laminate.

6.4 NO96 system


6.4.1 Impact areas The strength of the NO96 insulation system should be checked for the impact load foot prints identified in Figure 6-5 and Figure 6-6 and defined in Table 6-3. Foot print 1 covers the entire insulation box surface and is not shown in the figures.

M ( p DAF F )

Mc

where: M is the calculated dynamic bending moment per unit plate width at the most highly loaded cross-section of the bottom plywood plate for a quasi-static pressure p DAF . This is most likely at a cross-section adjacent to the mastic support. The calculation of this quantity is described in more detail in 5.2.4. Mc is the bending moment capacity of the plywood plate for the considered material orientation, as defined in 6.6. Unless a specific orientation of the plywood plate with respect to the mastic support can be guaranteed the value should be representative for the weakest cross-sectional direction of the laminate.
6.3.7 Shear failure of secondary barrier level plywood plate (corner/knuckle) The maximum shear force in the bottom plywood plate in the most highly loaded cross-section adjacent to the mastic support should fulfil the following requirement:

The foot prints have been carefully selected to obtain lower bound strength values for all relevant components of the primary and the secondary insulation box. Foot prints 2-8 are selected to primarily cover the strength of the secondary box. Of these, foot prints 2-5 cover the strength of the external members of the insulation box, whereas foot prints 6-8 cover the internal members. Foot prints 9-13 are selected to primarily cover the strength of the primary box. Of these, foot prints 9-11 cover the strength of the external members of the insulation box, whereas foot prints 12 and 13 cover the internal members.
Table 6-3 Definition of load foot prints Foot Print Width (mm) Length (mm) FP1 (full box) 1 141 990 FP2 188 249 FP3 188 488 FP4 187 990 FP5 249 990 FP6 353 249 FP7 353 368 FP8 481 990 FP9 188 118 FP10 530 118 FP11 1 140 118 FP12 530 249 FP13 1 141 251 Area (cm2) 11 296 466 915 1 849 2 466 876 1 298 4 759 221 625 1 346 1 320 2 864

Q DAF F

Qc

where: Q is the calculated static shear force per unit plate width at the most highly loaded cross-section of the plywood plate for a quasi-static pressure p. The calculation of this quantity is described in more detail in 5.2.5. Qc is the shear force capacity per unit plate width of the plywood plate for the considered material orientation, as defined in 6.6. Unless a specific orientation of the plywood plate with respect to the considered cross-section of the plate can be guaranteed, the value should be representative for the weakest cross-sectional direction of the laminate.
6.3.8 Bending failure of secondary barrier level plywood plate (corner/knuckle) The maximum bending moment in the bottom plywood plate in the most highly loaded cross-section should fulfil the following requirement:

8 7 5 4 6 2 3

M DAF F

Mc

where: M is the calculated static bending moment per unit plate width at the most highly loaded cross-section of the plywood plate for a static pressure p. The calculation of this quantity is described in more detail in 5.2.5. Mc is the bending moment capacity of the plywood plate for the considered material orientation, as defined in 6.6. Unless a specific orientation of the plywood plate with respect to the

Bulkhead direction in primary box

Figure 6-5 Load foot prints primarily intended for assessment of secondary box members

DET NORSKE VERITAS

38

Classification Notes - No. 30.9 June 2006

vertical bulkheads:
11

Q DAF F
where:

Qc

Q is the calculated section shear force per unit plate width at the most highly loaded cross-section of the cover plates. The calculation of this quantity is described in more detail in 5.3.2.
13 12

Qc is the section shear force capacity of the plywood plate for the considered material orientation.
6.4.4 Bending failure of cover plate(s) The strength criterion for this failure mode is a stress control criterion for the bending stress in the cover plates in the most highly loaded cross-section adjacent to the vertical bulkheads:

10

M DAF F
where:

Mc

Bulkhead direction in primary box

Figure 6-6 Load foot prints primarily intended for assessment of primary box members

6.4.2 Failure modes The strength assessment of the NO96 insulation boxes is based on the following set of critical failure modes (see Figure 6-7):

M is the calculated bending moment per unit plate width at the most highly loaded cross-section of the cover plates, most likely to be at a cross-section adjacent to vertical box bulkheads. The calculation of this quantity is described in more detail in 5.3.2. Mc is the bending moment capacity of the plywood plate for the considered material orientation.
6.4.5 Buckling of plywood bulkheads

1) Shear failure of the cover plate(s) of the primary insulation box. 2) Bending failure of the cover plate(s) of the primary insulation box. 3) Buckling of internal and external bulkheads of the primary insulation box. 4) Buckling of internal and external bulkheads of the secondary insulation box. 5) Through thickness crushing of the bottom/cover plates of the primary/secondary boxes at the intersection between the bulkheads of the primary and secondary boxes. The crushing deformation will lead to failure of the same plates in bending or shear.
Bending failure of the cover plates Buckling of primary box bhds

6.4.5.1 General This section presents the formulas and procedures to be used for buckling strength assessment of the internal and external bulkheads of both the primary and secondary insulation box. The section is organised into three subsections describing the assessment of the static buckling capacity of the plywood bulkheads, how to account for the additional edge bending moment load potentially experienced by the external primary bulkheads, and last how to utilise the bulkheads in the dynamic and unstable load regime during rapid load events. The buckling strength check should be based on the following strength equation:

DAF F
Shear failure of the cover plates

cD M

where:

is the nominal in-plane stress in the bulkhead plate, as described in 5.3.3. D is the nominal in-plane buckling capacity of the bulkhead
considering both dynamic and temperature effects.

