catzooks.com

Fairing bulkheads (2)

Well, I took a good look at how the straight epoxy turned out today. You can see a number of spots where the epoxy isn't very thick at all, it's just a very thin layer over the wood strips, and the texture of the wood still shows through. I kept the spreader pretty vertical for the entire time, and wasn't too careful about if the sharp side or rounded side of the spreader was pulling; as a result, it looks like there would have been a lot of resin starvation were I actually glassing fabric. There's more instructions here; in particular when spreading glass on fabric, you should hold the spreader at around 45°.

Another note: Bram indicates that the first resin application is mostly to fill in staple holes, so the order isn't especially important.

Reading through the The Strip-Built Sea Kayak, I found out that having a grooved aluminum roller would help reduce the amount of excess resin used to hold the hybrid on to the bulkhead materials. I don't have one, so I ordered a couple from LBI Products (2x ES-422-100-300 and 1x ES-782-100-300).

So, to keep things moving along, I decided to flip the bulkhead strips over and get started on the other side. I made a couple of changes this time - first off, I was much more aggressive about getting the glue off. The paint scraper is like the first step of planing - it pulls off some of the high spots of the wood as well. You may leave a rough surface on the wood, but this will get cleaned up quickly with the sander. In fact, with the scraper, I didn't even bother to use the plane.

Next, I tried the random orbital sander with 40-grit, but it didn't seem to be doing much. So I switched to using an 10-inch strip of 60-grit on a sanding block. This made a big difference. I generated a lot of dust quickly and really smoothed things out. And I worked up a nice sweat.

Since this worked out so well, I started looking into buying or building a fairing board. Fisheries Supply sells a "3M Hookit Marine Fairing Board System", but the board is $50 and the sheets (10 of them supposedly) are $50. $100 seems a bit steep, so I asked the list, and got a reply from Rob Deane. He suggests just building one with 1/4" plywood board, and using 3M Repositionable Spray Adhesive to glue plain old aluminum oxide paper on.

In hindsight, because the strips hang over about 12 inches on either side of the tabletop, it might have been helpful to take two pairs of straight pieces of wood, wrap them in plastic, and clamp them on the top and bottom of the strips hanging from the edge. A lot of the fairing needs to be done at the end of the strips - they tend to be like "fingers", each piece not quite in the same plane. The boards would have helped with this.

Another excellent reference - One Ocean Kayaks has a bunch of pages with color pictures and good ideas on how to build a kayak. Excellent site.

Bram posted a pdf with instructions on how to glass out the entire hull:

When wetting the s-glass, start by dumping about 10 oz. of epoxy on the center of the bottom. Squeegee it from the center toward the sides, bow and stern. Dump the second 10 ozs. in a circle on top of the outer edge of the wet-out surface - not on dry material. Continue to squeegee in an outward direction in ever larger circles. Never leave a wet edge very long before extending it further. When you get to the sides, continue to work toward the bow and stern without letting any edge set very long. When the epoxy becomes tack free, about four hours, apply a second coat of epoxy with a brush. When that becomes tack free (another 4 hours), look it over. If you can still see cloth dimples, apply a third coat immediately. You will save yourself a lot of work and get a higher quality finish by applying these additional coats without allowing the epoxy to cure more than 6 hours between coats. This hull has 3 coats of epoxy on it and gave me plenty of material to wet sand without getting into the fabric.

(9 images)

Fairing bulkheads

After planing, I spent some time sanding down to get things a bit smoother. I was mostly trying to eliminate the rough feel of the edges between the strips; it takes a lot to get rid of those edges, however, so I figured I would go ahead and fair things out.

Bram's instructions aren't very clear about the order in which you put the epoxy on; it implies you start with a straight epoxy sealer coat and then put the fairing compound on. The strip-built boat says to scrape, plane, sand, fair and then seal. That's what I choose to do.

For five squirts of epoxy, I needed around 6-7 heaping teaspoons of 407 to get a heavy cream consistency epoxy. I spread the compound over a large part of the surface. The notches between the strips were filled, and in places there was a very thin film of fairing compound over the surface itself. I hit almost all of the strips. For strips which were low, I put a large amount of compound in the low spot to bring it up to the level of the surrounding strips. 407 is a pink-purple colour, but it will never show anyway. For the entire board, I used 8 squirts of epoxy.

I made a couple rookie mistakes:

• my first batch of epoxy was too much. I mixed five squirts, and should have started with three. By the end of the batch, the epoxy was pretty hard.
• I turned on a heater at the start of the process and let it run after I was done. The result was that the air trapped in the wood expanded and blew bubbles in the fairing compound. I should have turned it on long before I started and then turned it off when I was done; the air would then contract, thus no bubbles.
• it's really important to take off the extra spills. In my case, there were a couple "lines" of compound which weren't filling anything, I just didn't take them up. However, the epoxy doesn't flow, so they stay exactly where you left them, and it dries so hard they are very hard to sand down.

