May

After that initial sand, I still had some problems with several low areas: around the attachment of the foam bow, and the transition of more round/full keel towards the sharper entry area. I decided a skim coat would be easiest. I also spot-filled some bubbles.

(This is an area that has been problematic in varying degrees on all four float sides, and I've seen others comment also. I have two ideas. The reflection of extra glass at the bow entry leaves a low spot behind, and I've been too aggressive at sanding down the stripes in this transition area from full/round  to flat/vertical. Or the root problem may also be how the forward end of the float is constructed. The battens near the keel make a difficult twist in a compound curve as they approach form #1 - this may make this area shallow. A more careful investigation when attaching the foam cap and fairing the foam could alleviate this problem.)

I have a constant supply of used, but clean syringes from work. I cut off the Luer-lock to make a fine point, use a couple of reusable ergonomic modifications to save my hands, and fill them with a Ziploc bag. With these I can fill the small Dremel bur depressions left from attacking the bubbles with minimal effort or material.

I did the deck in a similar fashion as the first float. First I put on a single skim coat, working the material with significant force to work out any bubbles. I'll admit this is more successful than on the sides with the candy stripes.

I modified my sanding by trying the 'cheese grating' technique described by Wayne Hicks. Essentially you begin sanding while it's still soft - when it deforms with a thumb nail, but isn't tacky.

The idea is different enough that I was sure it wouldn't work . . . but it does.

Shortly after I finished applying the QuickFair I was able to begin sanding with 36 grit (at the bow where I had started) and slowly worked towards the stern while it continued to cure. The material came off in soft clumps that did not clog the sandpaper and were easily brushed off. Interestingly, the heat of the sanding seemed to speed the cure as I moved along. It was a great time saver to remove the majority of excess QF - and there was no dust! Unfortunately, I can't see how I could incorporate this technique to larger curved areas, as I couldn't apply it fast enough with the short cure time of QF.

Nearly finished except small detail areas.

After working with the soft QF during the 'cheese grating' I tried several other variations on this idea. One was obvious - cut off the tops of any filled Dremel divots with a wide chisel. This negated any effort trying to sand them without lowering my finished QF level.

I also found a couple of areas that were low from the edge of the trowel during the skim coat of the deck. Attempts to fill this type of area before had either left it low by trying to minimize the amount, or required extra effort to sand down without hurting the surrounding material. To fill these high, I put tape around the defect and skimmed over the top. This left the QF patch elevated by the uniform thickness of the tape. When strong enough to not tear, but soft enough to cut, I used a chisel and easily removed the excess.

Both these techniques took care of the last minute touch-ups before I began the final sanding to get it to that 'thin as possible' state.

I used the same skim coat followed by two coats of raw epoxy as required to seal the QF.

I have been working hard to get the floats to a stopping point before June. So, though they have a ways to go, this is how they'll sit.

Between steps I also continued to work on smaller parts. Here are the gudgeons as made in the mold.

My previous experience of vacuuming after wrapping carbon around an edge had not been perfect. As the excess resin is removed and as the fibers are compressed, they are not pulled tighter around the edge, but instead make a small wave or wrinkle. I decided to do a hand lay-up and let gravity pull the peel ply over the carbon. I've also found that making the part slightly larger and cutting to size makes the finished carbon uniformly thick out to the edges.

Gudgeons being fitted on rudder web.

April

The trip and getting back up to speed cut into the month, but I decided to start by using the Dremel and filling obvious bubbles before I finished sanding. I like to finish sanding, not filling, but of course if you fill too early and find some more . . .

I was pleased with the result. Hopefully the picture shows how you can see through the QF as it thins (yellow foam/glass color).

A close up with pattern from the stitched glass tow's and even a vague stripe to the right which is a seam between foam strips.

I do continue to learn about fairing and the rigid longboard (LB). I started out thinking: sand end to end, then back, then repeat. But there are a few nuances I've picked up over the last two sides.

First, there is more than one fair shape. We would like the one chosen by the designer and the one that has the least added fairing weight. I know my first side was fair, but likely not the best I could do now at achieving the other two goals.