Buckling of secondary box bhds

Crushing of plates at bulkhead intersections

The temperature dependent material stiffness and strength parameters should be evaluated at the mid-height of the primary and secondary bulkheads according to the procedure described in 5.4 and 5.6. 6.4.5.2 Static capacity assessment The proposed buckling strength design curve is shown in the figure below, including results from finite element analyses for room temperature and uniform load for different slenderness. The buckling strength design curve is given as:

Figure 6-7 Failure modes considered for the NO96 system

6.4.3 Shear failure of cover plate(s) The strength criterion for this failure mode is a stress control criterion for the through thickness shear force in the cover plates in the most highly loaded cross-section adjacent to the

c 1.05 + 2 (1.05 + 2 ) 2 42 = F 2 2
where c is the static buckling strength at room temperature,

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

39

is the reduced slenderness defined as

F = E
F is the material compressive strength in the direction of the
load (measured as average stress over the plate cross-section) and E is the elastic buckling strength of the plate. A graphical presentation of the buckling strength design curve is shown in Figure 6-8.
Design curve

, minimum 1.0 + 0.0042 (h B) where h is the height of the bulkhead, B is the width of the bulkhead, and b is the width of the loaded area (all given in mm). The k1-factor as a function of load width b is shown in Figure 6-9. The width b may be taken as the width of the zone where the vertical stress is larger than half of the maximum stress at mid height of the bulkhead, as illustrated in Figure 6-10.

k1 =

0.007

1 0.9 0.8 0.7


1.7 1.6 1.5 1.4

c/f

0.6 0.5 0.3 0.2 0.1 0 0 0.5 1 1.5 2 2.5 0.4

k1

1.3 1.2 1.1 1 0.9 0.8 0 0.001 0.002 0.003 0.004 0.005

Figure 6-8 Graphical representation of the buckling strength design curve

b/(h*B)

The elastic buckling stress for a simply supported bulkhead subjected to loads over parts or all of its edge can be expressed as: E = k1 E ,1 + k 2 F where: E,1 is the elastic buckling stress for a simply supported bulkhead subjected to uniform load, k1 is a factor depending on the size of the load exposed region of the bulkhead, k2 is a factor to account for the increased buckling strength of slender bulkheads caused by the rotational restraint resulting from the finite thickness of the bulkheads, and F is the material compressive strength as defined above. The factors k1, k2, and E,1 are described in more detail below. The elastic buckling stress for a simply supported bulkhead subjected to uniform load, E,1 may be taken as:

Figure 6-9 The part load elastic buckling stress correction factor k1

load width b
=0/2 =0/2

=0

E ,1 =

2D * t
12 h

Figure 6-10 Illustration of the criterion for selection of the load width b

Where t is the thickness, h is the height of the bulkhead plate, and D* is the generalised bending modulus of the bulkhead plate, defined as:

E, E 2 D* = h 4 b4,1 + b42 + 2 2 ( 12 E b, 2 + 2G12 ) h B h B


where Eb,1, Eb,2 are bending stiffness properties of the laminate, 12 is the in-plane Poissons ratio, and G12 is the in-plane shear modulus of the plate, all defined in Table 5-1. When the bulkheads are subjected to compressive loading over a limited area only, the slenderness is modified using the factor k1 defined as:

The case of a part load located towards the end of the bulkhead has been considered, but found not to be critical. In this case, the redistribution of the load will be smaller, since the stress can redistribute in one direction only. However, the restraining effect of the end bulkhead will be larger when the load is close to the restraint, and some of the loading will be taken up by the end bulkhead. If E is less than 0.5F, the elastic buckling stress may be increased by k2F. The factor k2 is defined to account for the increased buckling strength of slender bulkheads caused by the rotational restraint resulting from the finite thickness of the bulkheads, as illustrated in Figure 6-11.

DET NORSKE VERITAS

40

Classification Notes - No. 30.9 June 2006

Figure 6-11 Illustration of the end restraint effect experienced by slender bulkheads during buckling deformations

The rotational restraint is a function of the through-thickness stiffness of the top and bottom plywood plates connected to the bulkheads, and is taken to be linearly decreasing with the applied stress, as follows:

6.4.5.3 Correction of static capacity due to bending moment The outer bulkheads in the primary box will be subjected to bending moment due to the deformation of the top plate, while the outer bulkheads in the secondary box will be subjected to bending moment due to deformation of the entire primary box. The bending stress in the bulkheads is calculated by linear finite element analysis. If bending is significant, the critical buckling stress must be reduced as a result of the bending moment using a buckling-bending interaction equation. In this case, the rotational restraint correction should not be applied, i.e. k2 = 0. The bending stress resulting from linear analyses indicates that the bending moment in the outer bulkhead in the secondary box is relatively small. For uniform load, the bending stress in the outer bulkhead is approximately 10% of the total stress, which means that it may be neglected. This is also supported by the capacity tests carried out by GTT, which gives a capacity in excess of the one calculated by the design formulas proposed, without considering the bending moment effect. For the end bulkheads in the primary box, the bending effect is much larger. For a strip load on the edge of the box, the bending stress is approximately 37% of the total stress, which means that the bending moment must be accounted for. The following interaction equation is used for the external bulkheads in the primary box:

k 1 E ,1 > 0.5 : k 2 = 0 F k 1 E ,1 < 0.18 : k 2 = 0.112 F k 1 E ,1 else : k 2 = 0.35 0.5 F


The k2-factor as a function of the normalised elastic buckling strength is illustrated in Figure 6-12. In the case that dynamic strengthening is accounted for, as defined in 6.4.5.4, the rotational restraint effect must be re-evaluated after calculation of the dynamic strength. The k2-factor must then be re-calculated, using the dynamic strength d instead of E,1 in the formula above. Using the modified k2-factor, the modified static and dynamic strength is calculated. Further iterations are not necessary.
0.12 0.1 0.08 0.06 0.04 0.02 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