Once the fairing compound was dry (overnight), I went back and sanded more. I used 40-grit paper on my random orbital sander. Sanding at right angles to the direction of the strips seemed most effective. I had to turn my sander up to "maximum" to be able to make much of a dent - the epoxy is very hard. Tipping the sander on it's side helped take off some of the "mountain ranges" of epoxy blobs in certain spots.

Once things were sanded down a bit more, I mixed up some straight epoxy and spread it over the entire surface. I needed 9 (4+3+2) squirts to cover the whole thing.

Further research says that the Gougeon method of fairing is to cover the surface in plain epoxy, and while it's still wet (gelled if you're working overhead), go back and cover the required spots in fairing compound. Having done both, I can see why this is preferred - the initial epoxy application fills in some of the smaller holes so you don't have to go back and do as much work with the fairing compound. I'll try this on the other side.

(8 images)

Building bulkheads

The next steps are to build the bulkheads, partial bulkheads, and tee soldiers. To do this, you build a special plywood, with glued together strips laminated with kevlar/carbon hybrid and epoxy.

Normally the lowest-quality strips would go into the bulkheads, but in looking through one of the packages of wood, I can't find any knots or blemishes, so I just cut whatever was there. I cut 26 strips - 24 into thirds, and the last two into 20" pieces. Each strip is about 16'6" long, so each third is 65" long. With 72 of these, I have a flat board 65" tall by 54 1/2" wide.

Because the strips must run vertically in the bulkheads, stations 6 and 8 (the two bulkhead stations) must fit across the width of the board. Station 8 is about 69" wide, so to accommodate the extra width, I cut another two strips into 20" pieces and glued those on to build another 15" of width.

When gluing strips, you press them together tightly and squeeze all the glue out from between the strips, then staple the strip down to your workbench. The perfect glue line squeezed out should be a thin sliver of glue. Any more than this and you'll spend a lot of time with a paint scraper pulling the extra glue off (what I did). To get the perfect glue line, hold the glue bottle vertically and use the edge to control how much glue can escape - you should have a bead of glue down the middle of the cove as it comes out, not touching the edges of the cove.

For most of the board I used too much glue, and so had to scrape it all off. First, the best time to do this scraping is after the glue has started to set, but before it's become rock solid - if it's still white in the middle but yellow around the edges, it's a good time to scrape. Second, once it's dried, scraping glue sounds different than wood - glue has a plasticy "screetch" to it. Keep scraping until you don't hear that sound. The plane has a much worse time taking the glue globs off, so the scraper is best here.

After this, plane the surface down.

[Update 2008 Dec 20: I was talking to Matt Broughton about how to go about building these bulkheads, and an idea came up. Because you're building on a flat surface and don't need to shape the bulkheads, it might be easier to just rip the cedar for these bulkheads yourself, and save the money on buying strips. Get a couple 2x8s of cedar planking at the local Orange Borg, rip it into 1/2” thick strips, and glue together edge-wise. There's no bead and cove necessary, and the widths don't have to be perfect so no planing is involved (in fact, having more width is probably better because you save the glue weight). Doing this saves on the cost of the strips. Even I would consider doing this next time - the effort involved is minimal.]

(21 images)

Strongbacks cut

I've finally completed the messy task of cutting the strongbacks.

What I did was take the 1/4" hardboard templates I built, clamp them down to the particle board, then use a template router bit to rout out the actual pattern. When you're done, you have an exact duplicate of the hardboard pattern.

Doing the actual routing just means clamping together the pattern and the board, and continually moving the particle board around to expose another six inches to rout, while ensuring there is enough support for the weight of the router. The pattern should always be securely clamped to the board by at least one clamp. I used between two and four clamps on a pattern, two holding the pattern on, the others holding the board to the table. I also had two spring clamps which I used to clamp the particle board to the workbench.