What I've recognized is that the curved areas thin first. (I've mentioned not feathering at deck and keel till the last step.) A more subtle area is the transition from bow section to center hull. I think my impression on the first side was that this meant I needed the slightly thicker QF between two thinning areas (thinning must mean it was high, right?). After all, when I checked it was 'fair'.

I've come to a different conclusion. The LB wants to remove high areas, easy to visualize on a flat surface. But on a complicated surface, say convex changing to flat changing to convex, the game changes. Let's assume that with a given grit and force, an equal amount of material can be removed. On a flat surface, the whole LB touches (for example 1 unit force/100 sq inches). On a curved surface, the area of LB that touches significantly decreases, and with equal force the amount of removed material increases significantly (1 unit force/5 sq in).

Now what happens is that instead of creating an ultimately flat surface, the LB actually accentuates the curve! More is removed at the convex curve area, and less at the flat area since more is left behind. It's fair, it just doesn't meet our other two goals.

The floats have convex curves and lots of flat (fore and aft), but no concave. This means that you have to compensate for the above phenomena by either decreasing the force over the convex areas, or doing a 'double stroke' (or triple) and sand the flat areas for more time. Note: this phenomena is even more pronounced when you are sanding the candy stripes by themselves, so that you start at a disadvantage which only becomes more pronounced.

(Likely, someone will make the argument, "that's what a flexible board is for." I suppose. But then you have still more variables to control: stiffer for subtle curves? more flexible for tighter curves? should you still vary the force applied? . . . And I don't like flexible boards on the flat – which is really what you have a lot of.)

In retrospect, it seems so obvious. I think when I started that I was blinded by the sand, sand, sand, and the idea that "if it's fair, I must be doing it right."

There is also variable hardness in the QF depending when you get back to sanding. Although it cures enough to hand-sand at ~3 hours, it continues to cure to a harder state over the next several days. You can use this to your advantage by going after the mini-bubble fills early. But it can also be a disadvantage if you're sanding a larger fill patch where it is removed preferentially in comparison to the surrounding material with an advanced harder cure.

This is just a picture on the second side of how I try to put down the stripes. I use a Ziplock bag with the corner cut off at 45º. I've found that the sides have a seam, but the bottom doesn't. If you hold the bag with the bottom against the hull, the seam then makes the opening a triangle, giving the bead a nice wide base of attachment and a narrow top. The QuickFair is so good at keeping its shape that if you were to do it so the bead was round, it would not slough down, but cure with a narrow attachment and overhangs above. These would later just be voids, as it is very difficult to fill such small spaces.

Another time saver that gives superior results (gotta love something like that!) is to use a paint scraper or similar tool along the edge of the stripe instead of sanding. It straightens out imperfections, gives a perpendicular edge, and seems to be a perfect bonding surface. At least with my tool, I didn't notice any damage to the laminate. Remember, leaving the stripes high does not mean the final layer will be that thick!

I reread the email from Meno (posted in 'January', I recommend it). It's interesting that he went from stripes to simply troweling on material, and I've done the opposite. Of course it was his description of, "if you want to try candy-bag . . ." and listing some of the difficulties he encountered that has helped me feel that I've become successful enough with the candy-bag method. For me, an extra time investment to get the material on has been easier. Will I still think so after the main hull?

I proceeded to fill, then started making dust.

March

This is obviously my second time around for the shaping and glassing of the float, so I have just a few pictures and comments.

In the last post I forgot to mention that I had constructed the second bow slightly different. I made the central core of four sheets at the same time. But on the second bow I attached the outer pieces with contact cement, like Ian suggested. On the first one I had used QF (thinking, 'how could contact cement be strong enough?'), which worked and shaped easily, but the contact cement is lighter and was easier to construct. There was no tendency for the pieces to slip in the garbage bag vacuum clamp. I think it takes longer to cure to the same strength, but it held up great while shaping. As they say, listen to the designer.