+ c

b F (1

) E

=1

where b is the bending stress in the bulkhead, F is the yield stress of the outer veneer, and is the critical average stress under combined axial load and bending. The values to be used for the bending stress and the yield stress in the above equation depends on whether the plywood plates are modelled as laminate plates or equivalent homogeneous orthotropic plates, as summarised in Table 6-4.
Table 6-4 Definition of bending and yield stress values depending on plate modelling approach Parameter Plate modelling approach Equivalent Laminate plate homogeneous orthotropic plate

b F

b = max av

E veneer E average
E veneer E average

b = max av
avg F =F

veneer avg F =F =F

k2

The parameters referred to in the table are defined as follows: Eveneer and Eaverage are the elastic moduli of the surface veneer and the entire laminate, respectively, in the material direction coinciding with the load. max is the maximum calculated stress at the extreme fibre of the plate. av is the average stress over the plate cross-section, calculated t as av = dz in the case that the z axis is oriented in the 0 through thickness direction of the plate. Table 6-6. Since the bending moment is proportional to the load, b may be replaced by b = Cb, where Cb is the ratio between bending stress and average stress in the bulkhead, as determined from finite element analysis. The solution to the interaction equation is then:
avg F is the material strength of the laminate plate specified in

E / F
Figure 6-12 Graphical presentation of the end restraint correction factor k2 as a function of the normalised elastic buckling strength

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

41

= ( E + c + Cb
where

1 2

E c 1 )+ F 2 F

exceeding 1.0 ms.

K = KA + KB
2 2 2 2 K A = F E + 2C b F c E 2 E c F 2 2 K B = C b E c2 + 2C b E F c2 + c2 F 2

Table 6-5 Relationship between impact pressure rise time and vertical stress response rise times for the primary and secondary box bulkheads t rresponse [ms] t rload [ms] Primary box Secondary box 0.5 0.75 0.7 1.0 1.0 1.0

The bending moment applied to the external bulkheads of the primary box due to bending of the cover plate need not to be taken greater than the value implied by the rotational restraint limit as defined by the above. This means that the maximum moment is defined for the load level giving a nominal compressive stress in the bulkhead plate equal to 0.5F.

DSF f1() f2() 1.0 0.55 1.5 tr/Te

6.4.5.4 Dynamic strength correction Buckling failure is a dynamic event, and it is clear that inertia will allow a buckling exposed structure to sustain in-plane loads beyond the static buckling capacity without suffering damage given that the load event is sufficiently rapid. This effect is accounted for by multiplying the critical buckling stress by a dynamic strength factor (DSF= cD / c ) defined as a function of the slenderness of the bulkhead and the ratio of the rise time of the sloshing impact stress response (tr) and the natural period relevant for lateral oscillation of the bulkhead plate (Te), as follows:

Figure 6-13 Definition of the DSF as a function of the ratio between the rise time of the sloshing stress response and the natural period of the bulkhead plates in the lateral bending deformation mode

The dynamic buckling strength is then found as:

cD = c DSF
The dynamic low temperature buckling strength is limited by the material compression strength: Max( cD ) = F
Dynamic factor

DSF = f 1 ( )

f 2 ( ) f 1 ( ) t r t r , 0.55 Te Te 0.55

t 1 tr DSF = f 2 ( )1 ,0.55 < r 1.50 Te 0.95 Te tr DSF = 1.0, > 1.50 Te


f1() and f2() are given as:

4.5

f1

4 3.5 3 2.5 2 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 3

f 1 ( ) =

f 3 F

cr

, maximum 4 (see Figure 6-14)

f3 = safety factor =0.95 f2() = min(f1(), 1.5) A graphical representation of the dynamic strength factor as a function of the ratio between rise time and natural period, tr/Te, is shown in Figure 6-13. The natural period relevant for lateral oscillation of the bulkhead plate can for all load cases be calculated as:
Te = 2h 2 12 t Eb

Figure 6-14 Graphical representation of the slenderness dependent factor f1

where t is the thickness of the bulkhead plate, h is the height of the bulkhead, is the density of the plywood, and Eb is the bending stiffness of the plate about its stiffest axis, ref. Table 5-1. The relationship between the rise time of the impact pressure and the rise time of the stress response in the bulkheads of the primary and the secondary insulation boxes are summarised in Table 6-5. Response rise times for impact pressure rise times between the values given in the table should be obtained by linear interpolation. The response rise time can be taken as equal to the impact pressure rise time for impact pressure rise times

6.4.6 Crushing of plates at bulkhead intersections The crushing strength is measured in terms of the nominal stress in the vertical bulkhead of the secondary insulation box, av , measured against a critical value of this parameter, cav associated with through thickness crushing of the insulation box cover plates and more importantly the associated bending failure of these cover plates. The capacity has been determined from laboratory tests and numerical simulations. The strength check should be carried out using the following equation:

av DAF F

cav M

av

is the nominal stress in the bulkhead of the primary insu-

DET NORSKE VERITAS

42

Classification Notes - No. 30.9 June 2006

lation box in the region of the bulkhead intersections.


av c

is the through thickness compressive strength of the plywood plate measured in terms of the nominal stress in the secondary box bulkhead.
6.5 Plywood strength data
Strength data for 9 mm and 12 mm thick plywood plates are summarised in Table 6-6. The first subscript c identifies that this is a limit value (critical), and the second and potentially third alphanumeric subscript refers to the material directions of the laminate, as defined in Figure 5-19. For bending moments the subscript refers to the vectorial direction of the moment, i.e. M1 denotes bending about the 2-axis and M2 denotes bending about the 1-axis.
Table 6-6 Mean strength properties for the 9 mm and 12 mm plywood laminates 9 mm 12 mm 20C -163C 20C -163C c,1t, tension (MPa) 70.0 60.0 70.0 60.0 c,1c, compr. (MPa) 43.0 65.0 43.0 65.0 c,2t, tension (MPa) 54.0 46.0 54.0 46.0 c,2c, compr. (MPa) 32 32 32 32 cav compr. (MPa) Mc,1 (Nmm/mm)* 1 100 935 1 850 1 580 Mc,2, (Nmm/mm)* 760 650 1 380 1 180 Qc,13 (N/mm) 59 59 79 79 Qc,23 (N/mm) 43 43 58 58 * Calculated based on the veneer tensile strength.

