(referencing a Yamaha 2005 AR230HO where water penetrated the unsealed table base)

This is the condensed version - the original was written daily, so includes various drying and repair ideas along with rationale for my repair. My short opinion on penetrating epoxy for wood rot: unless you can dry the wood completely (which I learned is impossible without exposure), the strength of such repair will be very weak and the rot will continue. Your money and time would be better spent on a proper repair. Note that FRP simply refers to fiberglass. Though my results looked amazing, I covered the enitre deck with SeaDek following the repair.

REMOVING THE TABLE BASE: I shimmed the base with thin wood to prevent the screws from spinning and drilled them out with a bit just big enough to cause the heads to pop off once i got deep enough - this bit was somewhere in the realm of 5/32". I did this to initally inspect the damage - I had to progressively gain more access as I tracked the edge of the rot.

I later discovered the seat is not mounted like the table base, as it does not have bolts that go through the floor to nuts. Instead, the floor has a square metal plate sandwiched below it rather than wood. This plate is tapped - so removal/install of the post seems simple (though I imagine an impact screwdriver would be needed ...and some luck). The plate is 12-13" square and does not appear to be aluminum or stainless. I marked the floor where these plates were located and avoided them.

CUTTING THE ACCESS HOLE: Tapping the floor with a long piece of metal rod and listening to the sound, I knew the damaged area was large. I chose the original cut out (access area) with the plan to cover the saw/repair lines with traction mat - at the time, I assumed the rot had not spread beyond this. I laid done some masking tape and drew a line 1" from the smooth channel between the traction areas. I used a 1x3 secured with c-clamps as a guide for my circular saw. The blade I used was a new, cheapo 7.25" Irwin Classic 140T (the more teeth, the better) mounted backwards to prevent chipping. I set it at a depth of at least 1/4". The floor is a sandwich: there's the actual gelcoated fiberglass topside (referred to hereafter as the "floor"), the wood core (referred to as the "core/coring"), and the bottom most part that sandwiches the underside of the coring (referred to as the "backing"). The backing is a thin sheet of fiberglass presumably there to seal water from the backside of the wood. I was careful to leave the backing intact to support my repairs.

DETERMINING THE REPAIR AREA: I used a metal paint scraper to carefully remove most of the rotted wood off the backing. There were some chunks still mostly intact that required the use of a limb saw (electric chainsaw) to break them up enough so I could pry it from the backing, a grinder spinning a 40grit flap disc, and a chisel. I found that there was more damage than my new "access hole" was allowing me to get to. I learned that wood sandwiched between FRP is impossible to thoroughly dry - penetrating epoxy would not work correctly. The access hole wasn't big enough, so I began to use a piece of metal rod to probe the recesses searching for solid wood and would mark the deck with a Sharpie. Then I drilled holes in an attempt to find the edge of the rot - grey is rot, tan is good. A trick to getting at rot under FRP (ex: for the layup bevel) is to place an Allen wrench in a drill, route the bad wood, and vacuum it out. I later mapped the damage on paper in order to calculate supplies.

I laid down some 2x3 studs to provide support so I could cut the rest of the floor away. I used the circular saw with the Irwin Classic 140T blade and a grinder with a cut off wheel to remove the floor. I left an area around the perimeter of the damage for my fiberglass bevel (the slope needed to overlap new glass on top of old glass). The rule of thumb is a 12:1 chamfer - which is very big. The thickness of the original floor glass is around 5/32", so it would require 1.875" of work space beyond the cut (5 divided by 32 times 12). Space is limited in some areas, but I have confidence a solid core bedding, proper surface prep, and meticulous layup will compensate for it. So, I opted for a 10:1 chamfer with an averaged depth of 5/32" which is a 1.56" bevel (5 divided by 32 times 10). I rounded down to 1.50" because I was confident in my layup skills. So to maintain the factory edge/lip of the non-skid, this 1.50" was measured back from the outside edge of the non-skid perimeter in those areas.

