Bonding of lightweight materials for cost effective production of high speed craft and passenger ships (BONDSHIP)

Constitutive laws to be used in numerical FE calculation can be directly obtained from experimental tests on material samples: some tests have been performed within the project, providing relevant data on different types of flexible adhesives. Experimental behaviour of elastomeric materials (traction, compression, shear), have been implemented in the constitutive laws of FE material models following several approaches, as for example Mooney and Ogden formulations. The effectiveness of the methods for material modelling, in linear and non linear hyper-elastic analyses, has been assessed by participating in different benchmark studies, proposed and faced during the project. Approaches followed by different partners of the project and corresponding analysis results have been compared each other and with experimental results, showing the truthfulness of the selected method of representing elastomeric materials in FE models. Effective models for elastomeric materials lead to interesting follow-ups in the FEM techniques used to analyse, for example, anti-vibration tools and equipments.

The behaviour of bonded structures subjected to cyclic loads has been investigated through a series of experimental fatigue tests performed at the CETENA laboratory at Riva Trigoso Shipyard owned by FINCANTIERI. Tests have been performed on two full-scale fairing samples (Fincantieri AC1). After fabrication, in January 2003, samples have been left exposed to weather conditions (sun, rain, salt environment), without any protection, for about six months. The achieved results can be considered very satisfactory from the point of view of the joint resistance and contribute to improved the knowledge on the fatigue resistance of bonded structures.

FiReCo have performed work in the BONDSHIP project related to design and analysis of bonded joints. The work carried out in cooperation with the project partners has documented possible applications for adhesive bonding in ship structures. The possibilities and limitations of different adhesive types with respect to structural strength have been verified by modelling and by testing. Different modelling techniques have been investigated, and the limitations of simplified models is determined and validated by direct analyses and complex material models. FiReCo has extensive knowledge and experience in analysis of complex composite structures and adhesive bonded joints and laminated joints in such structures. In combination with the existing expertise, the results and techniques investigated in the Bondship project will contribute to new solutions to the benefit of our future customers.

Superstructure modelling. The superstructure modelling has helped to investigate the general behaviour of the global superstructure and outline hot spots and load levels in specific bonded areas. This load response of the structure was performed according to DNV rules, thus defining a worst-case situation for the ship. Adhesive modelling approach. As the adhesive chosen by VT presented non-linear behaviour, it was proposed to use a non-linear finite element model to study the strength of the different joints. Several approaches available with the ANSYS package were considered but only one was sustained: a multi-linear approach. This method presented good agreement between experimental and numerical modelling for butt strap tests with 1-, 3- and 5-mm adhesive thickness, whereas for the 10-mm adhesive thickness, divergence occurred when load-displacement curves were compared. Stress-strain approach. This approach assumes the adhesive joint to be perfectly made, without any flaws. To assess the stress state in the adhesive layer, 3 approaches were considered: analytical method derived from classical theories, linear elastic finite element method and non-linear elastic finite element method. Analytical methods proved to be quick to implement, giving a first estimate of the stress state in the adhesive although not as accurate as the finite element formulations. In the case of structural joints, the finite element formulation has enable to predict stress concentration in the bondline and give input for fracture mechanics study as cracks are likely to propagate in such areas of high stress. Fracture mechanics / damage tolerance approach. The damage tolerance approach considers that a damaged structure is acceptable but to a certain level. Therefore, the investigation focused on the prediction of the maximum acceptable size of crack for a given load applied to the bonded structure. After analysis of DCB tests results, the latter were compared to numerical values obtained from FE models of structural joints with different crack sizes. It has been found that the fracture energy of the DCB tests was far beyond the fracture energy level reached in the numerical model of damaged structural joints. However, these results should be considered carefully since DCB tests were performed with a 1-mm adhesive thickness whereas the structural joints specification states a 3-mm adhesive thickness. Potential applications for these results are for the design of large bonded structures. End-users of these results would be naval architects, structural designers and owners who would be interested in implementing this technology in the construction of new units. The main innovative features of this work is a validated new approach for assessing adhesively bonded structures. Success factors will be a gain in weight in the structures, the possibility to fasten different materials and cost savings.

