When going through chop, I can feel the boat flex. Almost like a "shiver" under my feet.
Walk up to a new GMW and knock on the hull. The solid "thud" it makes shows the knowledge GMW has gained over 30 years of building Thistles. The stiffeners Doug uses are about 3" wide and 1/2" high. They spread their benefit over a larger area and don't trap water or trip the crew. Unfortunately, 3521 is older than these refinements.
NW1D made a stiffener kit for a while, and a number of people in the PNW still perfer their approach. Based on the fact that deflection varies with the square of the thickness (ME's feel free to chip in here), this approach tries to achieve maximum stiffness. Its supporters also praise the usefuless of the stringers as a "ladder" when tacking the boat.
Perhaps the most common other approach is to slice wrapping paper of paper towel tubes in half then glass over the halves. Those who do this like the readily available forms and the fact that they're about the lightest form you can have.
I don't even think about starting something like this wihtout checking with the Chief Measuerer. Especially when the CMR that governs these includes such precise wording as "approximately parallel to the keel".
On examining my boat, I quickly determined that if I was to run the stringers all the way to the stern as the CMR allows, they would need to vary from parallel by 3-4" at least. Or else I'd have to stop them forward of the stern tanks. While I was waiting for an answer to my email, I realized that the CMR's length allowance covered boats without stern tanks, and that I should end my stringers forward of the tanks to keep them parallel to the keel -- the tanks already stiffen up the back end. The response from the CM confirmed this reading, but didn't really clarify the limits of "approximately parallel".
Why do what's proven when you can reinvent the wheel? In looking at the established approaches, I wondered how much stiffness was enough. After all, once you stop the boat from flexing, any extra support is redundant. Another thing I considered was that the narrow but tall stiffeners may be too focused, and only stiffen part of the hull.
Both the GMW and NW1D approaches were simple shapes and had advantages. I decided to try and get the best of both by approximating an inverted catenary (that's the shape a cable makes when suspended from the ends. Think the St. Louis Arch). So with a little time and an online graphing program I came up with this curve. I then approximated it linearly so I could form the foam on my tablesaw. This gave me the height of the NW1D approach while nearly giving me the width of the GMW approach. It also eliminated any sharp corners which would make it easier to lay the fiberglass and prevent banged knees. Here's a word document with my notes from making the foam.
I cut the angles of the foam on my tablesaw. The challenges here were that some of the angles required using the fence as the "bottom", and that the foam was so light it would "float" up off the table and get cut too deep. The first just required patience and care to handle, while the solution to the second was to use a featherboard. A traditional featherboard would have crushed the foam, so I ended up using foam scraps- they had just the right combination of flexibility and strength.
This was pretty simple. Measure out from the keel to set a few points, and mark the line using a straight stick and a sharpie. 
In some of the pictures you can see 2 sets of lines about an inch apart. This is because I set decided my original placement got too close to the bailers.
I hate grinding fiberglass. Messy, itchy, and generally a pain. This was no different. I initially tried using my 4inch air sander, but quickly decided that would take a month. So I switched over to my angle grinder with a 5inch 80 grit disk on it. This was much more aggressive and worried me that if I slipped I'd go straight through the hull. But it did cut through the interior gelcoat, and keeping the grinder as flat as possible prevented problems. Each stripe took about 1-1.5 hours to grind. I did them over several days to break it up.
Trey came over and helped me install the foam. I could have done it myself, but there were a few places (like handling the long pieces after they were coated with epoxy) where a second pair of hands was useful. He also helped me with the back cradle for aquarius, and that definately needed 2 people. Some simple math showed how far from the edge of the sanded area the edge of the foam went, we used the same long board and sharpie to mark a line as I had for laying out where to sand and marked where to put the foam.
Before I glued down the foam, I took a sanding block and lightly sanded the corners of my foam. Just enough to round the foam so it was more of a curve. Took maybe 5-10 minutes for all the foam using 100 grit paper.
The foam pieces were prepared by rounding their outer ends and mitering the inside ends (where the 2 pieces that define each stringer meet) so the fiberglass would lie nicely and so there wasn't a vertical joint. Then we put some epoxy down on the boat, covered the bottoms of the foam with epoxy, and stuck it down. A little blue tape held the foam down well except at the front where the twist in the hull required some extra weight to hold things in place. Overall this took maybe an hour.