* * * * *


0.50" Balsa Core
Arjay J-Core Bonding Compound 4501
(J-Core alternatives: Core Bond or poly resin thickened with fumed silica (Cab-O-Sil))
Vinyl Ester Laminating Resin (2.25 gallons)
MEKP-925 General Purpose
Fiberglass Cloth (8oz) and Chopped Strand Mat (0.75oz & 1.5oz)
Chopped Glass
Gelcoat (FGCI's FGCS White - at least a half gallon)
Mixing Containers and Mixing Sticks (dozens of each)
Sanding Aid (wax additive; surfacing agent)*
Naphtha (for removing the wax additive)
Acetone (lots)
Paper Towels (several rolls)
Kaizen Foam (I used foam that was almost 1.25" thick)
* I used a wax additive for this project, but PVA is an option. PVA will require an HVLP gun with a 0.8-1.0mm tip for a project this size.

3M 6200 Face Mask (P100 particulate filters for grinding; 6001 organic vapor filters for mixing)
Cheap Layup Brushes
Caulking Tubes & Plungers and Gun
Cheapo 9" x 3/8" Nap Paint Rollers and Disposable Roller Pans (for resin)
Wooster Tiz 7" x 1/8" Foam Rollers and Disposable Roller Pans (for gelcoat)
9", 7", and 4" Paint Roller Frames
Various Sanding Blocks and Boards
Sandpaper (40grit to 1500grit)

Ryobi 4.5" Grinder (fiberglass disc and flap discs)
Makita 9403 (4x24") Belt Sander
Ingersoll-Rand IR-4152 6" Orbital Sander
Husky 30gal Compressor
Shop Vac (6 gallon)

SUPPORTING THE REPAIR: The wood core the factory used was 0.50" balsa. Most coring options are not very rigid on their own, so I supported the repair with Kaizen Foam between the backing and the cross members above the fuel tank to prevent sagging while the resin cures. There are two thick metal bars that span the tank to hold it in place, about 1.50" wide. The fiche doesn't show them, but there are pieces of rubber between the metal and the tank, acting as cushions that will help to keep the Kaizen in place on them (the fiche also doesn't show the other two bolts and nuts on each end of the cross members). After a few trial fitments, I found that the desirable thickness of foam between the cross members and the underside of the backing for my boat was almost 1.50". The test-fitment in the picture is 20" long, but I cut two lengths of the foam to be 1.50" tall and 26" long (max inside width of the fuel tank enclosure is 28"). I added a bolt, lock nut, and fender washers as a handle to them so I could move them around easier with my forearm through the table base hole.

For additional layup support, I added two more strips: one between the metal cross members and one further to the bow (basically inline with the table base hole and another hole I made for foam insertion). After some trial and error, along with some shaping from a grinding wheel, I had two pieces 27" long and approximately 3" tall. I hole sawed an access point for the bow-side support that was large enough to insert it in one piece. For the other one that is pictured below, I cut it into 9" increments. I securely attached an anchor string to them so they could be pulled out later through the table base hole or the fuel pump hatch. A great way to cut Kaizen is to attach a sharp blade to a magnet along with a strong clamp which will give you a 90 degree cut. I also added some foam under the corners of the fuel pump hatch to keep it level until the repair was complete.

CORING: I went with old fashioned balsa since the rest of my balsa floor is still intact. Some will say that's a terrible idea and recommend the 2-3 times more expensive Nida-Core/Carbon-Core (plastic honeycomb, which seems to be iffy depending on brand), Coosa or Divinycell (foam). Their issue with balsa is that it can rot if it gets wet. However, the only way it gets wet is with poor manufacturing choices (ex. Yamaha not sealing the table base hole) or poor owner practices (ex. installing carpet buttons without properly bedding the core beneath). Balsa is strong, economical, easy to work with, and it matches what had been on my boat the previous 12 years.

