International Journal of Small Craft Technology RESEARCH INTO SECONDARY BONDING OF YACHT BULKHEADS WITH EMPHASIS ON DEFECTS AND METHODS OF IMPROVEMENT K J Duggan, Falmouth Marine School/University of Plymouth, UK SUMMARY The scope of this article is to introduce and explain my research and findings into the process of secondary bonding of yacht bulkheads. The case study was a newly laid up Rustler 42 designed by Stephen Jones and built by Rustler Yachts who are based in Falmouth with their factory located next to ponsharden boatyard. Research came from a variety of sources including books from the Marine School’s library dealing with composite construction, manufacture’s websites, on-line videos and other various mediums. The practical work was done on-site at the Marine School’s composite laboratory with is located behind Rustler’s factory hangers which meant easy access to materials as well as information and practical advice from Rustler’s experienced workforce. Test pieces were constructed over the Easter period in roughly two weeks all built to the same standards and specifications in three groups of three to allow for continuity and comparison of results. NOMENCLATURE T - Laminate Thickness W - Fibre Density SGF - Specific Gravity Fibre SGR - Specific Gravity Resin U - Void Content WF - Fibre Weight Fraction 1.
INTRODUCTION
This journal article was designed to be a continuation of a previous assignment designated from the same degree module which was a Literature Review which was a 3000 word essay outlining current research and writings in the field which could be used as research for this project. The literature review that was conducted originally focused on the growing development of bioresins and bio-adhesives which were in application in many forms of industry ranging from engineering to medical applications. The research taken from the literature review could then be used to develop ideas about how to conduct tests and experiments to demonstrate findings and results. The focus of this project from the start was on secondary bonds used in the marine composite industry and the growing need to develop alternatives to petro-chemical resins which still dominate the industry because of their excellent mechanical and ease of building properties. When it came time to start building the test laminates in the school’s composite laboratory it became apparent that testing bio-resin’s abilities as a bonding applicator was an apparent dead-end concept. Areas of difficulty which are associated with bio-resins are its ability to bond certain areas where there is enough exposure for ultra violet light to penetrate the laminate and cure the resin. This presents problems in the fact that secondary bonding requires contact with two surfaces meaning that ultra violet penetration of the laminate would be next to impossible compared to a petro-chemical resin such as
polyester which uses a catalyst to start a thermo-reaction to cure the bonding surfaces. The sectors of the marine industry that use bio-resins are mainly dingy builders and small manufacturers of surf boards thus making bio-resin application a very niche market. This change in focus led to changes across the board thus changing the outcomes and objectives of the original project idea. The journal that was originally proposed to publish the findings had to be changed as it was a journal which focused on polymer sciences. With the focus shifting away from bioadhesives and their potential application in the marine composite industry the area of interest focused on the defects caused by secondary bonding. This project was also conducted in conjunction with local Cornish industry to act as a benefactor to the local industry and economy of not only Falmouth but the County of Cornwall. The aid of the local yacht builders of Falmouth in this case Rustler would be beneficial to both parties as it would allow useful guidance courtesy of Rustler, but most importantly the final results from the experiments could be hugely beneficial in highlighting potential areas of concern within Rustler’s build mythology and a means of improving not only the final product but decreasing their build time. The choice of case study happened almost by accident, on one visit to Rustler to speak with certain heads of department and management the composite team had just finished hand laying up their mould for a new Rustler 42, this included the instillation of the internal longitudinal stiffeners which run from aft of the transom to the bow stem and the flow coat starting from the bow. The Rustler 42 is moulded in accordance with Stephen Jones construction drawings. The lay-up is comprised of isophthalic gel coat (double gel), colour white, with complexes of chopped strand matt and woven rovings, hand laid with iso-phthalic resin on the first two layers. Bulkheads are of 18mm marine grade plywood bonded to
the hull and deck and longitudinal stiffening stringers are foam cored. The hand laid up system, using glass fibre re-inforced polyester (GRP) is to Lloyds Register specifications and uses Lloyds approved materials. Once the yacht is completed, Rustler’s overview of the 42 is that of a classic looking yacht with a moderate height to the topsides, a lovely sheer line, pleasing overhangs and a long, wide and low cabin top. The Rustler 42 adds a performance element to other essential blue water attributes like directional stability, stiffness and good load carrying ability. Below the waterline she looks conservative, with a deep canoe body, long fin keel and a big skeg hung rudder. The Rustler 42 combines style that is traditional yet modern, with a proven cruising layout, resulting in a live aboard yacht that has stability and elegance with the same unique sea-kindly characteristics as the Rustler 36. 2.
to help reduce the amount of heat produced from the thermo-reaction caused by the catalyst when curing the polyester resin and to help reduce overall shrinkage of the hull and post-curing when left in the mould for a set amount of time. After some extensive discussions with Richard Garland regarding the build methodology of the Rustler 42 as well as other vessels in the Rustler fleet the key area of interest focused on what they use as gap filler when installing the bulkheads in any vessel. A small strip of foam shaped to a rectangle which measures 15mm in width and changes in thickness depending on the size of the marine ply used to shape the bulkheads. The foam strip is used as a gap filler in-between where the bulkheads hug the longitudinal stringers which means that the bulkheads are never glassed in hard against the hull because shape corners of any material act as stress inducers.