I'd bought several pieces/weights of fiberglass tape for this. After checking the width and thickness of several, I settled on 2 layers. The bottom layer was a 4inch wide 8.7oz biaxial tape (West Systems). The top layer was a 6inch wide 17.8oz biaxial tape. I got both from Jamestown Distributors.
A day after installing the foam, Trey was back to help put in the glass. I'd already sanded off any excess resin from the day before and touched up the sanded areas around the ends (accidentally reshaping the back curve of the port stringer in the process). We measured each piece and cut nice round ends to match the area I'd ground off. We wanted everything pre-cut as accurrately as possible so we didn't have to cut wet cloth. Then we vacuumed the boat to clean up any stray dirt or fiberglass parts.
To wet the cloth out we used a paint tray. Trey wet the resin out while I used a plastic scraper to squeeze out the excess. The length was neatly folded onto itself until the entire piece was wet. We used about 25 pumps of resin (West System) 
for the 4" pieces and 22 pumps for the 6" pieces. We were a little more aggressive squeezing out the resin on the 6" pieces, and had a little resin left from the 4"
Installing the glass went way too easily. We just unfolded each piece along the foam and worked it so it lined up on each end. Everything laid down beautifully. We laid in both of the 4inch pieces then worked them down- mostly with our hands, but we also used a small foam roller. Then we went back and did the 6" pieces. The back curves just laid down perfectly , the fronts required about a 1" cut at the front to get them to lie flat. The aft port curve was the nicest- my "accident" with the sander earlier gave it the gentlest slope.
By this time I was wondering what was going to go wrong. The glass was in and lying flat, we'd rolled it down and then used our fingers to get the nice curve right where the foam and hull met. The 6" (which was actually more like 6.25" wide) just hung a little over the area I'd ground off and the total cloth almost perfectly matched the thickness of the gelcoat I'd ground off. About 1.5 hours after we'd started, we were cleaning up the few drips of epoxy we had.
On checking it the next day, I found one small spot that had a bubble under it. I groundd it open then filled with epoxy+microfibers. There were a couple of rough spots near the front ends (which had the steepest dropoff and had been cut and folded over themselves to make them lie flat). I sanded them a little to take off any rough edges. The most significant complaint I had was that the biaxial cloth I'd used had significant print-through. Live and learn I guess. I'd picked the biaxial cloth because some materials I'd read implied it provided superior strength. But the extra fairing will be extra weight.
I sanded the edges to fair the new glass into the old gelcoat (which basically required sanding off only where it overlapped the old gelcoat. The 17oz cloth was almost exactly the same thickness as the old gelcoat.). Then I mixed a batch of resin and added Microlite® to make it about as thick as honey and applied with a foam roller. After several hours I mixed up more resin and thickened this until it was more like mayonaise and applied with a plastic spreader. The first patch went on real thin with no significant drips. The second created some serious puddles at the edges, but they sanded away nicely. I'd built up the epoxy with the intent of having just enough that I could sand it all smooth without cutting into the glass fibers. After sanding with a palm sander and then foam sanding block, everything was nice and smooth, and I I stayed mostly out of the fibers.
After waiting more than the recommended 3-5 day minimum cure time, I prepped the surface as directed then applied Interlux® epoxy barrier kote to the areas that were new epoxy. After that dried I put on a finish coat of Interlux Brightside to match the existing gelcoat. To make the surface non-slip, I added Intergrip to the Brightside. Salt, sugar,and sand all sound like great ideas for non-skid, but I'd bought these components before I encountered those suggestioins.
After finishing this, I raced in Alamitos Bay, where there's always plenty of wind. I could still feel some flex in the boat, but nowhere as much as before. The non-skid finish was also great- not abrasive to scratch anyone, but a nice grip for the feet.
After I was done I talked to Doug Laber at GMW on other matters. We briefly touched on my stringers. He wasn't surprised by the print-through from the biaxial glass. If you're considering adding stringers, Doug would gladly talk to you and the time would be well spent. He's forgotten more about thistles than I know.
Copyright
Notice:
All
content and photos: © 2002-2024 Jason and Brenda Rziha . All rights reserved.
Permission is granted to electronically copy and to print in hard copy the images shown here exclusively for personal and private use.
Contact us for higher resolution images, and to obtain permission for any other use of this material.