Tips & tricks:

• This is messy. Essentially the router bit removes 3/8" of material all the way around the edge of the pattern. That's a lot of dust. There's no dust control on a hand-held router. Make sure you put up plastic sheeting, because dust will be everywhere. Wear your mask.
• You only need two sheets of particle board (and likewise you only needed two for the hardboard). With a little creative fitting, you can squeeze everything on two sheets. Sheet #1: stations 1/2, 1, 2, 3, 4, 8, 9, 10, 11. Sheet #2: stations 1, 2, 5, 6, 7, 12, 13, 14. (notice the repeats. ;-)
• Put the patterns down with the paper squashed between the hardboard and the particle board (paper side down). I tried doing it with paper side up for one station (the first, station 10) and found that the router caught the paper and ripped a bunch off. Another option is to spray the templates over with polyurethane (smelly, but protects them from everything).
• Be sure to have stations 10 & 6 at the bottom of the two sheets of particle board - I forgot about 6, so I only have the negative of station 10. I'll probably have to rout another negative for 6 (or come up with another approach). These negs are useful when it comes time to build the boat support.
• Be absolutely sure that when you're routing, the router is flat on the template. I ruined a couple templates this way - the router bit sneaked up on me and ground out a chunk in the pattern as well as the particle board.
• Before taking the router out of it's slot after doing a chunk, wait for it to stop completely. I pulled it out early for one and it bounced around against the sides, ruining another template.

I botched three stations - 1, 2 & 6, and had to create new ones from hardboard. 2 & 6 Bram sent me new copies, 1 I used the kinko's copy of (it's narrow enough that the horizontal inaccuracies are acceptable).

To transfer the control lines (centerline, waterline, and the positions of the strongback holes) from the templates to the particle board, I drilled out 1/64" holes at all the corners. At first, I tried to put a marker with a 1/4" long nib down the hole and through to the pattern. This didn't work out very well, because there's usually a little extra wood at the exit point of the drill hole, and also because the holes fill up with dust from the router. What I did instead was use small finishing nails to transfer the points. This worked great, and seemed more accurate to boot - you can tell the nails are vertical, and they lined up more accurately than some of the marker points I had transferred. They even create a short pilot hole so that when it comes time to saw out the strongback holes, they will be more accurate.

Greg suggested that it was worth creating another template 1/4" shorter for the interior bulkheads. I'm thinking it might be easier and faster to just take the current station template and shift it down 1/4" when routing the top edge of the bulkheads. This way, the offsets for the step-up/step-down will be correct as well.

[Update 2006 Aug 28: I was watching How It's Made and they were building something which followed a process similar to this. Their trick was to scribe the outer edge of the template, use a jigsaw to cut to within 1/4” of the scribe line, and then use a router to trim off the excess and get a perfect shape. Duh. Don't do this.

Another thought: If you did do the jigsaw method, it would be a *lot* easier to manage if you could use a router table with a large "table" surface and just slide the piece around. With two people man-handling each station, it seems like it would be a lot easier, instead of positioning and re-positioning the piece you're working on.]

(20 images)

More construction notes

Class rules are complete.

Here's some more construction notes which have come up from the list (click here to see other notes):

• (link) On page 44 the manual tells you to cut the strips 64" long. That is correct. On page 45 the manual tells you to cut them off at 55" overall. That is the length I used on the first boat but decided to make them a bit longer on the second one. My preference is that the final length should be 62 inches so that they go from about 1/2" aft of station 8 to station 13. This is not a class controlled length but rather a preference. I recommend that you make them 62" long.
• (link) Station 10 is 74" wide at the widest point.
• (link) An interesting note on why it's a good idea to get to weight minus tolerance - if you undershoot, then weights will be added to bring you up to weight, not weight minus tolerance.
• (link) It's a good idea when you bond deck & hull together that you include some "proof" that it happened on a certain day (important for measurement).
• (link) Bram posted a PDF indicating that it was worth making the stations, tee soldiers, and centerboard trunk 1/4" shorter than the stations. Take this from the TOP of the bulkheads and soldiers. In email, Greg suggested creating a duplicate hardboard pattern for the bulkhead stations (6 & 8), take 1/4" from the top of the duplicates, and then using those shortened patterns to rout the bulkheads.

  There is a tendency to cut the stations just a little bigger than the patterns, and the bulkheads a little bigger than the stations. When you add this growth to the growth caused by the hybrid and epoxy, it is very likely that your bulkheads will be 1/4" larger than they need to be (you will be at the maximum of the depth measurement at the centerboard trunk). I think everyone will save time by simply cutting your bulkheads, centerboard trunk, and soldiers about 1/4" short. That 1/4" should be taken off the top of the bulkheads etc. Remember, the bonding flanges will easily deal with 1/8" of fill anyway. This will save having to cut them down after they are installed, as I've had to do on both boats so far.

More interesting links

A few interesting links. First, there is the Cherub class, which is interesting because many of the boats are home-built. Lots of good reading there. Another good page is there links page.

Another interesting class is the Merlin Rocket. The Rocket is a class which has been around since 1946 in England; the boats used to be constructed as a clinker hull, and more modern versions are built with multi-chine plywood construction. They're also pretty fast, and less wet than the modern trapeze boats. They've a very appealing look about them. Here is a link for plans; they are expensive at £80.

Finally, there's the International 14 site. Bram has been racing the I-14s.