E3 is the relevant temperature dependent through thickness elastic modulus of the reinforced polyurethane foam as defined in 5.4. t rresponse is the estimated rise time of the average through thickness stress, and should be calculated as follows: For the upper part of the foam the response is considered to follow the load, and the response of the rise time should consequently be taken equal as the load rise time. For the foam adjacent to the bottom the response rise time should be taken equal to the load rise time for load rise times larger than 1.5 ms. For shorter load rise times the response rise time should be obtained by interpolation between the values given in Table 6-8.
Table 6-8 Relationship between load and response rise times for bottom foam for load rise times below 1.5 ms load [ms] t response [ms]

tr

0.250 1.500

0.500 1.500

7. Stiffness and Strength of Hull Structure


7.1 General
The sloshing impact loads acting on the containment system inside the cargo tanks must be transferred into the supporting hull structure. It is therefore necessary to ensure that the hull structure has the sufficient stiffness and strength to carry the sloshing loads. The hull structure assessment should be carried out in two steps in order to make sure that: 1) The stiffness of the hull structure is sufficient to provide adequate support for the containment system. 2) The strength of the hull structure is sufficient to carry the sloshing loads without suffering large permanent deformations. For assessment of the hull structure stiffness, it is sufficient to consider the plate stiffness only. The stiffener span is much longer than the stiffener spacing, which means that the curvature of the plate during deflection is much smaller in the stiffener direction than in the cross-stiffener direction. The assessment of the hull plating stiffness is discussed in 7.3. For assessment of the hull structure strength, it is sufficient to consider the stiffener strength only. The stiffness criterion presented in 7.3 will be dimensioning for the hull plating. The hull plating will not be critical from a strength point of view, since the containment system will redistribute the sloshing loads into the stiffeners when the plate is deflecting. Hence, the load will eventually be carried by the stiffeners. When the plate deforms, the containment system will re-distribute the sloshing impact load from the plating and directly into the stiffeners, and the strength of the plating will therefore not be critical. The assessment of the stiffener strength is discussed in 7.4. The hull stiffness and strength should be considered for the most sloshing exposed regions of the tank. For normal filling operations, particular attention should be paid to the following areas, Figure 7-1: the inner deck and chamfer area at locations close to the transverse bulkheads the transverse bulkheads at locations close to the inner deck the inner deck and chamfer area close to the upper chamfer knuckle along the entire length of the tank the chamfer area close to the lower chamfer knuckle along the entire length of the tank.

6.6 RPUF strength data


The temperature and strain rate dependent through thickness crushing strength of the reinforced polyurethane foam is given in Table 6-7.
Table 6-7 Temperature and strain rate dependent strength for through thickness compression of the RPUF Strain rate Temperature 20C -163C 0* 1.2 2.0 0.1/s 1.35 2.2 4/s 1.6 2.3 100/s 1.65 2.3

* Quasi-static loading.

6.6.1 Evaluation of temperature dependent strength It should be assumed that the temperature at the hull structure side of the containment system is 20C, with a linear variation down to -163C at the primary barrier of the system. The material strength can be taken as a linear function of the temperature between the limits given in Table 6-6 and Table 6-7. The material strength for temperatures between the limits given in should thus be obtained by linear interpolation between these limits based on the through thickness coordinate at the location being investigated. 6.6.2 Evaluation of strain rate dependent strength The strain rate dependent material strength should be determined based on linear interpolation between the values given in Table 6-7 using a strain rate calculated as follows:

& =
where:

DAF
2 E3 t
response r

is the average through thickness stress in the foam as defined in 5.2.3.


DAF is the dynamic amplification factor as defined 5.6.1.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

43

CL

Upper chamfer knuckle

Lower chamfer knuckle

1 Frame spacing

Figure 7-1 Part of inner shell which should be specially considered with respect to sloshing (top view)

The extent of the areas is to be determined based on results from sloshing tests for the actual tank configuration. As for the membrane containment system, the assessment of the hull structure is carried out using a comparative approach. Hence, the stiffness and strength of the hull structure for the target case is compared with the corresponding stiffness and strength for the reference case. The reference case to be considered is defined in the next Section.
Figure 7-2 Illustration of stress concentration in the foam of the Mark III containment system above the stiffener

7.2 Comparative basis (reference case)


The basis for the comparative assessment of the hull structure of the new LNG carrier should be taken as the reference case vessel specified in 1.6.2. Since tank fillings between 90% and 98.5% of the tank height are assumed for the reference case, a typical deck panel is considered as the basis for the comparative assessment. A stiffened panel with the dimensions listed in the following is taken as representative for the reference case: Vessel with Mark III containment system:
t (mm) 13 s (mm) 880 L (mm) 2 800 Profile (mm) 250 90 12/18 Figure 7-3 Illustration of stress concentration in the secondary bulkhead of the NO96 containment system above the stiffener

Vessel with NO96 containment system:


t (mm) 14 s (mm) 810 L (mm) 2 800 Profile (mm) 250 90 9/14

where t is plate thickness, s is stiffener spacing, L is stiffener span, and Profile is stiffener scantlings. The yield stress for the hull structure material for the reference case is 235 MPa.

7.3 Stiffness of hull plating


7.3.1 General In the response and strength assessment of the membrane containment system, described in Section 5 and 6, it is assumed that the structure supporting the containment system is flat and rigid. In reality the hull structure is flexible, and the hull deflection will affect the stress distribution in the containment system.

The stiffness of the hull structure must therefore be assessed in order to ensure that the sloshing impact loads do not lead to unacceptable deflection of the hull, which will lead to stress concentration and possible damage to the containment system. As the shell plate below the insulation boxes deflects due to the sloshing loads, the stress in the containment system will redistribute towards the stiffeners. An illustration of the stress concentration occurring in the containment system at the position of the stiffener is shown in Figure 7-2 for the Mark III system and in Figure 7-3 for the NO96 system.