Free-hand shaping with an electric hand planner is quick and more controllable than you would first think. You can see the desired profile penciled along the stem. Final shaping was with the longboard.

A small hand router with a guided trim bit cuts off the excess deck along the hull. After this I longboarded the upper hull to make sure I was happy with its shape since it will serve as final guide for the next step.

A 45-degree bit with roller guide starts the deck radius. This is quick and makes hand sanding faster and easier to control. I used the same radius sander made from PVC pipe to make most of the edge, but my final radius does change along the length of the boat and is finished with the longboard.

I think I learned on the first float to fair the foam more aggressively. Three-eights foam seems thin to be thinking of sanding too much, but it's the high spots that make the low spots, which will need filling. The risk seems to be that the surface can flex, specifically between the bulkheads, with only the interior layer of glass. Hence the possibility of taking off more over the bulkheads -- while in between, the material flexes away instead of being sanded off. Theoretical or real, I'm not sure.

Finished shape.

I glassed in my usual fashion without any incidents. (Both pictures with plastic still on.) I'm not sure I've mentioned it, but somewhere along the way I've started blowing off the surface with compressed air. Even after vacuuming several times it makes an enormous difference, though it does foul the air for a while.

The skirt of painter's plastic allows me work out the excess resin and not worry about messing the floor, etc. From my experience last time, I would say that the extra effort of rebating the foam for extra layers of glass is well worth it.

Trying to go another step in removing bubbles that develop while mixing the QF, I 'worked' the QF before putting it in the bag used to make my stripes. It helped, but at the cost of a smaller time window to apply. I'm not convinced I'll always do this.

The instructions say to apply multiple thin coats so you can minimize the bubbles. Of course I'm trying to avoid the multiple applications, and in particular the stripes are not 'worked' after applying. Would putting the material in a brief vacuum reduce them? In the end, fixing the few that are apparent after sanding is not really that big of a deal.

I advance guides forward to help keep them straight and the proper distance for my trowel. Quick and easy, but the real value is how this helps in later stages.

This was the real test for rigid vs. flexible longboarding. I went after the surface with the rigid board, then checked it with all my straight edge tricks. **SUCCESS!** Fast, easy, and 100% fair! None of the fiddle from before. I would strongly endorse using a rigid board.

'Working' the material seems to remove the bubbles. A 45º action decreases the chance of getting air trapped along the stripe. A word of caution – it doesn't take long for the QF to become 'dry' in how it behaves with the laminate. What I mean is that it is still workable, but if you spread it out and then remove it to examine the surface, it will not have as wet an appearance as it does when spread just after mixing. I again recommend reading Wayne Hick's treatise, part of which I will quote:

"The most frustrating part is getting the micro to spread out evenly without it tearing, lifting up, or rolling up into a ball behind the trowel. The trick is to go slow enough to give the micro time to spread out from under the trowel and stick to the surface. You'll notice that the micro doesn't want to stick at first, but give it a few seconds. The epoxy within the dry micro will migrate outward and will wet out the surface. Once this happens, the micro will adhere and will spread out easier. (This is why some builders like the epoxy wipe.)"

So, I guess by his description, I'm suggesting that the epoxy 'migration' decreases the closer you get to gel, and may happen while it still seems to be spreadable. I think his material has a much longer work time.

((DISCLAIMER: My guess is that all builders will experiment with different techniques, and likely come to different conclusions. Experience, materials, tools, personality, and goals will influence choices. The value in reading each other's experience is not to find a recipe, but to have a multitude of potential approaches and 'tricks' to springboard from (or things to avoid). This became very evident when I bought a new batch of QuickFair. I'm not sure what the reason was, but it had slight, but important differences in how it handled. For a while I felt like I was learning all over. Each builder will likely have slightly different experiences and draw different conclusions about how they want to approach similar problems -- by sharing them we all win. My own techniques are a first time builder learning from reading and doing. They will evolve as I push forward, and this commentary is on my experience and choices – which are not to be construed as the right or better way.))

I'm still marking the top of the stripes so that I can spot-sand any areas where the trowel may have floated high, as you can see here.

