by

Eur. Ing. Jeffrey N. Casciani-Wood

Chartered Engineer

F.R.I.N.A., F.C.M.S., M.A.S.N.A.M.E., M.I.I.M.S., M.A.E., F.LL.A., F.I.Diag.E., F.I.Corr.

The introduction some forty years ago of plastics to the boat building industry produced, in its wake, the problems of how to survey such vessels and what the Surveyor had to look for.   The intervening years have produced a wide variety of often conflicting surveying experience and widely differing answers to these problems.

Boats are manufactured as a stressed skin monocoque structure from a thermosetting resin strengthened with various forms of fibre usually of a glass or carbon type and fitted, for cosmetic reasons, with an outer gel coat usually not more than one or two millimetres thick.   The mixture has a number of additives to give colour, antislip properties and the like and these vary from design to design.   The chemistry of these plastics is complex and is often not fully understood by the Surveyor in the field but the resins used are basically of two types:

1. Polyester (which has three sub types) and,

2. Epoxy.

Both the polyester and the epoxy materials generally are a mixture of an acid and an alcohol with styrene as a diluent in the presence of 1.5 to 2 % peroxide catalyst usually Methyl-Ethyl-Ketone Peroxide (MEKP).   About 1% cobalt is also added as an accelerator to start the curing process at a lower temperature than would otherwise be needed.   All of these are organic compounds i.e. they are based on the nearly unique ability of the carbon atom to form complex long chain molecules. made up usually of isophthalic (also called metaphthalic acid) which is a saturated acid.   The purpose of the former is to control the amount of cross linking that occurs in the cure while the latter is used because it increases the resistance to water and chemicals.   It is somewhat dearer than the orthophthalic acid used in general purpose non-marine polyesters and it should be remembered that the orthophthalic acid polyester is more prone to blistering than the isophthalic form   The difference in the two resins lies in the other constituent - the alcohol.   This is usually propylene glycol in the main laminate resin as this has good water resistance properties but sometimes diethylene glycol is added in order to save cost and to increase the resilience of the cured laminate.   For the gel coat resin neopentyl glycol is used.      The physical characteristics of polyesters depends upon the precise nature of the components and the proportions used which vary from manufacturer to manufacturer.   Since glycols have three hydroxyl groups to form these esters, any resin involving it can form massive cross linked three dimensional polymers of the thermosetting type.   This process is called esterisation.   Such a resin slowly hardens with time by an exothermic reaction so that in a matter of hours the moulded hull is effectively in its final condition

 Polyester.

 Polyester as used in boat building is of three distinct but related types the chemistry being different for material used for gel coats when compared to that for the main laminate. For both laminate and gel coat resins the acid is usually a mixture of 25 to 40 % maleic (unsaturated) acid with the balance being although full cure does take a number of weeks depending upon the ambient conditions.   Both the diluent and the catalyst ultimately unite with the esters.   The by-product of the process is mainly water with some styrene which evaporates off.   The material is relatively cheap as it is easy to manufacture, transport and handle and it forms the basis of most fibre reinforced plastic boat structures today.

Its downside is twofold:-  

1. It is not as impervious to water as was originally thought, indeed, it absorbs water at a predictable and steady rate.

2. The esterification process is reversible in the presence of absorbed water under given conditions by the process of hydrolysis and the by-products of this reduction, which vary from case to case, can be ethanoic (acetic) acid, styrene, glycols, free amines or a mixture of these.

 Epoxy.

 Unlike the linear chain polyesters, epoxides are complex molecules of multi-directional, interlocking ether chains.   Epoxides have greater strength; they are more impervious to water; have better resistance to fatigue and are less flammable than the polyesters. Further, etherisation is not so readily reversible as esterisation.   Epoxides are cyclic ethers containing three membered rings; the simplest and most important of these being epoxyethane.   They are saturated aliphatic ethers and form a homologous series which correspond to the molecular formula CnH2n+20 and are isomeric with aliphatic monohydric alcohols.   However, epoxides are more costly than polyesters hence their lack of general use as a boat building plastic.

Surveying the Outside.

 In surveying the outside of a plastic hull there are a number of defects that the Surveyor has to look for but these can be classified under three main headings:-

 Contact Damage.

  This usually is in the form of scores or scratches of varying depth and importance, contact gouge marks and breakage of the gel coat due to the presence of underlying voids.

 Crazing. 

 This is usually of two general forms, star crazing and long line crazing.  This defect is usually found at the deck edges, under deck fitting feet and at points of sharp contraflexure or high local stress.

 ‘Osmosis’.