Starting from the reality of the shipyard, Fincantieri indicated, during the first part of the project, eight candidate application cases and, after this, it mainly focused its interest, for further study, on two of them: fairing on the cruise ship (AC1) and external casing (AC6). The first application case AC1is a fairing on decks for safety and comfort; the idea was to bond the aluminium alloy or composite GRP panels to steel bulwark stays, welded or bonded to the deck. The second application case AC6 is an external aluminium or composite casings to be bonded to a steel or aluminium superstructure in the upper deck. For each of the two selected Acs, a set of information were collected in order to have a complete scenario for the initial design of the join.; Possible joints solutions/configurations were designed on the basis of the following information: -The characteristic of the materials; -The evaluation of the self-weight and external loads; -The knowledge of the structural properties of the adhesive and the structural behaviour of basic simple joints. The details of the application cases were given to the experts within BONDSHIP, which made recommendations for the design of joint details. Additionally Sika in Zurich made screening tests with metallic structural materials and relevant "protective paints" usually adopted in FINCANTIERI. This was made, in order to determine suitable material/adhesive combinations and to select suitable cleaner and primer (surface pre-treatments). In this phase strength calculations, for each of the two ACs, by finite element analysis were made in order to verify the structural feasibility of the design solutions; it was verified to have acceptable levels of stress and strain obtained for all loading conditions for each AC. Considering also the benefits from both the economic and production points of view, the joint design was finalised. The main partners involved in this activities were Fincantieri, Cetena and SiKa. Result of these activities was an official Report: Deliverable Nr. 4-41-D-2001-01-0. What it was to investigate about the ACs was the production feasibility and how to realise them. In the middle of year 2, according to the recommendations of the partners, it was defined the production drawing/plan need to the completion of prototypes. In the same time the production procedures for Fincantieri ACs were defined; result of these activities was an official Report: Deliverable Nr. 4-42-D-2001-01-0. The fairing mock-up is constituted of three steel stay webs (Fe 510 /) height about 1 meter and two closed profile extruded panel. The casing mock-up is constituted of an aluminium deckhouse (Alloy 5083) of dimension 5504mm x 2100mm bonded on a steel deck portion (Fe 510) of 5904mm x 3900mm. The different phases followed during the assembly of casing and fairing mock up are contained in an official Report: Deliverable Nr. 4-41-D-2002-02-0. At the beginning of year 3 the prototypes were completely realised; the bonding was realised with support of Sika experts. Starting from this moment, several test on prototypes were performed (static,fatigue ,vibration and fire test) and some repair trials at AC6 were accomplished in order to be able to develop information and recommendations for users on the correct procedure to be used during the repair process. Official Reports: : Deliverable Nr. 4-42-D-2002-03-2: Test results from testing of prototype (Cetena); Deliverable Nr. 1-12-D-2002-02-0: Final report on fire testing (FiReCo); Deliverable Nr. 4-43-D-2002-02-0: Final repair guidelines (completed repair of prototype). At the end, on the basis of the experience done during the prototype assembling, It was evaluated the economical benefit in using bonded joint instead of other typical joint usually used in the yard. Document Nr. 2-23-W-2002-03-0: Cost comparison (Fincantieri).

BONDSHIP is involved in preparing a special issue of the JOURNAL OF ENGINEERING FOR THE MARITIME ENVIRONMENT - PART M with the aim to make adhesive bonding in shipbuilding accessible to a wider audience. A number of articles in the special issue will be written by project members. In addition, other recognized experts will contribute to this publication. Scope of special issue: This issue will discuss the use of adhesive bonding in shipbuilding. A key question in shipbuilding is the approval of such joints. A number of important questions are linked to this and will be presented in this issue by the various authors. Central to this topic are the possibilities and limits of predicting the joints strength and lifetime using simple or advanced numerical methods and the need to use prototype testing in addition to or instead of modelling. Another critical issue is the production of the joints. By and large shipbuilding is a steel based industry with a long tradition of welding. Hence there are special problems, which need to be addressed when introducing adhesive bonding into shipbuilding. Finally, examples of successful application of adhesive bonding in ships will be shown.

CTA has carried out a series of tests to evaluate rigid adhesives (2-component polyurethane, acrylic or epoxy adhesives). These tests have been chosen in order to select correctly the adhesive systems with the following criteria in mind: the strength of the adhesive assembly, the resistance of the assembly to temperature, its durability in a humid marine environment. Thermo-mechanical properties are useful to know the state of polymer matter (glassy or rubber-like state): the glass transition temperature (Tg) et the ratio of the shear modulus at 80°C divided by the shear modulus at 20°C are assessed at the initial condition (15 day storage at 20°C) and after a post-curing (624 hours at 70°C, 95% relative humidity). The single shear lap test is performed according to ISO 4587 with 4 different adherents (steel, alu5093, alu6082 and GRP): two parameters are deduced the mean shear strength of the assembly and the type of failure (cohesive, adhesive, composite). The wedge test, derived from the ASTM standard D3762-79, is aimed to screen different surface preparations of various adherents (steel, alu5093, alu6082 and GRP). Once adherents are identified, it is possible to select the appropriate adhesive system (e.g. an adhesive and a surface preparation).