I chiseled out the remaining chunks of balsa still left in the now larger space. Then used the drill with the Allen wrench trick to route wood out from under the perimeter as I planned to cut the core big enough to slide under the old fiberglass "overhang" so the balsa and new FRP layers are staggered to reduce shear points. Once the larger chunkcs were out of the way, I used the grinder and a 40grit flap disc to carefully remove any remaining old balsa or scrim. "Scrim" is the fabric on the backside of the balsa that holds the wood pieces together. I then switched to the 6" orbital with 40grit to level the backing better and get a little under the FRP overhangs. Note that 80grit is the finest you ever want to prep any surface for resin application.

Anyone claiming that polyester or vinyl ester isn't strong enough for boat repairs (versus epoxy) should note the picture below. This boat was 12 years old at the time of repair and was built with balsa and polyester, plus this shows it wasn't built perfectly. Aside from the issue from not sealing the balsa at the table base, the boat is still very solid. A proper build or repair with poly/vinyl is more than sufficient ...and won't interfere with gelcoat adhesion like epoxy (amine blush) sometimes can. I taped the imperfections that I could reach from below before bonding in the core to prevent drips.

I cut the balsa with a utility knife so dimensions were big enough to allow almost 1/2" to slide under the floor overhangs. To get them to actually slide in under the overhang, I had to chamfer the balsa edges a bit with the flap disc. Where the table base and bolts go, I cut blocks of the balsa off out - this area will be filled with J-Core to prevent water infiltration (something that should have been done at the factory). Before pulling the test-fitted balsa, I outlined the overhang edges onto the wood with a Sharpie - this gave me an idea of the balsa's placement/depth when I bonded it in so there was room for the next puzzle pieces. I also marked the pieces with an arrow pointing to the bow and a few with "S" or "P" (starboard; port) to make sure they get to the proper side. The second picture shows the tools I used to prep for coring. That bottom grind disc is a 36grit fiberglass disc and it's the equivalent of a light saber! I was reluctant to buy it since it is $11, which is 2-3x more than a flap disc. I used it to rough-in the layup bevel into the remaining floor around the perimeter, then refined the slope with a 36grit flap disc, followed by a hand block with 40grit to get it perfect. You can see an example of the 1.50" FRP chamfer near the port-side seat post (this was done around the entire repair perimeter).

BONDING IN THE CORE: I could've measured for the center of the table base hole relative to a few points so it could be located again, but instead used the SeaDek to relocate it later. I added backers (cardboard covered in wax paper) under the table hole and the extra hole I cut to support the glass repair there - secured a string to the one closer to the bow so I could pull it out through the fuel pump hatch after the repair. I taped from the underside of the poorly-glassed factory backing areas near the fuel hatch and the ski locker to prevent resin dripping through, but left the dozens of pinholes as-is since the 0.75oz mat will keep the resin from dripping through. I taped off the beveled edges and cleaned the backing with acetone. Drill 3/16" holes through the taped/beveled overhang every few inches to allow trapped air to escape as the J-Core fills the space below when the balsa slides underneath.

To help strengthen the balsa bond and fill the holes in the backing, I applied a single layer of 0.75oz mat with vinyl ester (VE) resin: I precut the pieces, painted the backing with VE, laid in the dry glass (it's way too thin to wet out before placing), added VE and stippled out the bubbles with a brush (I do not like finned rollers on finer mats). As soon as that was done, I mixed up a large quantity of the J-Core. Working one section of balsa at a time, I poured the J-Core onto the still wet layer of glass, and used a spreader to get it where it needed to be. Make sure to get enough under the overhangs to fill in the gaps - if there's enough, it should ooze out the drilled holes and can be wiped off. Then I put each piece of dry balsa in, shimmying it a little to let the J-Core settle in. If the J-Core oozes out on top of the balsa, either wipe it away or give it a nice, gradual transition to the bevel while it's still workable. I added some chopped glass to the remainder of the J-Core to fill the table base area I cut out. When all this was done, I covered the entire area with wax paper to prevent J-Core from sticking to the weighted panels if it seeped up. I got the old panels of the floor (I already ground the back of them flat) and set them back in on top of the new balsa to distribute the weight of the cinder blocks placed on top while the J-Core cured overnight.