METHODOLOGY
The methods employed in the construction of the test laminates incorporated information obtained from Rustler regarding the construction and lay-up schedule used in the building of the Rustler 42 and Hugo du Plessis’ 3rd edition revised Fibreglass Boats (2002). His book is relied upon by yacht owners, surveyors and yacht builders who are keen to understand how fibreglass behaves. Hugo du Plessis himself has spent a lifetime working with fibreglass, five decades ago he was one of the pioneers in the industry, starting out as a moulder and boat builder; later in his career he became a yacht surveyor specialising in fibreglass gaining experience in identifying the common defects as well as the elusive one.
Hugo du Plessis states that bulkheads or any other type of partitions must not be a tight fit against the inside of the hull as it will cause distortion as well as a hard spot. Therefore there must be a gap which is then filled so it can be moulded over and avoid a line of weakness. This will spread the stress so the bulkheads blend into the hull. Other types of gap fillers are also mentioned which include using resin putty, trapezoidal sections of plastics foam or wood, but for a production vessel which is on a strict build time and constantly trying to improve on build time materials such as putty and wood would be unsuitable, where as foam can be quickly crafted into different shapes in small or large sections and moulded to fit the curvature of the hull for a specific job make it an ideal material.
Gaining information regarding the lay-up schedule from Rustler regarding the 42 was a fairly simple task, Richard Garland; Rustler’s head laminator was more than obliged to disclose the necessary information required to build the test laminates. Because the case study was the newly laid-up Rustler 42 the test laminates should be built to exactly the same standards and specifications as the 42 in order to make these tests as accurate and close to the real thing as possible.
In a ideal world the test laminates would be crafted as smaller but to scale cross sections of the Rustler 42 hull complete with a section of bulkhead where they would meet as specific points, but because of the limited resources available at the time of preparation the laminates were prepared on small 50 square centimetre plates of glass which in principle would represent a small sectional area of the 42’s freeboard located atop of the sheer line.
The test laminates were engineered using the same laminate schedule as the Rustler 42 thus creating a monolithic panel which would finish to roughly the same thickness about 0.8mm which included the layers of gelcoat applied beforehand. The lay-up schedule consisted of four layers of 850 quad matt which is designed to give hull strength in four different directions at 0/90 degrees and +/- 45 degrees. The quad matt also has a layer of 225 chop strand matt sown into the weave which is required by law to prevent two shinny surfaces from touching thus reducing the risk of surface movement within the laminate which could lead to de-lamination. Because of the large surface area of the Rustler 42 the laminators have to apply the four layers of fibreglass in sets of two
The laboratory work was aided by Falmouth Marine School’s resident lab technician Simon Combe who also does part time work for Rustler who offered valuable input and direction into the building process. The test laminates were constructed using the same hand lay-up techniques which are employed on most Rustler vessels. 2.1
WET LAY-UP/HAND LAY-UP PROCESS
The hand lay-up process involves impregnating the glass fibres by hand which can take the form of woven, knitted, stitched or bonded fabrics. The consolidation of the resin is usually accomplished with rollers and brushes with an increasing use of nip-rollers which forces the
resin into the fabrics and the same time removes trapped air pockets for a better consolidated laminate and reduced void content. 2.1(a)
2.1(b)
Main Advantages: Widely used for many years. Simple principles to teach. Low cost tooling if room-temperature cure resins are used. Wide choice of suppliers and material types. Higher fibre contents and longer fibres than with spray lay-up. Main Disadvantages: Resin mixing, laminate resin contents and laminate quality are dependent on the skills of the laminators. Low resin content laminates cannot usually be achieved without the incorporation of excessive quantities of voids. Health and safety considerations of resins. The lower molecular weights of hand lay-up resins generally means that they have the potential to be harmful than higher molecular weight products. The lower viscosity of the resins also means that they have as increased tendency to penetrate clothing as well. Limiting airborne styrene concentrations to legislate levels from polyesters, vinylesters is becoming increasingly hard without expensive extraction systems. Resins need to be low in viscosity to be workable by hand. This generally compromises their mechanical/thermal properties due to the need for high diluents/styrene levels.