 Osmosis is one of the processes by which water is absorbed into the laminate and it can be and is reversed when the vessel is left to dry out on the hard.  The word is, however, among the boating fraternity, aided and abetted by the popular yachting press, usually and totally incorrectly used to describe gel coat blistering  formed as a result of the chemical and structural break down of the resin matrix as a consequence of water absorption whether this is by true osmosis or not.

The first of these main defects is fairly obvious and is usually found by simple observation of the hull and any voids in the lay up can be found by the familiar technique of the hammer test and a well tuned ear.   The second defect is also found by simple observation and experience soon shows the Surveyor where to look.

By far the most important defect in plastic hulls, however, is gel coat blistering, popularly but incorrectly called osmosis, or, more vulgarly, the boat pox!

Gel coat blistering can be found by simple observation if it is a fairly advanced stage but the early stages are often difficult to locate and, if not noticed, can develop with remarkable rapidity later leaving a very unsatisfied client and a possible law suit on the Surveyor’s hands.

If the boat is located inside a shed or the side of the hull is dark, a bright torch shone along the hull will often show up the blisters by means of the shadows cast.  It is also useful to carry a hand mirror so that the sun’s rays can be deflected along the hull to highlight blisters.

Finally it should be noted that all plastics suffer from ultraviolet light deterioration and boats left lying in the sun soon loose their pristine shine on the side or end lying toward the south.

 Surveying the Inside.

On the inside of the vessel different defects may be found.   The usual primary and secondary stiffening for this type of vessel is of wood or other material encapsulated to form a top hat type section.   Care must be taken to check that no cracking has occurred along any of the frp covers thus allowing water to enter the core and introduce rot.   Similar defects may be found at the bonding of joinery work or plywood bulkheads to the main hull.   More subtle defects are due to the down draining of the liquid resin during the laying up process and ‘hinging’.   The former tends to leave the top sides of the vessel starved of resin and not fully wetted out while the lower sections of the hull have excess resin.   The latter which usually shows itself as long closely spaced cracks occurs where the plastic laminate ‘hinges’ about a hard spot for example the shell ‘hinging’ due to hydrodynamic pressures about the line of a bulkhead or the cockpit sole about the vertical cockpit coamings.   A properly trained and experienced Surveyor will be able to find and report these defects in a ‘glass’ hull but every boat is a new experience and brings new problems and knowledge.

One final point is that, often such boats are fitted with built in fresh water tanks of frp construction with a gel coat on the inside.   As this inner gel coat is not visible it is often made of cheaper and inferior polyesters and blistering is a very common defect of such tanks which should be opened for as full interior inspection.   It should be borne in mind that, if ignored, such internal fresh water osmotic blisters will eventually burst putting their contents into the fresh water.   These chemicals not only have a very unpleasant taste but are also toxic! 

 Moisture Meters.

 These days it is de rigeur for a Surveyor to carry one of more moisture meters to measure the water content of the hull.   The results obtained are subject to interpretation and different Surveyors have different ways of effecting this.

The machines are of two basic types:-

1. The first works on the principle of capacitance and measures accurately the local dielectric constant or permittivity on an inverse logarithmic scale numbered 0 to 25. The readings can be affected by the local thickness of the gel coat, the density and quality of the underlying matrix, the presence of extra layers of reinforcement or structural items, chain cables, ballast, bilge water, copper, fuel or water tanks, gas cylinders, batteries and electrical wiring and similar items on the inside of the hull and even the Surveyor's body and clothing.   This type of machine is to be preferred as is does not damage the gel coat surface of the hull.

2. The second type does not suffer from the above restrictions as it measures the electrical resistance between two electrodes but, as it is used by inserting pointed needles into the surface in order to measure the material’s resistivity, thereby damaging the surface, it can only be safely used at the bottom of broken blister pits.

It should be remembered that moisture meters do NOT provide either and absolute or a percentage measure either by weight or volume of actual moisture content!   It must also be stressed that there is no direct relationship between moisture content and laminate condition and a high level of readings on their own do not necessarily mean that the laminate is suffering damage due to penetration by permeation or absorption or chemical or structural breakdown induced by hydrolysis leading to the process of osmosis developing with subsequent blistering of the gel coat.   Such readings, therefore, cannot by themselves and in the absence of other information be used to make a diagnosis of structural breakdown.   The danger sign is when readings taken at successive times a week or so apart remain persistently high and do not fail appreciably within two or three weeks of lifting the vessel from the water in accordance with Newton’s exponential law as this possibly indicates that the laminate is retaining breakdown products and justifies a more detailed examination such as the taking of surface hardness measurements, core samples and similar actions.