SMOOTHING TRANSITIONS: The next day, everything looked great - it still had some give, so I did not put all of my weight on any single area. I used the hand block and orbital, both with 40grit, to level a few high spots in the balsa. I used compressed air and a vacuum to get up the dust. The core may not be perfectly level, but my new floor was more level than the factory's. Where the thinnest part of the bevel meets the balsa, there has to be a smooth transitional slope between the two materials. With another batch of J-Core, I used a plastic spreader to give it a nice, gradual slope between the balsa and the bevel ...similar to a fillet. This transition is very important for a strong layup. You can see that I addressed a few small gaps on the rest of the balsa to reduce chances of air pockets and ridges when I applied the glass.

PREP FOR GLASS: Be prepared for everything to take longer than expected. I carefully tape off the outside perimeter of the non-skid - there's that "ledge" that I was able to maintain. I did this by stacking the tape at least as high as the edge and leaving it there until the glass was in, removed it, and retaped it the same way for the gelcoat (it's a pain, but worth it). To make the pieces for layup, I use waxed paper to make a pattern: laying the paper out (tape together for larger sections) onto the deck and tracing the outline (where the bevel ends and old gelcoat begins). There were three main sections making up the large area: the sections near the fuel hatch and the ski locker were 11" wide and spanned from one side to the other; the section in between was the largest, about 14" wide. Then I transferred the patterns onto the dry fiberglass with a Sharpie and marking them which side up, where they went, and what layer they were (any ink bleeds will be covered with gelcoat). Each new layer was about 3/16" smaller around the edges than the previous - just continue to cut 3/16" off the original pattern for each smaller layer. With this type of sectional layup, avoid overlapping with the mat (try to butt the edges best you can), but be sure to overlap the cloth sections (maybe 1/2"). The cloth is thinner, so it won't build up so fast, thus trapping air at the overlaps. I cut all the layers and kept them in separate, organized piles (layers 1, 2, 3, etc) that were ready to go. My layup schedule was (first to last; bottom to top): 1.5oz mat, 8oz cloth, 3oz mat, 8oz cloth, 1.5oz mat. However, I found that 3oz was too thick to saturate for a project this size. 3oz is basically two sheets of 1.5oz stuck together - so I seperated it during application to ensure it wetted out properly before it started to kick.

Final prep sanding and masking off parts of the boat that I didn't want splattered. I draped a tarp over the work area to prevent the sun from speeding up the kick and to keep any falling leaves off. I set up a fan to keep me cool and help prevent dripping any sweat onto the work area, then set up a few lights in case I went into the evening (which I did). I kept a timer running non-stop for three reasons: 1) to make sure I stirred my resin/MEKP batches for 1-2mins; 2) to know when the balsa-sealing layer should kick; 3) to see how long it took me to complete.

This was a much larger area than I had ever done. So I decided to try using a paint roller to wet the glass out evenly and quickly. The roller worked so well that I also ended up rolling the gelcoat with great results. The only downsides are that it absorbs/wastes product (you'll need to factor that into measuring your batches) and it can leave product behind at the roller edges if you get ham-fisted which will need sanded away (try to roll it out in such a way that any extra ends up on the masking tape and pull it up at the end). I used a fresh, cheapo 3/8" nap roller on a 9" handle along with a disposable paint pan for each layer. I mixed resin in 20oz batches for the 1.5oz mat (took two batches per mat layer) and 12oz batches for the cloth (only one batch needed per cloth layer). I believe I used a little over a gallon of VE resin on this day of layup. The paint pan spreads the heat out, so it doesn't kick as fast as it would in the mixing tub (16oz will overheat in the standard mixing containers). I pre-measured the resin out in separate mix tubs and covered with lids before starting - I also had a tiny cup with pre-measured MEKP on stand by for each container. This eliminated the need for a helper and allowed me to get the next layer on before the previous could got too dry. Once the first layer of dry glass was stacked within reach in the boat and my other supplies where ready, I started the clock and mixed the first batch to seal the balsa core...