The basic equation for working out how thick a laminate will finish is:
T=
W [SGF/WF – (SGF – SGR)] SGR x SGF x 1000 x (1-U)
2.2
BUILD PROCESS
The decision the build sets of three test laminates was to aid the comparison of results generated by having one laminate built using current Rustler practices and the other two using alternative ideas, for gap fillers in this case foam sections shaped in trapezoids and squares suggested by Hugo du Plessis as a means of combating the effects of “hungry horse” distortion of the polished topsides of the yacht. These distortion marks found on the free board of yachts is quite obvious when looking down the sides of yacht hulls that are parked around ponsharden boat yard whether they are displacement hulls of plaining craft.
Since composite materials are often the key choice in weight critical applications the process of joining and fastening should not add unnecessary weight to the structure. Reasons for selecting adhesive bonding over mechanical fastening or other forms of joining include cosmetic, technical, economic etc. The building of the test laminates was conducted using glass plates that measured 50 by 50 centimetres as a work surface but the overall size of the laminates were 40 by 40 centimetres leaving a 10 centimetre gap around the edges to allow for overlaps. Before the laying of the glass fibres can begin a release agent has to be applied to the work surface in the same manner as when waxing the inside of a mould which prevents the laminate from gluing itself to the work surface. Two layers of wax were applied to the work surface making sure to wax on and off. The first layer of 850 quad matt/225 chop strand matt was laid onto the work surface and wetted out using a roller, a consolidation/nip roller and a pot with which could hold 900 grams of resin catalysed with 2% of the weight of resin minus the weight of the pot which weighted 40 grams. The layers of glass were applied in groups of two with four layers in total which complied with Rustler’s build practices as a means of reducing the amount of heat generated by the thermo reaction between the catalyst and the polyester resin. Once the three bases where finished three pieces of marine grade ply representing bulkhead sections measuring 20 by 40 centimetres were glued to three sections of foam, one shaped the same size as Rustler’s current build practice the other two were the same thickness 15 millimetres but one was a trapezoid the other a rectangle. The tops of the foam sections had to finish the same thickness as the plywood sections to allow for ease of build but not in the case of the rectangular section. Another of Rustler’s practices regarding bulkhead installation is to cut an “L” shaped rebate at the base of the bulkhead to allow one side of glass fibre to finish flush with the surface which was also employed in the laminate construction. Before the foam and bulkhead sections could be bonded the panels had to be “keyed” using 120 grit sand paper which helps to improves adhesion between two surfaces and then wiped down with astatine to remove the dust and other impurities which might poison the chemical bonds. Adhesion of the foam/ply sections to the laminates was aided by gluing them roughly in the centre which would help to reduce movement when glassing in the sections to the panels. Glassing in the bulkhead sections was done using hand lay-up using the same techniques as Rustler buy using pieces of 600 chop strand matt that measured 6, 8 and 10 inches long by 40 centimetres wide done on both sides. Masking tape was anchored at the base of the work surface to the top of the plywood on both front and back to help check the ply at a 90 degree angle and to help with ease of build.
The first group of test laminates were put aside for a week giving plenty of time to cure after which they would be removed from the work surfaces and checked for distortion marks to see if there was any differences in appearance or improvement in overall quality. The process was repeated with the other two sets of test laminates the only difference was that groups two and three had gel-coat applied before laying up the fibreglass which would help to highlight the distortion marks better and add to the realism of the build process. The second group of laminates were only left on the work surface for less than 24 hours to see if the effects of bond burn would be greater than groups one and three which were left on the work surfaces for 6 to 7 days each. 3.
RESULTS AND DISCUSSION
Once the test laminates hand been completed and left to cure for a curtain amount of time ranging from 24 hours to 6/7 days depending on the build specifications, the gel-coat surfaces were given a light polish using putty required give the freeboards of yachts a shiny finish as well as a protective surface.
Figure 1 shows one of the laminates taken from the second group of three highlighting how they all came together in the build process with the only differences being the foam shapes used as gap fillers between the panels and the plywood bulkhead sections. This laminate was the first of the second group which had the gel-coat applied before the laying up of the glass fibres and all the laminates in the second group were only left to cure on their work surfaces for less than 24 hours. With these laminates if was decided to apply the bulkhead sections straight away, this was to record how great the effects of bond burn and shrinkage would be compared to the other two sets which were left on their work surfaces while the secondary bonding was being applied.
Figure 2 is a side view of figure 1 highlighting the importance of accuracy with regard to the foam section’s width. As is shown there is a size difference between the foam and the ply thickness which has created a gap and undulation in the glass sheathing which could allow water and other impurities to penetrate the laminate and cause de-lamination if this sloppy workmanship was done on a real yacht.