The aim of WP5 was to develop design and production techniques as well as repair schemes for bonded joints in large passenger and cruise ships. MONTHS 1-6 (04/2000 - 10/200): During the first six months of the project, a total of ten application cases for bonded joints were defined and described in detail: AC1 load bearing secondary structures with large dimensions AC2 foundations on decks AC3 fixturing profiles for internal outfitting walls AC4 lightweight sandwich structures (in cooperation with the SANDWICH project) AC5 cargo lashing devices on RoRo-decks AC6 doors, windows and hatches AC7 gutterways on decks AC8 joints of large external structures into load-carrying primary structures AC9 new modular structures for balconies AC10 attachements below decks Based on existing mostly welded solutions, detailed requirement sheets were proposed by IFAM and filled by Meyer Werft and IFAM. These sheets are detailing loads, environmental conditions, production requirements etc. for all application cases. After defining the AC's, the principal design of a large prototype structure was started. This structure (sized 7.5m x 4.2m x 2.5m) was designed to be as close as possible to a real ship structure and to accommodate all application cases. The principal design of the prototype was discussed with the partners and the details of the application cases were given to the experts within BONDSHIP, which made recommendations for the design of joint details. The main partners involved in this activities were Meyer Werft, IFAM, R&R and SiKa. Results of joint designs and calculations have already shown, that a rethinking is required by the shipyard towards a Design for Bonding. This however could decrease the economic benefits obtained from using adhesive bonding techniques and will limit the applicability for onboard changes which is one of the main goals of the shipyard. Result of these activities was an official report (Deliverable): Deliverable Nr. 5-51-D-2001-01-0: Basic specification of application cases MONTHS 7-18 (10/2000? 09/2001): After the applications were defined the planning of the prototype began. The prototype consists of a loadbearing welded ship steel structure (L=7m; B=4,2m; H=2,5m), in which the applications for bonding are accommodated. We got with it both experience in production & repair of bonding components and knowledge regarding stiffness, fire safety, resistance to environmental conditions etc. by tests. Planning comprises also strength calculations for each of the 10 applications by finite element analysis (FEA) at the institute Ifam in Bremen. Additionally Ifam in Bremen and Sika in Zurich made screening tests with materials, which were used on Meyer ships. This was made, in order to determine suitable material/adhesive combinations and to assign these suitable surface pretreatments. Official report deliverable NR. 5-51-D-2001-01-0: Detailed Design OF Meyer throwing prototype MONTHS 19-23 (10/2001? 02/2002): In the middle of year 2 planning could be finished and it began the building of the prototype. Application case 4 in WP5 comprises a number of solutions based on the application of sandwich panels, which are being investigated in close cooperation with the SANDWICH project. A cooperation agreement between the two projects were obtained allowing both projects to use the results in their consortia. Tests have clearly shown the need to further improve the adhesive bonding techniques, the results of first strength tests within the SANDWICH project have been encouraging showing clear increases in the stiffness of sandwich panels due to the implemented filling materials in the range of 30% transverse to the stiffeners. The production of the prototype joints and improvements towards a high quality bonding process will be the main tasks for the second project year. Official report at DNV in Høvik: Deliverable NR. 5-52-D-2002-02-0: Prototype - JLM MONTHS 24-30 (03-09/2002) to the end Year 2 (03/02) production was finished and it began with the bonding of the components. For this preliminary manufacturing processes on basis of the screening tests were designed and the steps of bonding were stipulated. Deviating steps were taken up during the bonding and the bonding steps were accordingly corrected. The bonding was realized with support of Ifam and Sika. Planning for testing began from this moment. Testing of the components began in 08/02. Stiffness -, creeping- and load-tests were carried out. In addition repair trials at two applications were accomplished, in order to be able to develop repair guidelines. Official reports: Deliverable NR. 5-52-D-2001-01-0: First draft of production procedure Deliverable NR. 5-52-D-2002-01-0: Final production procedures Deliverable NR. 5-53-D-2002-01-0: Final repair guidelines.

The adhesive bonding process on the shipyard is much more challenging than in other areas of transportation industries. One reason for this is the unprotected work place because normally ships are assembled outdoor. And if the assembly is performed inside a hall it is also influenced from the outside temperature. And in both cases there is a remarkable amount of dust and dirt around. Another reason for difficulties are the big tolerances in the components. The big dimensions force this and the welding process, which twists the components, and there are also the tolerance problems of a single production. To produce components that are well suitable for bonding the workers must be more sensitive for precise dimensions. For the repair of bonded joints a good access to the joint is essential, this, from both sides. That is needed for the operation of the cutting tools, a special cutting wire or a trimming knife powered by compressed air or electricity. For the successful cutting the adhesive joint needs a minimum thickness of several millimeters. For the repair the adhesive application with big pumps or mixing systems is very often impossible, because they are not available. Therefore a simple to handle handgun system is needed, and if necessary also a different adhesive system. The learned lessons of the production and repair trials will be discussed at internal information meetings about the results of the BONDSHIP project. This includes training and promotion activities for the relevant Sika employees in technical service, marketing and sales. Also marine guidelines for bonding techniques with Sikaflex adhesives will be created. This is all for the support of our marketing and sales, and they also will use the learned experience for external conferences and publications.

Protection of Results: Meyer Werft will not develop entirely new solutions in the project, which could be protected by IPR. However, the shipyard will gain valuable experience which will be used to strengthen competitiveness on the world-wide market. Details about the application cases related to Meyer Werft will be confidentially kept within the BONDSHIP consortium. Functional Analysis of the main deliverables to identify applications: A main reason for Meyer Werft to participate in BONDSHIP is the acquisition of knowledge on the design, production and repair of bonded joints. All generic deliverables produced in WP 1 and 2, will therefore be carefully studied to retrieve the know-how necessary for practical applications of bonded joints in cruise vessels. Moreover, the experiences gained by other shipyards like VT (WP3) and FC (WP4) and described in the deliverables of the corresponding work packages will also help the shipyard in practical applications. The main focus of Meyer Werft is however on the deliverables of WP5. In this work package bonded joints will be designed, produced, tested and repaired for a total of ten application cases, defined by Meyer Werft. Those application cases have been selected with the clear intention to directly apply them on the products produced by the company, if the project results are satisfactory. Targeted audience: Meyer Werft is more a recipient of information in the BONDSHIP project, than a developer. Experiences gained in the prototype production will be shared with other consortium partners. It is not foreseen to disseminate those experiences to external parties. Exploitation and dissemination strategy, plan of action, resources: Application of Bonding Techniques to new application cases: In addition to the prototype production, where the application cases defined was tested under practical conditions, the competence centre has planned and prepared a repair case. This case is related to buckles appearing in the sun deck of a cruise ship delivered some years ago, which led to a separation of ceramic tiles from the deck. Additional stiffeners were to be placed under the deck to retain flatness and thus repair the defects. Welding could not be applied to avoid thermal destruction of the tiles, coated surfaces and equipment installed. Tests have been carried out and a combination of stud welding and adhesive bonding have been selected to install the additional stiffeners. The personnel who will carry out the repair have been trained and repair guidelines have been prepared. Approval of bonded joints: Approval by a classification society is a precondition for practical applications of adhesive bonding. Based on the application cases defined in the project and on the experiences which were gained on the builtprototype , Meyer Werft will apply for approval of the most promising cases by the end of the project or after its termination. Application of bonded joints in cruise ship construction: It is foreseen to practically use the successful application cases as soon as they have been qualified. It is expected, that successful applications can first be used one year after the termination of the project.