Finally, the type, shape and distance apart of the machines internal electrodes all have a significant effect on the level of actual readings given and, in considering the readings taken, this fact must be taken into account when comparing readings taken with a different meter - even one nominally of the same make.

Some years ago one of this Company's Surveyors was requested to act as a mediator in a minor dispute between another Surveyor and a boat yard.   The details do not matter but, when our Sovereign machine was tried together with that belonging to the other Surveyor and that belonging to the yard (both also being Sovereigns), the differences between the three machines were negligible.   A number of readings were taken, however, on a separate occasion which were clearly out of step with those previously taken.   After analysis, it was decided that the odd readings were due to atmospheric conditions; in this particular case, high ambient temperature and very low humidity. Theoretical considerations would suggest that the corrections for these factors should be quadratic but, within the fairly narrow limits of ambient conditions normally experienced in the UK, it has been found that a simple linear correction is within experimental error.

From this Company's experience, Surveyors may be expected to pronounce an opinion on boats that have been out of the water anything between half an hour and several years.   Obviously the time out of the water affects the readings as the hull has had time to dry out (including the process of reverse osmosis) and no published data on this drying out effect could be found.   With the permission of the owners and the help of a local boat yard, a number of experiments were made to record the wetness number on several vessels over a several month period over a number of fairly widely separated days starting from the day they were taken from the water.   It was found, as one might expect, that the readings corrected for ambient conditions varied in time according to one of the solutions of the Newtonian differential equation, i.e. the drying out was exponential and asymptotic to the x axis.   It was from these experiments that it was decided that the ambient corrections could be linear without loss of accuracy.   En passant, arrangements have been made to carry out similar experiments on several other vessels and, in due course, it is hoped to publish the results on this web site. Following on from this it was found that when the readings for a particular day were plotted individually, in general they followed, again as one might expect, the usual mesokurtic curve associated with the Normal distribution. It was also found, however, that the range of raw data lifted was often very wide while Pearson’s coefficient of variation rarely exceeded 35%.   The value of the skew coefficient, which could be positive or negative without any apparent reason why, was also usually very low.

A number of references were studied and several sources were contacted to ask if there were any data on the interpretation of the numbers lifted in the field with a totally negative result.   It was therefore decided to carry out some experiments to see if a reasonable correlation factor could be established between measured wetness number and the mass of water absorbed by a given laminate.   These experiments were carried out using a number of specially laid up samples in controlled ambient conditions and were able, from the results, to establish two empirical formulae:-

1. The first gave a simple polynomic relationship between wetness number and absorbed water as a percentage of weight.

2. The second - which was more complicated than the first - has enabled a 'prediction’ with a fair degree of accuracy of the order of wetness number to be expected on the survey of a boat from the known number of days she had previously been afloat and, later, out of the water.  

As a background to this experiment, the only known published data of a similar nature were those given in an article on the comparison of moisture meters in the article by Mr Staton-Bevan published in PRACTICAL BOAT OWNER number 319, July, 1993 under the auspices of the YDSA and Southampton University.   When the results of this Company's experiments were plotted, the mean YDSA line as analysed out of that article was used as the basis as this had suggested a simple polynomial relationship and it was found that our results sat very closely on the line drawn from the YDSA experiments.

Most Surveyors will be aware that often anomalous readings taken in the field can be put down to the machine ‘feeling’ the presence of extra thicknesses of lay up, water tanks, piping, wiring, gas bottles etc.   We have, however, in our Company a long suffering lady who seems to develop an inordinate amount of static electricity in her clothing.   We were quite surprised to find, one day, when she was with us that readings she took were measurably higher than those taken by the writer or his professional colleague.   The Company has several machines of different types and the anomaly was clear on all of them.   She was finally persuaded to remove her shoes and earth herself out.   This brought her readings down to the level of our own and clearly her ‘static’ was affecting the results.   This showed that the presence of the Surveyor’s body and clothing could possibly affect the readings on the machine.

 Interpretation.

 Intuitively, the probability of osmotic blistering developing on an frp vessel should vary with the measured wetness number Wnm since it is usually assumed that this value varies with the amount of water absorbed by the laminate by the process of osmosis among others.

By wetness number is meant the arithmetic mean of the readings taken on the wetted surface of the hull corrected for the ambient temperature and relative humidity:  the minimum number of readings taken being directly proportional to the wetted surface area of the vessel under survey.   These readings should be measured in an identical manner and to a defined standard so that the results may be accurately and scientifically compared.