Figure 3 shows a side view of another laminate from the second group this time with the trapezoidal section of foam which measured 25 millimetres wide by 15 millimetres thick. A theory was put forward early into the build process that because this shape had reduced angles of 45 degrees instead of shape 90 degrees application of the glass fibres would be easier and quicker than figure 1 which represents Rustler’s current build practice. Using the trapezoidal shape proved to be the most affect in terms of ease of application and speed of build time, because glass fibres follow the shape and size of the consolidation roller it is very difficult to get the fibres in to very sharp 90 degree corners meaning a penny roller might have to be used to reduce both the curvature and possible gaps when glassing in secondary bonded structures.
the possibility of reducing the amount of time required to sand back the hull surface once it has been removed from the mould. The results were later explained to Richard Garland Rustler’s head laminator who went through the findings and was able to point out the pros and cons of the results. The current choice of build methodology as shown in figure 1 is something born out of tradition and familiar practice because it allows for simple and quick installation. The two experimental ideas regarding the choice of size and shape of the foam would lead to a slight cost increase when purchasing the foam because more is required for a given area.
Figure 4 shows the side view of the third laminate from the second group using the rectangular section of foam measuring roughly 35 millimetres wide by 15 millimetres thick. This shape was the second option suggested by Hugo du Plessis as gap fillers which he described would be just as effective as the trapezoidal section. This was by far the hardest and most difficult the glass up using the hand lay-up technique because of the two 90 degree corners which make getting a tight fit more difficult than as show in figure 2. Theoretically this rectangular foam section would be just as effective in absorbing the compression loads exerted onto the hull as would the trapezoid because they both have a wider surface area than figure 2 thus helping to reduce the hungry horse distortion marks. Overall there were clear similarities between all three groups of test pieces in terms of build time, efficiency and most importantly how they helped to highlight and combat the effects of print-through. Because only groups two and three had gel-coat applied to them identifying the distortion marks became a lot more easily, when compared to figure 1 which represents Rustler’s current build practice figures 3 and 4 displayed a slight reduction in the amount of distortion, this helps to support the hypothesis that because the trapezoid and rectangular foam sections have a wider surface area they were able to better absorb the compression loads and cope better against figure 2 in terms of overall shrinkage of the foam filler. The significance of the results were further highlighted by Simon Combe (Falmouth Marine School’s lab technician) who described the process of how himself and other workers for Rustler must spend roughly two weeks of constant sanding of the gel-coated freeboard of all newly built yachts to remove the hungry horse distortion marks as well as other defects which are attributed to the quality of the mould tool surface itself. Because figures 3 and 4 showed a slight reduction in distortion marks compare to figure 2 it is entirely possible that if Rustler were able and willing to adopt new build methods similar to those test pieces it leads to
Richard also highlighted how the installation of the interior joinery would be affected by the two experimental choices, if a trapezoid section was introduced into the build process then the joiners would have to either scallop around it or box in around it to help with the installation of cabinet units and other utilities. 4.
CONCLUSIONS
Although limited, there has been some significant work undertaken to demonstrate why mechanical bonding is the preferred method of installing secondary internal structures in the case of this study bulkheads. While is it fair to say that a small cross section measuring 40 square centimetres cannot truly represent the overall interaction between the hull skin and the bulkhead, the test pieces that were produced did in fact highlight a small amount of the distortion that affects the aesthetics of the polished topsides. Other aspects such as time constraints and limited resources in the case of the latter did have an effect on the ability to produce test pieces that could have been larger in scope and scale and unfortunately there wasn’t enough time to conduct destructive testing to gather information regarding the laminate’s compression and sheer tolerances. Another key area of significant success would be the build guidelines outline by Hugo du Plessis for bulkhead installations and proving some of the points he made with regard to filler types and shapes along with highlighting other defects caused by secondary bonding such as bond burn. 5.
ACKNOWLEDGEMENTS
Firstly I would like to thank my course lecturer and project supervisor Alex Whatley for his patience and valuable direction and input. Thanks to Simon Combe for his supervision and input during the build process, thanks to Rustler Yachts and Richard Garland for allowing me access to their facilities and materials and finally a big thank you to Falmouth Marine School for giving me the opportunity to undertake this project.
6.
REFERENCES
1. Plessis, H., (2002), Bulkheads, Fibreglass Boats 3rd Edition, pages 94 to 97, London, Adlard Coles Nautical 2. Rustler Yachts., (2010). About the Rustler 42,[online], available at: http://www.rustleryachts.com/42.php, accessed 30/04/ 2010. 3. Gurit, S.P., (2010), Guide to Composites, SP Gurit, United Kingdom 4. Gurit, S.P., (2010) Superyacht Structure Insights, SP High Modulus, United Kingdom. 5. Larsson, L., (2007), Principle of Yacht Design 3rd Edition, International Marine/McGraw-Hill, Maine, USA