A consistent procedure has been developed to design adhesive joints for ship building applications. At the beginning of a design process lists of requirements for the specific application case have to be worked out by the shipyards. These lists have to be filled out with a specific set of main characteristic features which yield full information about any aspect of the specific joint, if the lists will be carefully worked out. Finally the lists of requirements might provide a complete technical description of the specific application case. With this method the shipyard is provided with a methodology to find a safe approach to a new technology. The list of requirements is the basis of any further work from testing to design, production and manufacturing etc. According to the requirements of the application cases screening tests have to be defined and to be done in order to select adhesives, adhesive primers and surface preparations. One severe influence to the screening tests and to the enormous amount of tests, which might be necessary, is determined by requirement to bond on painted surfaces. In parallel to the first working steps in the design process of the joints a test programme has to be worked out to determine necessary data for numerical and analytical design approaches with consideration of possible effects of ageing and fatigue. The types of tests are determined by the requirements of software for finite element analysis on one hand side and by the concept of knockdown and safety factors to be used for the design process of adhesive joints on the other hand side. This concept needs data about the strength of the adhesive joint before and after ageing, fatigue, temperature or any other influence to determine specific knockdown factors. The data have to be determined for any surface of paint, which should be used. It turned out, for example, that the fatigue behaviour of rigid joints is significantly influenced by the surface, especially if the surface is painted or not. Besides the data for the work-horse adhesives and selected surfaces which can be obtained from the project partners the main result is an experimental methodology to determine characteristic values and knock-down and safety factors, as well as mechanical data for finite element analysis. The experimental method will be one backbone of the guidelines. For the modelling of joints different approaches have to be used for flexible and for rigid adhesives. For flexible adhesives the "energy approach" with hyperelastic non-linear material law was generalised in order to be able to optimise any local joint geometry. The hyperelastic analysis was compared with global and local linear elastic analysis. While the global results agree within a tolerable error range, the local analysis shows significant differences. Since failure occurs locally, non-linear analysis is recommended for the prediction of the locus of failure and failure prediction as well as for geometrical optimisation. Furthermore, a pragmatic linear elastic approach for strength prediction was developed for joints with rigid adhesives. The approach uses a failure criterion developed from coupon testing, including knockdown and safety factors. The method is conservative and was verified by component tests for bondship application cases. The design methods for flexible and for rigid adhesive joints will also be parts of recommendations in the guidelines provided by DNV. The methods can be directly used at current shipyard conditions. The education programme for the European Adhesive Engineer will use the results of the project for further spreading the information to other application areas. The implementation of the results to the specific sites needs some further effort, which can be supplied by IFAM within further co-operations and partnerships.

In the screening tests 5 adhesives on 19 surfaces with different pretreatment methods were tested, in respect to the demands from the shipyards. For the most applications a practical solution was developed. The flexible Sikaflex adhesives can bond on the most surfaces with the correct SikaPrimer. But the substrate needs always a corrosion protection; the bonding directly on raw metal is poor, especially for steel. Always very good results were achieved with the two component shop primers that are normal suitable at shipyards. The best way for qualification a flexible bonding system and correct surface preparation is the bead adhesion peel test. The results of this were also supported by lap shear tests, and later on with the ageing and fatigue tests. In the ageing and fatigue testing single overlap shear joints with 3mm bond thickness were tested with an alternating cyclic loading by a stress ratio of R = -1 with a test frequency of 25Hz. This results a negative minimum stress with the same amount as the maximum stress, and the mean stress is zero. The results show that the fatigue durability is above 0.2MPa, and below 0.45MPa for Sikaflex-552, respective below 0.25MPa for Sikaflex-292. After two million cycles with 0.2MPa no weariness was detectable and the measured compliance was all the time constant. There was also no influence of the 1600 hours salt spray ageing detectable, compared to the not aged samples. If some corrosion occurs, it happens between the Primer and the aluminium (Anticorodal-110 EN AW-6082 T6), the adhesive was not influenced. The samples loaded with 2 million cycles at 0.2MPa and 1600h salt spray were ruptured in the tensile lap shear test and compared to not aged samples. The measured change of mechanical properties was very small. The test results will be used for further consultation in the transportation industry. Most of the examined surfaces are suitable in many other industrial applications, with exception of the two-component shop primer, which is typically only used in shipbuilding. The tested loading and ageing conditions are also very common.