It is worth noting that the stress concentration due to the hull deflection is not a very localized effect, compared to the dimensions of the insulation system. Since the stiffener spacing is large compared to the spacing of the mastic strips supporting the containment system, the stress concentration due to the hull deflection acts over a large area of the insulation boxes. The effect of the plate deflection may therefore be considered as relevant for several of the failure modes in the containment system. For assessment of the stiffness of the hull structure, it is normally sufficient to consider the stiffness of the plating between the stiffeners. For an insulation box located above a stiffener and close to the transverse bulkhead, the stiffener scantlings will also influence the stress distribution, since some of the load is transferred directly into the transverse bulkhead. The amount of loading transferred into the transverse bulkhead increases with increasing stiffener deflection, i.e. with decreasing stiffener scantlings. This means that increasing the stiffener scantlings will increase the stress concentration in the insulation box at the location of the stiffener. In practice the most critical case will be an insulation box located above a longitudinal girder, which may be considered as an infinitively stiff longitudinal. Hence, for the stiffness assessment, the insulation box may be considered as rigidly sup-

DET NORSKE VERITAS

44

Classification Notes - No. 30.9 June 2006

ported at the position of the stiffener/girder. The hull stiffness should be controlled by assessment of the stress concentration occurring in the containment system due to the flexible hull plating, relative to the case of rigid support conditions. The stresses need to be assessed by use of finite element analysis, as described in the next section.
7.3.2 Response assessment In order to assess the stress concentration in the containment system due to the flexible support, two sets of finite element analyses need to be carried out for each of the target case and the reference case:

Trv. Bulkhead

Load area

1) The first analysis is carried out with the actual hull plating supporting the containment system 2) The second analysis is carried out with rigid support conditions, i.e. with the containment system resting on a flat plate Both analyses should be static or quasi-static, and linear elastic. The non-linear effects related to the plate deflection are assumed to be rather small for the expected load magnitudes, and can be neglected. The dynamic effect is accounted for by applying a dynamic factor in the acceptance criterion described in 7.3.3. The finite element models should include one full box of the containment system. The models used for response assessment of the containment system may be used also for these analyses. However, it is essential that the mesh density is the same for the two analyses, since the stress will be mesh sensitive. If the same mesh density is used for both analyses, the calculated stress concentration will be less dependent on the mesh density. For the rigid support case, only the containment system itself is needed in the finite element model, since the bottom of the mastics are considered as fixed in the vertical direction. For the flexible support case, the part of the hull plating located between primary strength members should be included in the model. The mastic should be connected to the plating, and the stiffeners should be represented by fixing the plating in the vertical direction at the positions of the stiffeners. The containment system should be positioned as the first flat panel away from the transverse bulkhead, i.e. one corner panel distance from the bulkhead. The load area considered for the analyses should be the area that is critical with respect to the stress concentration. A quadratic load patch with width and length equal to the stiffener spacing may be considered, unless information from sloshing tests indicate that other areas are more critical. Uniform pressure should be applied over the load area. Since the analyses are linear, the stress concentration will be independent of the load magnitude. Hence, a unit load may be applied for all cases. The load area should be centered above the stiffener, as illustrated in Figure 7-4.

Stiffeners

Mastic

Figure 7-4 Illustration of location of containment system and load area for analysis of Mark III panel

In case of Mark III containment system, the resulting vertical stress is measured in the foam directly above the middle stiffener and middle mastic. In case of NO96 containment system, the resulting vertical stress is measured as the membrane stress in the mid part of the middle bulkhead of the secondary box. The stress concentration due to the flexible support is then calculated for the reference case and the target case as

flex SCFref = 3 fix 3


SCFtar
where:

ref
tar

3flex = fix 3

SCF = Stress concentration factor. 3 = maximum vertical stress in the critical part of the containment system, as defined above. fix = Containment system with fixed support. flex = Containment system with flexible support.
7.3.3 Acceptance criterion The stiffness assessment of the hull structure is carried out under the presumption that the strength of the containment system has already been assessed for the case of a rigid support condition. A potential load increase from the reference case to the target case should therefore already be accounted for by a capacity increase of the containment system. Hence, the hull stiffness criterion shall only ensure that the stress concentration due to the hull deflection is not larger for the target case than the reference case.

The dynamic amplification in the containment system is largest for the smaller rise times, but for such small rise times the effect of the hull deflection is found to be less important. The stress concentration due to the hull deflection is larger for somewhat larger rise times, where dynamic effects are less significant. For this reason, the assessment of the hull stiffness may be carried out using a static approach, using a safety factor to account for the potential difference in dynamic effects on the stress concentration between the reference case and the target case. Using the comparative approach, the acceptance criterion for the target case is determined by the reference case. The SCF calculated for the target case should satisfy the following reDET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

45

quirement:

SCFtar DAF
where:

SCFref

ing may be considered. Three stiffener spans should be included in the model, one on each side of the span where the load is acting. The number of elements over the stiffener web height and along the plate flange should be sufficient to properly capture the representative shear and bending stress distribution. The refinement of the model should generally be the same for the reference and target case. The sloshing load may be applied directly to the stiffener as a line load. The stress distribution through the containment system does not need to be considered for the stiffener assessment. The load area should be taken as the critical area according to the load-area curve determined from sloshing tests. A load width equal to the stiffener spacing and a length of approximately 1.5 m may be considered, unless information from sloshing tests indicate that other areas are more critical. The load should be applied at a location of one corner panel distance away from the transverse bulkhead. The transverse bulkhead and girders may be accounted for by specifying simple supports, as illustrated in Figure 7-6.