The vessel must be inspected in dry dock, on a slip, blocked up on a hard stand, or hanging in the slings.   Inspection between tides in a mud berth is not a satisfactory state in which to carry out measurements of the hull and is not recommended.   The hull must be superficially dry and thoroughly cleaned of all weed, crustaceans and other marine growth and the anti-fouling coat examined all over to see if there are any surface defects, cracking or flaking or other types of breakdown.   In the case where the weather is inclement due to precipitation it is recommended that the survey be postponed if practical as the rain water prevents the readings being taken.

Where a hull has absorbed water by permeation and subsequent hydrolysis and osmosis has taken place resulting in blistering a further defect can then occur which is known as wicking where the water and or the breakdown products creep up the internal fibres by surface tension and capillary action.   Wicking can only actually be seen when the gel coat is transparent.  In vessels where the gel coat is pigmented, current technology does not enable wicking to be clearly seen and identified on a normal superficial inspection of the boat’s skin.

 Blisters.

 There are basically four types of osmotic blisters one of which has three sub-types.

Type 1 blisters range in appearance from minute protuberances which appears as a general surface roughening of the gel coat in small blisters of about 1 mm diameter. The observed state could represent the early stages of other types considered later or the condition could remain static for a considerable period without any further development in either size or extent of the blistering.   This condition has no significant effect on the strength of the hull laminate, nor, unless widely spread, on the protection afforded by the gel coat and needs no remedial action.

Blisters of Type 2 characteristically follow the line of the glass fibre strands and may either be a train of small pinhead blisters or, alternatively, may form an elongated ridge. This defect can be anticipated to develop by a coalescing of the blisters into a ridge like form cracked at the apex.   Like Type 1 this condition has no significant effect on the strength of the hull laminate, nor, unless widely spread, on the protection afforded by the gel coat and, again, needs no remedial action.

Type 3a blisters can only occur on vessels built with a double gel coat.   They are generally 5 to 15 mm in diameter and on a boat that has recently slipped are typically dome shaped.   With time out of the water the blisters tend to flatten.   When not already broken, the blister may be ‘popped’ by applying pressure with a sharp pointed tool whereon a characteristically vinegar smelling clear fluid emerges.   Underneath the bottom of the pit will have a smooth, often glossy, appearance with no evidence of a glass fire or fibre pattern showing.   Blisters of this type if rectified may not recur or, again, may break out on a different part of the hull later.   There is some merit, therefore, if the defect is not widely spread in letting the defect develop to its full extent and undertaking a single repair at a later stage.   The defect, though unsightly, is not structurally significant and repairs may be safely deferred until a suitable time.

Type 3b is similar in appearance, when the blisters are unbroken, to Type 3a.   The difference appears when the bottom of the pit is examined in that the glass fibres are exposed though they may well be covered with resin.   No dry glass will be visible.   In general the pit will also be deeper than that for Type 3a.   This type of blister leaves the laminate susceptible to moisture ingress and wicking and is structurally significant and does need remedial action.

When unbroken, the Type 3c blister presents the same outward appearance as the previous Types 3a and 3b.   On breaking open, however, this type is characterised by the presence of resin free, often fluffy, glass fibre strands.   The cavity beneath is usually deeper than that found beneath the blisters of Types 3a and 3b and may contain foul liquids even for a considerable amount of time after the vessel has been slipped.   This type of blister requires early rectification and remedial treatment should not be delayed for any significant time.   With this type of blister progressive increase in the defect size may be anticipated and the possibility of an area of delamination developing cannot be discounted if the repair is unduly delayed.   This type is also frequently associated with wicking.   Whilst not of immediate significance to the hull’s integrity when first detected, leaving a blister of this type to develop over a number of months may lead to a localised weak area in the hull.

As opposed to the sharply raised dome shaped blisters considered earlier, the Type 4 blister is characteristically broad and flat, often most readily found when viewing along the surface of the hull laminate or by touch when running the finger tips over an apparently smooth surface.   Blister size is typically 10 to 50 mm in diameter and raised some 1 to 3 mm above the surface of the surrounding laminate.   This broad flat blister is typical of defect which lies beneath the gel coat and the first reinforcement layer or even deeper still.   This defect does not mean that the hull is in imminent danger of structural failure, although it is indicative of a void within the laminate where water may be accumulating and, by osmotic pressure, tending to extend to the affected area.  In time, laminate integrity will be affected and early remedial action must be taken.

From long experience it is found that blistering is more likely to occur on hulls with dark blue, red or dark green pigmented gel coats.   This defect is also more likely to occur in warm fresh water than cold salt water as both density and temperature affect the onset of the true osmotic process.