Several bonded solutions has been already used on passenger ships and fast ferries with good results in terms of performances and aesthetics. Others solutions can be adopted to increase the quality standards in designing and building ships; that s why Fincantieri participate in Bondship project. Fincantieri engaged itself principally in WP4. In this work package Fincantieri designed, produced, tested and repaired two of the eight application cases proposed by Fincantieri itself; for each of them a cost benefit was also evaluate and for the AC6 (the casing bonded to a steel superstructure) the results seem quite good. At least two new bonded solutions are available to be introduced on the products. Moreover the work directly done in WP4 and the know-how retrievable from the work performed in the project leaves to Fincantieri the guidelines on how to approach new bonded solution, starting from the design to the production. The deliverables produced in WP1 and WP2 can be useful to perform a first and preliminary design of bonded joints need to realise new application cases. The experiences done by others shipyard (JLM and VT) in developing their application cases will be useful for further applications.

From the design of an aluminium-framed superstructure presented by VT Shipyards as an application case for Bondship, a series of adhesively bonded joints were tested to assess their strength and failure mode in a marine environment. The test programme included static tests on standard single lap shear specimens and butt strap joints with different adherends corresponding the materials used in the application case (Al 5083, Al 6082 and steel) and structural joints made of extruded box sections: deck to superstructure and unit-to-unit joints. Also, fracture mechanics tests have been carried out to assess fracture energy of the adhesive system through Double Cantilever Beam specimens. The aim of the lap shear test programme was to assess the system adhesive / primer / surface preparation in two different conditions, namely: non-aged in laboratory conditions and environmentally degraded conditions humid and hot. Strength reduction from 8 to 28 % was observed depending on the type of adherend after ageing process. Due to the design of the VT application case, based on butt strap joints, and the low production tolerance outlined by the shipyard, influence of the adhesive thickness on the strength and mode of failure of such joints were investigated. Four different adhesive thicknesses were considered: 1-, 3-, 5- and 10mm. Also, influence of three different ageing conditions was investigated: ideal laboratory conditions, 3 and 6 weeks in a 100% rh at 40°C. Tests showed an almost linear strength reduction between a 1-mm adhesive joint and a 10-mm one. Strength reduction is more significant with respect to the adhesive variation than to the ageing process. It has been outlined the importance of surface preparation and adhesive application process specially in the case of aged specimen where a good bondline ensures degradation due to infiltration of water. Experimental modelling was carried out on structural joints to assess their tensile behaviour and flexural behaviour in a 4-point bending configuration. Deck to superstructure joints were shown to be fairly strong in tension whereas when tested in flexion, they were shown to have a similar strength to aluminium joints. These tests, together with butt strap and lap shear tests, have highlighted the fact that adhesion to steel presented a weaker bond than adhesion to aluminium, suggesting a more careful surface preparation concerning this adherend. Fatigue tests carried out to assess flexural behaviour have shown discrepancy in the results concerning deck to superstructure joints whereas unit-to-unit joints showed failure in the aluminium weld. A fracture mechanics approach has been proposed in order to evaluate damage tolerance of large bonded structures. To initiate this assessment, Double Cantilever Beam (DCB) tests have been carried out to estimate the fracture energy of the adhesive system. DCB with different adherends and ageing conditions have been studied and fracture toughness derived with different analytical methods. Results were shown to be fairly consistent although a slight difference exists depending on the type of adherend used for the DCB test. All failure occurred cohesively except in few cases after environmental degradation process. Potential applications for these results are for the design and construction of large bonded structures in shipbuilding. End-users of these results would be shipbuilders, repair yards and owners who would be interrested in implementing this technology in the construction of new units. The main innovative features of this work is a validated new approach for assessing adhesively bonded structures. Success factors will be a gain in weight in the structures, the possibility to fasten different materials and cost savings.

Guidelines and procedures: The objectives of this task were: to define guidelines and procedures for quality assurance in production, repair and inspection of joints, design and modelling of joints and acceptance tests. The following fields were to be covered in the guidelines work: - Quality assurance in production; - Repair and inspection of joints; - Design and modelling of joints; - Acceptance criteria and qualification tests for joints. The role of NDT in this task was to concentrate on the quality assurance and repair and inspection of joints. As part of the guidelines work European NDT standards that currently apply were reviewed; this included common inspection techniques and operator certification. The differences between manufacturing and in-service requirements were considered and recommended practices for the role of NDT were investigated. Extensive work was carried out at two shipyards (Vosper Thorneycroft and MeyerWerft) to gain practical experience of performing NDT inspection on marine structures in a manufacturing environment. An example inspection procedure for one of the bonded structures on the MeyerWerft application case was developed and documented. The principle results and findings from this work were: - Details of all relevant inspection technologies applicable for adhesive joint inspection. - Recommended inspection practices and operator certification standards for NDT of adhesive joints. - Example inspection procedures and documentation. - Practical experience of applying NDT in a shipyard environment. - Demonstration of NDT to support quality assurance during manufacture (Vosper Thornycroft) and support of repair trials (MeyerWerft).