3:30pm weather: started layup at 80*F with 55% humidity; finished at 75*F with 70% humidity

GLASS LAYUP: To get the best bond between materials, I waited to seal the balsa on the same day as the glass layup. The balsa quickly soaks up quite a bit of resin. I don't remember how much I rolled on ...at least 40oz. The roller gets it on very quick - I gave it 25-30mins to get tacky before I started mixing the first pre-measured layup batch. I rolled down a fresh coat of resin and then placed a piece of dry 1.5oz mat. I wetted it all the way out before moving to the next piece. I kept a spreader, a brush, and finned roller (in a shallow acetone bath) on hand for gobs, bubbles, or touch ups. I didn't need the finned roller (for bubbles) much as the paint roller does double duty if I rolled from the inside out. Only put down enough resin to fully wet out the glass and then roll over it several times to get an even distribution. As soon as the entire layer was done, I quickly cleaned and replaced the roller, binned the roller pan, and moved on to the next. The delicate 8oz cloth is a PITA and I quickly found the best method was to lay all the cloth pieces out without re-wetting the previous layer because they NEVER lay right (remember to re-wet before laying down the mat layers, though). This allowed me to be finicky with them without worrying about the resin gelling-over because they are delicate. Once all the dry cloth pieces were lightly laid into position, I mixed the resin and rolled the cloth outward easily. I again broke everything down and set back up for the next layer.

When the final layer starts to kick, pull the built-up masking tape around the perimeter - most masking tape has a 3 day window for easy release. The entire layup took twice as long as I had figured: four hours from the first mix to the last roll. I was two hours past sunset when I finished, which meant lowering temps and increasing humidity. Just in case, I placed one of those chemical dehumidifiers (small bucket with pellets - though there was no moisture in it when I checked 36 hours later) on the floor and set up a Pelonis space heater (low heat with the thermostat set around 75*F) overnight, then secured the mooring cover back on (this tactic was used a lot for this project).

ASSESSMENT & ROUGH SANDING: It had about 36 hours to cure before I got a chance to look at it and it was absolutely rock solid - I could bounce on it without seeing flex or hearing any creaking. In the next picture, you can see near the locker that the glass didn't fit exactly how I had planned. Not a big deal, just some finesse-sanding in that area to smooth it out. Started knocking down the high spots - there was a lot of sanding to be done. The palm block, the long board, and the pneumatic orbital weren't doing the job fast enough, even with 40grit ...so I brought in the belt sander. I made an adapter between my Makita and my shop vac with 3.5" of standard 1" PVC pipe. I used my bench grinder to thin the wall of one end enough so it would insert into the sander, then the shop vac hose slid over the other end.

The belt sander can be lethal. Used carefully, I was able to easily save myself hours of sanding everything level. I started out using 80grit in the belt sander and found that being prudent with the trigger, being mindful of it staying level, and always keeping the unit moving prevented taking too much off. Once I got comfortable with the 80grit, I switched to the 40grit. The 40grit did an exceptional job of also knocking down the remaining non-skid around the seat pedestals. Instead of replicating the non-skid, I decided to uniformly level the entire helm portion of the deck in preparation for all new gelcoat. Removing the non-skid pattern, I judiciously feathered the trigger and kept a very careful eye on what I was doing to avoid going too deep or knicking adjacent surfaces. I backed off when I could see the pink fiberglass below the gelcoat. The edges of the non-skid and the seat bases were a little tricky/risky, so I left them as these smaller sections were easily knocked down with 40grit on the hand block.

LEVELING & PREP FOR FAIRING: There was a very slight sag in the middle of the new deck repair. So, I roughed the floor with 80grit, wiped with acetone, and added a few pieces of 1.5oz mat to help "hide" the imperfection. Once cured, I went at leveling the floor with a little more patience: I laid the edges of a 4' and 2' ruler across the new glass, checking for high spots and knocking them down with various sanding blocks. The next picture shows the blocks I had gathered for my previous FRP project. Most of the hand sanding was done with the yellow 3M sheetrock block - it has a soft, forgiving pad. The pair of red blocks with felt pads are 3M Stikits and the yellow pair are Finishing Blocks with velcro from Shopsmith (the precut Shopsmith paper is expensive, but it cuts faster and lasts longer). I tend to use the Shopsmiths for resin and knockdown, but use the Stikits for gelcoat due to their available grits. Note the orbital seems better suited for fairing gelcoat rather than the much harder resin.