DAF is a partial factor accounting for potential difference in M is the partial resistance factor
The following values may be used, as specified in 1.7.8:

dynamic amplification between the target and the reference case

DAF = 1.15 M = 1.0


The suggested value of the factor DAF have been determined based on results from a large number of dynamic analyses carried out for various hull geometries. The factor may alternatively be determined by carrying out dynamic analyses for the reference case and the target case for various rise-times, and comparing with results for the quasi-static case. If the acceptance criterion is not fulfilled, the stiffness of the hull plating needs to be increased. The stiffness is a function of stiffener spacing and plate thickness. Hence, the plate stiffness may be increased by reducing the stiffener spacing or by increasing the plate thickness.

p Cofferdam

7.4 Strength of hull structure


7.4.1 General The strength of the hull structure must be considered in order to ensure that the sloshing impact loads do not lead to excessive permanent deformations in the hull. An illustration of plastic strains remaining in the stiffener after loading and offloading of the sloshing impact pressure is shown in Figure 7-5, where the pressure has been applied to the top of the containment system. The containment system is indicated with grey colour in the figure. For this specific case, the most highly utilized part of the stiffener is close to the transverse bulkhead, where large shear stresses occur.

Figure 7-6 Illustration of sloshing load applied to stiffener

The longitudinal stress L occurring in the stiffener due to the maximum global bending moment should be considered to act simultaneously with the sloshing load p. Hence, a uniform axial stress corresponding to the global bending stress should be superimposed on the stress due to the sloshing load. The value of the stress L should be taken according to the design bending moment both for the reference case and for the target case. The value of the sloshing pressure for the reference case and the target case may be taken as:

p ref = 0.5MPa p tar = load p ref


where:

load is the load increase factor for the target case relative to the reference case, determined from sloshing tests:

load =
Figure 7-5 Illustration of plastic strains in the stiffener resulting from sloshing pressure on containment system

p tar p ref

The pressures ptar and pref should be evaluated for the most critical load area. Boundary conditions may be applied as shown in Figure 7-7. The longitudinal stress L may be included directly in the analysis by prestressing of precompressing the model. When such applied, any corresponding stresses in the transverse direction should be eliminated. Alternatively the longitudinal stress may be superposed as follows:

The strength of the stiffeners in the inner hull structure is assessed using a comparative approach. Response analyses should therefore be carried out both for the reference case and for the target case, as described in the next section.
7.4.2 Response assessment For assessment of the stiffener strength, linear elastic and static finite element analyses should be carried out for the reference case and the target case. A single stiffener with attached plat-

x = x ,local + L
where x,local is the local bending stress due to the sloshing patch load.

DET NORSKE VERITAS

46

Classification Notes - No. 30.9 June 2006

z x y Clamped z-dir. fixed z-dir. fixed Clamped Symmetry

The strength of the pump tower main structure, the base support and the liquid dome area shall then be checked with respect to ULS and FLS. In addition, a vibration check of the main structure should be carried out.

8.2 Response analysis of main structure


8.2.1 Finite element model The finite element analysis of the main structure of the pump tower should be carried out as a 3D analysis. The analysis may be linear elastic. A principle sketch of the tower structure is shown in Figure 8-1.

Figure 7-7 Proposed boundary conditions for strength assessment of stiffener

The maximum von Mises equivalent stress e occurring in the stiffener web or the stiffener flange should be extracted from the analyses. Only membrane stress, i.e. the stress in the middle of the plate, needs to be considered. It should be noted that modelling only one stiffener is a simplification, and the response obtained from this analysis will not be realistic in an absolute sense. In the real case, the load will redistribute to the surrounding stiffeners when the loaded stiffener starts to deflect, and the stress and deformation will therefore be reduced. Hence, the simplified model is only relevant in a comparative strength assessment.
7.4.3 Acceptance criterion Using the comparative approach, the acceptance criterion for the target case is determined by the reference case. The von Mises stress in the stiffener for the target case should satisfy the following requirement:

Emergency pipe Discharge pipes Braces

Tubular joint

e,tar DAF

e, ref M
Base plate Guiding system

where: DAF is a partial factor accounting for potential difference in dynamic amplification between the target and the reference case. M is the partial resistance factor. The von Mises equivalent stress is calculated as:
2 2 e = x + y x y + 3

Figure 8-1 Illustration of pump tower main structure

The following values may be used for the partial safety factors:

DAF = 1.3 M = 1.0

The suggested value of the factor DAF have been determined based on results from a large number of dynamic analyses carried out for various hull geometries. The factor may alternatively be determined by carrying out dynamic analyses for the reference case and the target case for various rise-times, and comparing with results for the quasi-static case.

The main strength members of the pump tower are the emergency pipes and the two discharge pipes. The columns are connected by intermediate braces, and are fitted to the base plate at the bottom of the tower. The filling pipe and the float level gauge are connected to the tower by supports. The members that should be included in the finite element model are: emergency pipe discharge pipes (port and starboard) filling pipe float level gauge supports for filling pipe and float level gauge braces base plate.

8. Response and Strength of Pump Tower and Supports


8.1 General
Finite element analyses of the pump tower structure shall be carried out in order to determine the response in the structure due to the loads acting for the ULS and the FLS assessment. Separate analyses should be carried out for the main structure, for the base support, and for the liquid dome area, as described in the following sections.

The columns and braces may be modelled with beam elements, while the base plate should be modelled with shell elements. The beam elements should be positioned in the at the centre of each column/brace, as indicated by the dashed lines in Figure 8-2, with the finite element nodes at the intersection between the elements. The restraining effect of the joint on the braces may be accounted for by specifying rigid ends of the elements. Loads should only be applied to the free span part of the brace elements.