The contribution to the selection, design and construction of the FINCANTIERI AC1 fairings and AC6 casings bonded prototypes improved the know-how of Cetena in this field. Several FEM calculations have been performed on those structures, to improve their design and to evaluate the stress levels both in structure and in the joint under the selected loading conditions. The possibility to model a bonded joints within a global coarse FE model has been investigated. In 4-41-D-2001-01-0 Preliminary design of prototypes Fincantieri AC6 (external casings); the joint between the deck and the casing has been modelled within a complete 3D FE model of a ship, and analysed in the context of a global bending deflection of the hull. Detailed fine models of both ACs have then been built and analysed, focusing the attention on local effects arising from global and local loads on bonded joints. Different modelling approaches have been experimented investigating the effects of: - Different model refinements, - Different element types: 8 node brick, 20 node brick, - Different formulations: compressible and incompressible-Hermann formulation, - Different solution procedures: total Lagrangian, updated Lagrangian, - Different material formulations, Mooney, Ogden. Leading to a knowledge improvement about the behaviour of bonded structures and about modelling and analysis methods. Experience acquired within the project allows an accurate strength evaluation of bonded joints, by using finite element approach.

The guidelines sum up the collective know-how and experience of the BONDSHIP project partners. It is the first time that such comprehensive guidelines for adhesive bonding have been produced for marine application. The guidelines apply to all types of adhesively bonded joints in Ships. The guideline document is split into two parts: - Code (DNV Report No. 2004-0134): The objective of this document is to provide general requirements to ensure the reliability and safety of load-carrying bonded joints in ships. - Recommended Practices (DNV Report No. 2004-0193): The objective of this document is to provide guidance and examples on how to design, produce and inspect an adhesively bonded joint. Furthermore it shall provide the basis for meeting the general requirements laid out in the Code document. BONDSHIP GUIDELINES - RECOMMENDED PRACTICES: This presents methods and actual examples (including data where possible) for the design, production and in-service phase of an adhesively bonded joint. The following topics are discussed: - Specification of bonded joints; - Materials selection; - Failure criteria and characteristic strength values; - Design and analysis of bonded joints; - Testing of materials and structures; - Fire safety; - Production and repair of bonded joints; - Non-destructive inspection. Furthermore, procedures which are not easily found in the literature are presented in an appendix. The intention of this report is to enable naval architects and other suitably qualified engineers to design and produce safe and reliable adhesively bonded joints. BONDSHIP Guidelines CODE: This document provides general requirements to ensure the reliability and safety of load-carrying bonded joints in ships. This document applies to all types of adhesively bonded joints in ships. A novel approach for joint approval has been developed based on the following observations: Numerical analysis cannot reliably predict joint failure without additional large scale tests. While numerical analysis can give extremely useful insights into the behaviour of bonded joints, the approval of joint will only utilise representative tests as a cost-effective means of assessing bonded joints. Furthermore, the long-term performance of a bonded joint cannot be reliably predicted from the results of accelerated ageing tests without having relevant in-service experience to verify it. Therefore, requirements to the resistance of the joint are combined with requirements that limit the consequences of failure of the joint and that it must be possible to repair the joint using an approved repair method.

FiReCo have performed work in the BONDSHIP project related to design and fire testing of bonded joints. The work carried out in cooperation with the project partners has documented possible applications for adhesive bonding in ship structures. The possibilities and limitations of different adhesive types with respect to structural integrity during fire have been verified by testing. The application cases from each yard has been evaluated and a specific fire protection for each AC has been specified, built and tested. Test results shows the possibilities and limitation of adhesive bonding in ship structures. Possibilities and in some cases directly applicable solutions for fire protected adhesive bonded joints in ship structures has been designed and verified by testing. Results obtained may also be used for evaluation of adhesive bonded structures in general, as well as guidelines for further developments in adhesive bonded ship structures. FiReCo has extensive knowledge and experience in design and testing of complex composite structures and fire protection of such structures. In combination with the existing expertise, the results and techniques investigated in the Bondship project will contribute to new solutions to the benefit of our future customers.

This document contains the production procedure used during the construction of the Bondship demonstrator. This production procedure is based on the procedure outlined in document 3-32-D-2001-01-0 and work carried out in document Procedures for Bondship Test Specimen Manufacture. The procedure was also refined at the early stages of construction and the experiences leading to these changes have been included in this document. The bonding techniques used can be found in Tables 3.3.1 and 3.3.2. The demonstrator is constructed from aluminium box section, aluminium honeycomb panels and Plexus MA550 adhesive. This document includes the procedure used for the bonding of panels and box section to form the units that make up the demonstrator, bonding of units to form the superstructure and the superstructure to deck bond. Where deviations from this procedure have occurred during the build the reasons and new procedures have been included in this document. In addition suggestions on how the procedure, design or materials could be improved have been made. A build log for the demonstrator and a lessons learnt file form part of this document and can be found in Appendix A and B.

This benchmark study is to evaluate the capabilities of different modelling approaches (analytical, finite elements) used by the partners to predict the strength of a joint of simple geometry (a single lap joint). Modelling was undertaken by the University of Southampton and by CETENA. Independently, CTA conducted test samples to check the validity of the modelling. The key elements to keep in mind from this study are: the overall philosophy of dimensioning with rigid adhesives should be to design against yielding (required input data from the adhesive behaviour law can be restricted to the linear part), the non linear effects coming from the geometry of the joint must taken in account, with rigid adhesives systems, the yielding of the adherent is important (the adhesive is not always the weakest point in the assembly), adhesive materials are sensitive to pressure. The plasticity criterion (Raghava) must take into account the yield stress in tension and in compression, for a simple geometry joint, both analytical and finite element calculations have given equivalent results.