With the highest spots knocked down, I searched out and roughed-up the lowest spots by hand that seemed too low for fairing with microballoons. I also removed the gelcoat on the aft-most water channel with a stone on my Dremel, then vacuumed and wiped everything with acetone. After taping the channel off, I mixed up some "peanut butter." I make PB by measuring my resin and adding fumed silica (using a tongue depressor as a scoop, I add 2-3 large mounds per ounce of resin). I'll stir that for a while to get the lumps out, then add MEKP and stir for 2 minutes to ensure a good catalyst mix. Finally, I'll add some chopped glass (2-3 pinches per 2oz is what I like) for strength. The fumed silica gives it a more consistent texture, preventing it from being runny between the chopped fibers. I used a stir stick to spread it onto the deeper low spots and channel, then used a plastic spreader to even it out ...it takes a bit of practice and finesse to not drag it all back up. Once it had kicked, I carefully peeled the tape away from the channel.

I didn't dish the new glass out since I'm going to re-gel the entire floor - no need to account for heaight difference between old and new gelcoat. After sanding through the non-skid, it's obvious that there are undulations from the factory over the entire span of the floor, so my project will actually produce a more uniform floor than it had. I sanded everything down as level as possible while still maintaining enough of the defined edge around the perimeter of the old non-skid. I kept this edge/outline intact so it would still match the SeaDek that went down afterwards, maintaining a near-factory look. Before applying fairing compound, sand everything with 40-80grit, tape off the edges, and acetone the surface to remove dust and oil.

ROUGH FAIRING: Fairing compound is resin with microballoons (glass spheres that make sanding easier) and I also add fumed silica to mine. It doesn't have much tensile strength, so it is only used to smooth surfaces before gelcoat. I start with the vinyl ester resin and add silica (2-3 big scoops with the stir stick). As with PB, stir this for a while to get the lumps out, then add MEKP, and stir thoroughly before adding microballoons. Note that adding MEKP last will not produce a uniform cure. It takes a good amount of microballoons to get the right consistency - I liken it to the consistency of chocolate pudding before it has set-up. Make it as thick as pudding out of the fridge if the first coat needs to go on a non-horizontal surface. Do not stir the microballoons very fast as it traps air and will leave bubbles.

I used one of my rigid 2' aluminum rulers on its edge as my initial compound leveler. Doing it in sections, I poured out some fairing compound and used a 6" plastic spreader to coat the area. Then I scraped the ruler edge across this, between the outer edges. The excess scraped up with the ruler was used on the next section. The aft corners were easy (next picture), but it got messy close to the ski locker where the floor undulates a little more - the first layer is usually very ugly in instances like this. I picked a suitable edge-to-edge plane and did the best I could to get the compound where it needed to be. It will need a lot of sanding in these areas, but the next layer of compound will be much easier all around now that the basic shape is down. As it starts to kick, pull most of the masking tape.

For any carpet button holes that remained in the deck, I applied tape over each and enlarged them with a 9/32" drill bit. I mixed up some J-Core and poured it into a large syringe (I drilled the syringe hole bigger to let it flow out better). Leaving the tape to prevent a mess from overflow, I injected the J-Core. I wiped each one level, though most of them were slightly below level after cure due to shrinkage. You can either try to scrape it just before it cures or level it after it cures with dabs of gelcoat.

FINE FAIRING: Sanding filler down using a belt sander is going to gum up the sandpaper quickly due to the additional heat from the belt - I still used it. I feathered the edges with the sanding blocks, vacuumed, wiped with acetone, and taped the perimeter again. Mixed up about 4oz of the vinyl ester resin with silica and microballoons then spread it where it needed to be. This second layer of filler fine tunes the first rough coat that got the basic shape outlined. I did 1-2 more applications to fill any bubbles and smooth out the tougher areas near the ski locker. Since the gelcoat is going on with a roller that leaves a textured finish and will be covered with SeaDek, my new floor did not have to be buttery smooth - just remember that gelcoat is like paint and any prep imperfections will be noticeable. I did the final sand/fair (3M sheetrock with 40grit and the Shopsmith block with 80grit) before vacuuming all the dust that had built up and hiding since the repair began.