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

47

The filling pipe and the float level gauge should be connected to the emergency pipe and the horizontal back struts, respectively. The connection should be modelled so that vertical displacement and rotation of the pipes are allowed.
8.2.4 Material properties Characteristic values of the material properties are to be applied in the response analysis. The pump tower is usually constructed by stainless steel, and typically of qualities in the 300series. Properties for the 304L stainless steel are given in Table 8-1.
Table 8-1 Material properties for 304L stainless steel E (MPa) 200 000 Modulus of elasticity () 0.3 Poissons rato (kg/m3) 7 970 Density y,20 (MPa) 170 Yield limit at 20C y,-78 (MPa) 180 Yield limit at -78C y,-163 (MPa) 220 Yield limit at -163C 20 (10-6/C) 16.3 Thermal exp. coefficient at 20C -163 (10-6/C) 13.5 Thermal exp. coefficient at -163C

Figure 8-2 Illustration of beam element modelling

8.2.2 Load application Sloshing loads should be applied to the columns and braces located below the liquid dome. Gravity, inertia loads and thermal loads should be applied to all elements. The mass and load/drag force effect of additional elements should be included (such as pumps, platforms, ladders and valves), by distributing the masses and modifying the drag coefficients to the main members of the tower. The mass of liquid inside the columns must also be included for calculation of inertia forces. The mass may be applied to the columns as an additional distributed mass. However, the amount of liquid will depend on the filling level, and a separate model will therefore be required for each filling level considered. Therefore, the force due to the inertia acting on the additional mass may alternatively be included directly by a distributed load. 8.2.3 Boundary conditions At the top, the emergency pipe and the discharge pipes are connected to the liquid dome cover plate. The boundary conditions at the top of each column may be taken as: Tx = Ty = Tz = Rz = fixed Rx = Ry = free where T is translation, R is rotation, x is longitudinal direction, y is transverse direction, and z is vertical direction, ref. Figure 4-1. Alternatively, translational (kx and ky) and rotational (krx, kry, krz) springs may be used. The spring stiffnesses should then be documented from the finite element analysis of the liquid dome area. The sliding joints of the three upper horizontal struts should be modelled so that axial displacement and rotation about the axial direction of the strut are allowed: Tx = Rx = free where the x-direction is in the length direction of the strut. At the bottom, the emergency pipe and the discharge pipes are connected to the base plate. In order to account for the sliding pads between the base plate and the base support, the boundary conditions of the base plate are taken as: Tx = Ty = Rz = fixed Tz= Rx = Ry = free where x is longitudinal direction, y is transverse direction, and z is vertical direction. Alternatively, translational (kx and ky) and rotational (krz) springs may be used. The spring stiffness should then be documented from the finite element analysis of the base support.

For intermediate temperatures, the yield limit and the thermal expansion coefficient may be interpolated linearly between the specified values. The reference temperature for the thermal expansion is 20C.
8.2.5 Response parameters For tubular members, the following result parameters should be determined for critical sections along the length of the member:

axial force (or tension/compression stress) vertical bending moment (or bending stress) horizontal bending moment (or bending stress) torsional moment (or torsional stress).

For joints, the following result parameters should be determined: axial load horizontal bending moment vertical bending moment. In addition, total reaction forces at the top and at the bottom of the tower is to be determined.

8.3 Response analysis of base support


The finite element analysis of the base support of the pump tower should be carried out as a 3D analysis using shell elements. The analysis may be linear elastic. The finite element model should include the base support itself, and a portion of the double bottom. The portion of the double bottom should be large enough to provide realistic restraint conditions to the base support. Mesh density may be chosen according to recommendations given in /6/ and /7/. The reaction forces determined from the analysis of the pump tower main structure should be applied to the model, and combined with hull girder loads due to global bending moment, double bottom stresses due to external sea pressure, and thermal stresses due to the thermal gradient in the base support. The temperature variation through the height of the base support should be determined by a separate assessment. The resulting stress field in the base support should be determined, and applied for the ULS and FLS strength assessment. Areas of special concern for the base support is the connection between the base support and the inner bottom, the connection to the primary membrane and the connection to the secondary membrane.

DET NORSKE VERITAS

48

Classification Notes - No. 30.9 June 2006

In addition, the reinforcements in the inner bottom underneath the base support should be considered. The complete set of hot spots to be considered should be agreed with the Classification Society.

8.4 Response analysis of liquid dome area


The finite element analysis of the liquid dome area should be carried out as a 3D analysis using shell elements, as illustrated in Figure 8-3. The top cover plate and the connection to the pump tower should be included in the analysis. The analysis may be linear elastic. Mesh density may be chosen according to recommendations given in /6/ and /7/. The reaction forces determined from the analysis of the pump tower main structure should be applied to the model, and combined with hull girder loads due to global bending moment. The resulting stress field in the liquid dome area should be determined, and applied for the ULS and FLS strength assessment. Areas of special concern are the connection between the pump tower structure and the liquid dome cover plate, and the corner areas of the liquid dome.

eral sections along the length of each brace, in order to make sure that the most critical section is checked, i.e. the worst combination of axial and bending stress. The shear capacity of the tubular joints should be checked according to API RP 2A /9/. For each failure mode, it should be checked that the strength of the structure is acceptable according to the requirements given in 1.7.

8.6 FLS assessment


The fatigue damage should be calculated for relevant parts of the structure. It should be checked that the maximum damage is acceptable according to the requirements given in 1.7. For FLS assessment of a vessel with ordinary operation, loading and offloading at shore terminals, the ships operational profile through its lifetime may be assumed as 10% in harbour, 50% in full load and 40% in ballast. For the full load condition, a certain ratio of the life time should be assumed to be in the range of 70-98% filling. For ballast condition, filling ratios between 5-10% should be assumed. The ships operating profile could be taken from Table 8-2 for ships intended for normal world wide trading.
Table 8-2 Fraction of time at sea in loaded, ballast and part load condition Operating condition Fraction of design life Full load, 95% tank filling 0.40 Ballast, 10% of tank length filling 0.40 Part load, 70% tank filling 0.10 Harbour 0.10

All wave headings (0-180) should be considered, with equal probability. Long term stress distributions are to be determined for inertia loads and sloshing loads. The long-term FLS stress distribution may be described by a two parameter Weibull distribution, characterised by a reference stress range and the Weibull slope parameter.
Figure 8-3 Example of FE-model of liquid dome area