During the project the BONDSHIP NDT activity was disseminated at the Annual Conference of the British Institute of Non-destructive testing in Sept 2001 and Sept 2002. Synergy between BONDSHIP and a UK DTI project (UWASI) led NDT Solutions was developed. The BONDSHIP project provided many test assemblies for a new inspection system developed in the UWASI project. This synergy has led to the marketing of NDT Solutions ultrasonic equipment for the inspection of bonded composite components in the marine and power (wind turbine) industries. Experience gained and techniques developed during the BONDSHIP project have given NDT Solutions a competitive edge for the inspection of marine components enabling it to explore the marine and related markets. Experience gained in the BONDSHIP project has enabled NDT Solutions to successfully approach UK materials suppliers, manufacturers and operators for high performance carbon fibre composite racing yachts to demonstrate NDT technology for the inspection of bonded joints and carbon fibre pre-pregs.

The objectives of this task were: to extend existing ultrasonic NDT techniques for inspection of adhesively bonded joints during production (QA) and in-service testing for shipbuilding. Results summary: - Acceptance and qualification of joints: During the materials selection NDT was used to characterise the ultrasonic properties of the adhesive materials in order to investigate the inspectability of joints formed with these materials. The difference in ultrasonic properties between rigid and flexible adhesive was investigated and this information was used to assess the testability of new joint designs. - Inpsection of prototype structures: During this task NDT was used to inspect the completed prototype and to assess the quality of the bonding. New inspection techniques such as phased array ultrasound and standard NDT equipment was used to inspect a range of bonded structures developed by the BONDSHIP partners. For example, production butt strap joints for bonded steel and aluminium structures were inspected during manufacture and delaminations in the joint were detected. This is a signficant result since the delaminated areas could act as corrosion sites leading to degradation of the joint performance during the service life of the structure. The principle results and findings from sections 1 and 2 of this programme of work were: -- Characterisation of adhesives used for shipbuilding -- Development of simulation tools to predict ultrasonic testability of bonded structures -- Application of current and state-of-the-art ultrasonic NDT for the inspection of production prototypes -- Experience in the development and deployment of new inspection methods (low frequency techniques, and large area rapid scanning) for large bonded structures - Ultrasonic NDT to assess performance of joints and critical defects NDT Solutions worked with the University of Southampton and Vosper Thornycroft to support their work on joint performance and crictical defects. This included pre-screening of coupons before mechanical testing and monitoring of joint during testing. NDT was also used to investigate test joints with known defects to assess detection levels. For example, ultrasound was used to measure the change in thickness (displacement) of the adhesive bondload in a joint under load. Vosper Thornycroft manufactured a test joint with a range of defects that could be encountered during production or in-service. They aim of this work was to assess the sensitivity of the ultrasound inspection techniques to these types of defects. This joint was successfully scanned demonstrating that ultrasonic images of the joint could be obtained: The principle results and findings from this section of work were: -- Demonstrated use of NDT for pre-screening of test coupons -- Successful detection of displacement stresses in adhesive joints during mechanical testing -- Use of novel ultrasound imaging technology for the detection of simulated production and in-service defects - Use and exploitation of results Synergy between BONDSHIP and a UK DTI project (UWASI) led by NDT Solutions was developed. The BONDSHIP project provided many test assemblies for a new inspection system developed by NDT Solutions. The BONDSHIP results has helped NDT Solutions successfully market new ultrasonic equipment and test techniques for the inspection of bonded composite components in the marine and power (wind turbine) industries. This has given NDT Solutions a competitive edge in the inspection of marine components enabling it to further explore the marine and related markets. Experience gained in the BONDSHIP project has enabled NDT Solutions to successfully approach UK materials suppliers, manufacturers and operators for high performance carbon fibre composite marine structures to demonstrate NDT technologies for production QA and in-service testing.

The application of adhesive joints has been proven to be beneficial to lightweight construction in mass transportation by using hybrid concept, which allows for cold assembly and modular design. In the BONDSHIP project adhesive bonding has been introduced as structural joints into shipbuilding for joining of lightweight materials. Design process for structure with bonded joint consists of 3 major steps: Concept study, detailed design and final design step. In the concept study hand calculation and analytical analysis is necessary for the first joint dimensioning. In the detailed analysis design parameters such as adherent/adhesive materials, surface preparation, joint geometry, tolerance concept, securing mechanical system and UV-Protection have to be defined. General design rules with regard to joint configuration, geometrical and environmental factors, corrosion proctions as well as calculation base has been given in document No. 1-11-D-2001-01-0, "Outline of approach to easy-to-use design rules". Design method and design process has been discussed and documented in document No. 2-23-D-2001-01-0 "Outline of design approach".

A baseline design of a bonded aluminium superstructure is presented; it is designed in accordance with DNV Rules for Classification of High Speed Light Craft and Naval Surface Craft. The build standard incorporates aluminium alloy honeycomb cladding. The structure consists generally of a welded aluminium alloy extruded box section framework; the cladding being bonded in situ. Consideration has been given to unit construction as would be the case for a large superstructure and these unit joints have been designed in The superstructure to hull attachment, mast to superstructure and unit to unit joints are addressed. Structural design calculations and joint structural arrangements are contained in Appendices. Further work is discussed, including other cladding materials. A Macro FE analysis of the superstructure is being undertaken by the university of Southampton, using DNV loadings to analyse the structural framework as designed in order to compare the actual analysis with DNV requirements. And from a better understanding of joint loadings to consider the possibility of an all or part bonded framework.