FIGURING GELCOAT THICKNESS: This gets complicated... depending on who you ask, the final desired thickness post-sanding is between 15-25mils (thousandths of an inch), while some say to shoot for a final thickness of 30-35mils. The equivalent of 15-25mils being 1/64" to just shy of 1/32" or 0.38mm to 0.6mm. Generally speaking, gelcoat is applied in a thickness between 30-50mils if sanding is required (sanding is always required if a mirror finish is desired). The equivalent to these thicker figures being 1/32" (0.8mm) to almost 1/16" (1.3mm) before the sanding begins. For this project, I ended up rolling six coats with an 1/8" nap foam roller. Roughly eyeballing a core sample, this got me about 0.5mm with no sanding needed.

It is usually best to be on the thicker side of the window because this will allow more room for error during sanding and will also increase longevity of the finish due to the slowly decreasing thickness from maintenance compounding and polishing over the years. However, the thicker it remains above 25-35mils, the greater the risk of it eventually spider cracking as it shrinks with age. As for worrying about going too thin, the gelcoat really only has to be thick/opaque enough so the FRP beneath it is not visible. Gelcoat's job is to prevent UV and water damage to the substrate (though gelcoat is somewhat permeable to water - moisture is always coming and going). Keep in mind that applying a single, very thin layer of gelcoat may not cure correctly since the solvents can evaporate too quickly. Conversely, applying several heavy coats can trap solvents, thus preventing proper curing - allow the last coat to just lose its sheen before applying the next. Apparently you can let non-waxed applications harden for several hours between coats, but keep in mind that the prior coats' heat will help kick the following coats, so I would not wait too long.

PREP FOR GELCOAT: I opened the bimini and draped some plastic sheeting down the front and sides to keep the leaves and tree debris off/out of the work area. I pulled the port and starboard bench seats and placed plastic sheeting inside to protect the carpet. I rolled gelcoat from within the storage areas and the edge of the ski locker, using the seat posts to get back and forth. I cleaned the perimeter and near the seat post bases with acetone and began laying my masking tape. Masking took about 2 hours because I wanted it just right to maintain the non-skid edge. I taped the edge with at least three layers to help shape the "plateau" and then applied paper to the vertical surfaces to guard against splatter. I cleaned the entire floor area with acetone and did not step onto it afterwards. Here it is ready to for the gelcoat...

3pm weather: started layup at 81*F with 53% humidity; finished at 77*F with 54% humidity.

GELCOAT APPLICATION: For each coat, I used a fresh 7" Tiz foam roller along with a 3.5" (cut in half with a hacksaw) on the 4" roller frame. I started with the 3.5" to cut-in around the seat base and then switched to the 7" for the rest of the floor. My first batch was 10oz as I didn't want to go so small that I couldn't do a full coat. After the first application, I had a bit of waste left over, so the remaining batches were 8oz (perfect amount). Pouring it into a roller tray immediately after mixing in the MEKP gives enough time to roll a full coat with no worries of it overheating - each coat took about 13-16mins. It was 80*F, so I measured my MEKP at 1.25% to get a little more time as well. I chose the 1/8" nap foam roller since it would not soak up and waste a lot of gelcoat. Also, the foam does not leave behind lint and produces a very fine texture. By the end of each application, the foam was starting to let go from the cardboard tube, but it never became an issue. Forgoing additional microballoon applications, the gelcoat made some slight imperfections a little more obvious. Yet they weren't really that noticeable and I was never concerned since I applied SeaDek. The "worst" areas, actually, were the clusters of air bubble dimples in the factory fiberglass where I sanded down the non-skid - so I blamed the factory on these. I didn't even see them until the new gelcoat went on - all I could do was try to overfill/seal them during gelcoat rolling.