A reminder why the stripes have to be left high – they're the first to come off when the real fun begins.

Three long evenings after work: stripes, fair and fill, then fair. Looks like I'll make it with a single application/sand, though I won't know till later, as we leave on a trip tomorrow.

My wife and I took a trip to see our daughter in California. While last fall I drove straight down, then explored the Sierras on my way home, this time we worked our way down, then flew home and let my parents drive the rig back up the coast several weeks later.

We crossed to the central Oregon coast (reliving part of our honeymoon, this just happens to be our 24th wedding anniversary).

Then we drove diagonally across northern California. The west coast has a string of volcanoes – here is Mt. Shasta.

My goal was to see Yosemite again, this time in the spring with all the snow melt. So different than the fall! Water everywhere.

Yosemite Falls with Merced River.

Breakfast under El Capitan.

Rainbows at every falls.

Evening glow from Tunnel View, Bridal Veil Falls.

Breakfast looking across Cook Meadow. In the morning, Yosemite Creek was frozen slush, and we could hear loud crashes as large chunks of ice were carried over the falls. Above the 3000+ foot cliffs is nothing but snow and ice. Tioga Pass, which I drove last fall just before it closed, will not usually open till late May.

Hike to Upper Yosemite Falls with Half Dome in distance.

We crossed California yet again to drive the coast. The foot hills down to the San Joaquin Valley were full of poppies and red bud.

Walking against a strong wind at a beach along Big Sur.

One of the few mainland sea lion colonies. Mamas and young, but no bulls at this time of year.

February

Sanding gives you some time to think. So I was excited to start fresh on the second side. I wanted to try a couple of new ideas on top of what I'd already learned. My dream was to get all the QF on and then sand it off, all in two sessions (had I gone from depression to

mania?)

One of my first ideas was to fine-tune the guides. Several issues presented themselves.

1) I felt I was learning what the long board was capable of. 

It's a great 'averager' of heights. While it can do some amazing things, if you give it a variety of heights (or in the case of stripes, even a variety of widths) you will get an average that is some place lower than the lowest height. (Or, "if you give it junk, you get back below-average junk.")

2) Getting out the straight edge is a necessity.

The first time I sanded down the stripes, I sanded till the long board touched everywhere. What became apparent later was that sanding till it looks sanded isn't a very good endpoint. Many parts of the float are nearly flat fore and aft, and you can have a long 'wave' of high and low develop. And, with enough force it seems that the flexible board can dish-out areas (especially if you get confused and start sanding low spots, versus sand the high spots like you should). The straight edge doesn't lie.

I got the straight edge out near the end of the first side, but my thought for this side was to use it on the guides at the beginning.

You can sight along the straight edge (SE), but better yet, listen. After a little sanding, you can rock the SE along the stripes from deck to keel and get a smooth, non-clicking motion. This is the easiest dimension to achieve. Do this fore and aft, and at first you will get a non-uniform motion with one or more slapping sounds. Keep sanding and an eventual fair shape will produce an even motion with an even clicking sound as you rock over each stripe in turn. A slap indicates that the SE is rotating over a high spot or bridging a low spot.

Then I noticed that some graphite from previous markings along the SE was being left behind. I added more along the edge and marked the stripes by using a rocking motion starting at the deck and working down to the keel. The whole side was finished in a couple minutes.

Marks on all adjacent stripes means fair.

A stripe (or two) without marks means it's too low, or one of the adjacent sides are too high.

Sighting along the SE at these locations usually suggested the appropriate action and I developed an easy marking system so I could go along and spot-sand, fixing the problems. With minimal effort I now felt that I was starting with guides that were intentionally high, but fair.

I filled between the stripes as before. I think I forgot to mention that I would sand the edges of the stripes. Not really a fun way to spend an hour, but how could you leave a surface that you know will not bond secondarily?

In the picture below you can see a problem, similar to the other side, of a low area just above the keel. I could see this while I was planking, but any maneuver to fix it seemed to cause more troubles elsewhere, so here I am. Also (as seen in pictures further down), I discovered that feathering the stripes at the deck and keel like this before filling was a mistake. Leave them high until the end.