Establishing bonding know-how with shipyard personnel: To ensure, that the results of BONDSHIP can be efficiently used in the everyday business of Meyer Werft one additional engineer was employed, exclusively in charge for adhesive bonding applications. He was implemented in the work for BONDSHIP and receive an external training as a bonding specialist. Furthermore some workers of the shipyard were trained in adhesive bonding techniques during the production of the prototype by external experts. Internal Competence Centre Bonding Techniques: a new employee in the R&D department was in charge to collect pre-existing bonding know-how of the shipyard (eg. windows in cruise ships) and know-how gained from the BONDSHIP project. He was also in charge to coordinate the production of a prototype. This prototype will be produced under close to reality conditions by shipyard personnel. Thus, an internal competence centre on bonding techniques was created which was in charge to evaluate new applications.

In order to establish repair procedures for bonded superstructure joints, two contrasting rehabilitation techniques were investigated on representative joins. These consisted of reinstating a unit to unit join without damaging the surrounding framework; and reinforcing a s/structure to ship joint. The techniques were used on representative joints constructed in the lab, with the techniques being extended to a number of other joint geometries. Repair trials were not carried out on the s/structure due to space and time restrictions. The reinforcing technique involved using steel and carbon fibre plates to add strength, as well as a laminated carbon fibre repair. The joint reinstating technique could be implemented, but would prove costly and time consuming due to difficulties in removing old adhesive and reapplying new adhesive inside the joints. However the reinforcing technique was relatively straightforward with the carbon fibre laminated repair being the most versatile. Further work would involve the mechanical testing of these joints and establishing a range of procedures for other joint geometries.

In WP3 (VT), WP4 (CET) and WP5 (JLM) bonded structures for superstructure of ship have been designed, structural analyses have been performed and prototypes have been produced and tested. For CETENA fairing application case (AC1) flexible adhesives are chosen to join an aluminium fairing to a steel bulwark stay as an alternative to welded or bolted connection. For the joining of the aluminium extrusions the Alcan turn spring system developed in truck and bus constructions has been introduced. Compared with welding, the cold assembly with turn spring system and bonding reduced the overall production cost and provided a better surface as well. At the first design stage easy-to-use rules are invoked for the feasibility study of structural and joint design. Some examples of the application of analytical rules for the joint analysis are found in for JLM AC 1, 6, 7 and 8. For the pre-design of VT AC the loading of joints was analysed analytically. Also for the preliminary design of CETENA AC6 Casing and AC1 Fairing some analytical methods have been applied. The final design step assessed by component analysis and testing has the goal to prove the structural integration. A global FE analysis for CETENA AC6 Casing for prescribed loading cases. For the final design of CETENA AC1 Fairing the FE analysis has been performed. Finite element analyses have also been carried out to estimate the joint performance of VT application case. An overview of documents related to the design and structural assessment of application cases is given in document no. 1-14-D-2002-01-1 "Novel Joint Design".

The application of flexible adhesives to join structural elements with similar or dissimilar materials is becoming more and more common in the last years. The finite element method provides versatile numerical tools for the investigation of the behaviour of the bonded joints under operational loading conditions. Non linear finite element methods with hyperelastic material model are required for detailed joint analysis. For the design of large structures a cost efficient modelling method is necessary: for the modelling of adhesive layer in large structures, a substitute system of spring elements can be applied. In order to perform a global finite element analysis of a large bonded structure, a substitute system of spring elements has been developed to take into account the stiffness contribution of the adhesive layer. Three translational spring elements, connecting correspondent nodes of the adherent, constitute a discrete substitute model of the continuum adhesive in the global FE model. The stiffness moduli for each spring have been evaluated based on the test results performed on butt joint and lap shear joint. Assigning variable values to stiffness moduli, the complex dependency and the nonlinearity of the adhesive material on the loading rate, temperature, load type and cross section geometry can be directly accounted for. The analysis with the hyperelastic material model enables refined detections of local stress and strain gradients in the adhesive layer. Nonlinear analysis can be performed. On one side, this method requires minimum assumption, but on the other side large models and long computational time are required. The modelling method with a substitute system of spring elements allows for a reasonable element size of the adherents even with small joint thickness. Linear or nonlinear analysis can be performed. The deformations and forces across the joint are calculated. Numerical simulation of a benchmark test on a bond-ed structure with flexible adhesive has been performed. The adhesive layer is represented either by so-lid elements characterised by hyperelastic material, or by spring elements with equivalent joint stiffness. The analysis re-sults have been compared with the first loading curve of the mechanical test. Both modelling methods enable a sufficiently accurate prediction of the overall stiffness behaviour of the bonded structures, if compared with the test results. The stress evaluation of the adhesive is dependent upon the modelling method: the analysis with solid elements model provides local stress and strain. With spring elements model the nominal stress and strain is derived, which can be compared with permissible stress and strain provided by adhesive supplier directly. The correlation between local and nominal stress can be established by a sub-modelling.

Livrables

Résultats exploitables

Partager cette page

Télécharger