From what I had read, I had planned to do at least four coats - I ended up doing a total of six because the examples I saw were probably using rollers with a thicker nap. By the fourth application, I could no longer see the dark microballoons beneath. The first coat went on the thinnest because I was afraid of running through the first 10oz and not getting full coverage. I was not afraid to load up the roller with the subsequent 8oz batches. I rolled out each coat then replaced the pan and rollers, measured another 8oz of gelcoat, mixed the MEKP for 2mins, then went right at the next application. At this steady pace, the prior coat was just tacky enough to stay put when the fresh, loaded roller went over it. For the final coat, I added sanding aid (wax additive; surfacing agent) to the gelcoat mix - do not forget to account for the sanding aid volume when calculating MEKP.

As I did with the layup, I ran a stopwatch to ensure I mixed the MEKP for 2mins and to see how long it took. In total, it took 2.5 hours. I had planned for 2hrs, so not a bad estimate. I got the masking tape up while the gelcoat was still rubbery to prevent chipping the edges of a fully-cured gelcoat: perfect, defined edges. The results...

METRIC-ONLY CATALYST CHART: I made this simplified chart to make my life easier. Remember, if adding a wax additive, you have to figure and add additional MEKP for the increased volume.

* * * * *


Removed the Kaizen foam below the new deck that was placed there for support. Used Naphtha and a rag to remove the wax additive that had risen to the top of the cured gelcoat. I used increasingly finer grits of sandpaper to smooth out the area behind the fuel pump hatch so it could be polished to match the majority of the gelcoat surface. As for the new "non-skid" area, I did not sand or polish this - the roller left a subtle/attractive texture not much unlike an eggshell.

THE TABLE BASE: I used a couple carpenter squares and my 4' ruler as a straight edge to align my new SeadDek between the ski locker and the fuel hatch. With it taped into place, I traced the SeaDek's opening for the table base location with a pencil, found the center, pilot drilled it, and then used a 3.25" hole saw. There is no way to add nuts to the back of the table base bolts, so I made two plates that I can bring up from below. The plates are very roughly 7.5" x 7.5" and 0.25" thick aluminum. With my drill press and 3.25" hole saw (Yamaha uses a 4"), I put a hole in the center. I then added the six holes and used my homemade guide to accurately tap the 1/4-20 threads by hand. I made two plates since I added a base to the bow area.

MORE ROT FOUND: I removed the latches on the ski locker and cleanout hatch - not surprisingly, the wood around the latches was very deteriorated. I removed the swim deck's table base and found that the lip of the center hole had been sealed with white silicone at the factory (but not the mounting holes). I had to place a crowbar at the inside lip of the base and use a piece of wood to pry/leverage it off. There are only very slight signs of water intrusion due to the use of silicone. The swim deck core is not balsa, but plywood.

ROT PROOFING: Routed the latch areas with the Allen wrench and drill. Set up backers and poured in some J-Core, let it cure, and hole sawed it back out - I brushed PVA around the area before the pour to prevent any spill or overflow from sticking to the gelcoat. I applied wood sealer to add an extra barrier of protection to the treated plywood (it's way too solid to route out) of the swim deck and in the screw holes of all the hatch hinges using a needle syringe. Sand and water are able to get under the seat posts from the factory, so I sealed the bases with a small bead of white 3M Marine Silicone (#08017). Note that 3M 4200 and 5200 are overkill for most projects which are also a real pain to remove later ...and I usually like overkill.

With the new table base backing plate ready for install, I fished the strings through two holes in the deck and pulled the plate up to where it needed to be, then inserted the screws with some anti-sieze on them. Once tightened up, I ran a bead of white 3M Marine Silicone around the outside-bottom perimeter of the table base (as well as the one on the swim deck). With the SeaDeck installed, the project was complete. I actually prefer the texture of the area that I fixed over the non-skid area and wish I had sanded the rest of the deck smooth.

CrankyGypsy (established 2001)