For inertia loads, the ordinary Weibull-distribution for ship motion should be used. For sloshing loads, the long term distribution must be determined by combination of short-term results. In principle, the long term stress distribution for the combined stress effect should be established. However, if it can be documented that that inertia forces are the main contributor to fatigue damage for high fillings, and sloshing loads are the main contributor for low fillings, the FLS assessment may be carried out by considering the two effects separately. The fatigue life may be calculated based on the SN fatigue approach under the assumption of linear cumulative damage (Miner-Palmgren rule):

8.5 ULS assessment


The main structure, the base support and the liquid dome area should be checked with respect to yielding, using the von Mises yield criterion:

S F
where: S = e Rc = F

Rc

D=
i =1

ni Ni

where F is the yield stress of the material, and e is the von Mises equivalent stress:
2 2 e = x + y x y + 3

where D is accumulated fatigue damage, ni is number of stress cycles within stress range i, and Ni is number of cycles to failure at constant stress range i, according to the appropriate SN-curve. The cumulative damage for a design fatigue life of 40 years should not exceed the acceptable damage. The total damage is found by summing the damage for each sea state (according to the North Atlantic scatter diagram) for each sea state: all load conditions (full load/ballast) for each sea state and load conditions: all headings (0-180)

In addition, the tubular members of the main structure should be checked with respect to buckling and combinations of bending and buckling. The tubular joints should be checked with respect to shear capacity. The buckling/bending check of the tubular members may be carried out according to DNV Classification Note 30.1 /8/ or API-RP-2A /9/. Strength checks should be carried out for sev-

DET NORSKE VERITAS

Classification Notes - No. 30.9 June 2006

49

Basic design SN-curves are generally given as

support and the primary/secondary membranes.

log N = log a m log


where N is number of cycles to failure at the stress range , and m and a are constants. SN-curves for specific materials and details are obtained from fatigue tests, and are typically based on the mean-minus-twostandard-deviation curves for relevant experimental data, meaning a 97.6% probability of survival. For the tubular joints of the main structure of the pump tower, the SN-curve should be taken as the two-slope T-curve for air given in /7/: N 10 7 : log a = 12.164 , m = 3 N > 10 7 : log a = 15.606 , m = 5 This curve is based on a reference thickness of 32 mm. For other thicknesses, a thickness modification should be applied, as specified in /7/. It should be noted that the T-curve assumes full penetration welding. For simplicity, a one-slope curve may conservatively be used, with log a = 12.164 and m = 3 for the entire stress range. The fatigue damage for each load condition can then be expressed as

8.7 Vibration check


Modal analysis of the pump tower should be carried out, in order to check that the natural frequencies of the pump tower are outside the resonance intervals of excitations coming from the propeller blade and the engine room. The same model as used for the response analysis may be applied. The vibration check should be carried out for ballast condition, 10% filling, 70% filling and 98% filling. The margin between the excitation frequencies and the resonance frequencies is to be at least 10%. All relevant filling levels must be accounted for in the vibration analysis. This implies that both mass of liquid inside pipes and added mass effects are to be considered in the analysis. If the vessel is intended for ice-going operation, the excitation frequencies and forces from breaking the ice should also be considered.

9. References
/1/ International Maritime Organization (IMO), International Code for the Construction and Equipment of Ships Carrying Liquefied Gases in Bulk, IMO ,1993. /2/ Det Norske Veritas (DNV), Rules for Classification of Ships Pt.5 Ch.5, Liquefied Gas Carriers. /3/ Det Norske Veritas (DNV), Classification Notes No. 30.5, Environmental conditions and environmental loads. /4/ IACS Recommendation 34, http://www.iacs.org.uk/_pdf/ Rec34.pdf /5/ Pastoor, W., stvold, T.K., Byklum, E., Valsgrd, S. (2005) Sloshing in LNG carriers for new designs, new trades and new operations, GASTECH2005 conference and exhibition, Bilbao. /6/ Det Norske Veritas (DNV), Classification Notes No.30.7, Fatigue Assessment of Ship Structures. /7/ Det Norske Veritas (DNV), Recommended Practice DNVRP-C203, Fatigue Strength of Offshore Steel Structures. /8/ Det Norske Veritas (DNV), Classification Notes No.30.1, Buckling Strength Analysis of Bars and Frames, and Spherical Shells. /9/ American Petroleum Institute, Recommended Practice APIRP-2A-LRFD, Recommended Practice for Planning, Designing and Constructing Fixed Offshore Platforms Load and Resistance Factor Design. /10/ Efthymiou, M: Development of SCF Formulae and Generalised Influence Functions for use in Fatigue Analysis, Recent Developments in Tubular Joint Technology, OTJ88, October 1988, London /11/ NORSOK, N-004 Design of Steel Structures

D=

0 Td
a

m q (1 + ) h
m

where 0 is the long-term average response zero-crossing frequency, Td is the design life of the ship in seconds (= 1.26 108 for 40 years design life), and (1 + m/h) is the gamma function (equal to 6.0 for m = 3.0 and h = 1.0). Stress concentration factors (SCFs) should be applied according to recommendations given in Section 2.8 and Appendix 2 of DNV-RP-C203 /7/. The hot spot stress to be used in combination with the T-curve is calculated as

hotspot = SCF nom


The combined hot spot stress due to axial force and bending moments should be calculated for several points around the circumference of the intersection. Alternatively, the SCFs may be calculated using influence functions, according to Efthymiou /10/. For the base support and the liquid dome area, the appropriate SN-curve to be used for each detail should be taken according to the recommendations given in DNV-RP-C203, Appendix 1. Alternatively, the hot spot stress may be calculated by finite element analysis and combined with the D-curve, as explained in Section 2.13 of the RP. The procedure given in DNV Classification Note 30.7 may also be followed. Special attention should be given to the connection between the base support and the double bottom, and to the connection between the base

DET NORSKE VERITAS