You can see in the last picture how I marked with pencil along each stripe. After filling, there will be some high ridges that are easy to sand away, but I was concerned about removing as much of the junk as possible before I actually started fairing. My plan was to spot-sand down to the marked surface to increase my odds of starting fair.

You can see that after the first couple passes I need to correct some areas to be at the 'fair guide' height. You can also see why high guides are important. And you can see what is developing at the keel by feathering the stripes before filling. Too much material is getting removed too quickly.

I'm already concerned that I didn't leave the guides high enough. Only moderate sanding and I'm starting to mark spots where I'm concerned that I removed too much material. The glass showing along the keel is the edge that was wrapped around from the other side, so as it shows I'm really just feathering its edge. You can also see how just above the keel, feathering the stripes too low (to clarify, sanded to the correct final height before filling) has caused a row of low spots.

At some point I decided to stop and fill. The rows of little dabs of QF are filling bubbles that were in the stripes. Obviously they showed up during mixing and came out the Zip-lock candy bag. I enlarged these with the Dremel as before.

Below shows an idea for trying to fill and spot-sand a low spot. The goal is to minimize its effect on lowering the surrounding finished height during final sanding. (I find sanding addicting -- a little more . . . Now, before filling I again find myself almost, if not at, the stopping height.) I marked the surface with diagonal marks, sanded for adhesion, then marked the perimeter of the sanded area with a irregular circle. After filling a thin, but adequate amount, I removed the excess outside the circle with a putty knife.

Then I attempted to spot-sand, using the usual motions, until I was just sanding on the surrounding glossy surface left by the putty knife. I also spot-sanded the dabs filling the bubbles.

I think the above ideas were reasonable, but suffice it to say that all my efforts to do it in one big session -- or now, even one spot fill -- weren't going to work. I was forced to stop before I sanded on glass.

What I discoved (not a new discovery . . . sometimes we just have to experience things for ourselves) is that when you first sand through the fairing compound, it's tempting to continue on the epoxy. This hard spot can remain for quite a few passes depending on pressure and grit before any damage is done to the glass. You're thinking, "I'm not on the glass yet!" But of course you've disrupted your fair surface by adding in a new substrate with different properties. The surrounding QF continues to lower at a faster rate than the epoxy. The straight edge never lies. The only way I felt I could take care of little problems here and there was to skim coat again.

Here it is after a couple passes. What you see is a pencil mark through the skim coat, marking an area of concern, i.e. the top of a high spot. As the QF thins it becomes transparent and is a great help in knowing when the stopping point is getting close.

When it was close I changed to 60 grit and called it good. Not quite the one on, one off I had been hoping, but I was happy with the end result.

At this point, I had a major breakthrough.

I flipped the hull upside-down to fine-tune the keel. It seemed I had a couple of low spots, so I filled these, and re-sanded. This time it looked like I had different low spots. I know that one thing affects another, but this seemed ridiculous.

The day before I had purchased a RIGID 3M Hookit longboard, thinking that it would be useful for the flat surface of the deck. I gingerly tried using it along the keel and immediately felt like it might be giving me 'truer' results than the flexible.

Before going too far I stopped to do some research, wondering about this whole flexible versus rigid thing. I had started thinking that curved=flexible and flat=rigid. I came upon a website from the experimental airplane building community, part of Wayne Hicks' Cozy IV project. I read 'Chapter 25: The Contouring Process' several times. Part of me was sad because I had already spent so much time learning some of these things for myself (at least my thoughts weren't far off); the other part of me was excited to try all the new ideas.

I sent Wayne an email, and he responded in a couple of hours.

Wayne, enjoyed your site.

---> Thanx!

I am building a Farrier F-22 Trailerable Trimaran --

---> Cool! I've heard of 'em. I used to race Tornado catamarans.

My question was about using the flexible vs. rigid long board (3M Hookit). I have been exclusively using the flexible for the long convex shape of the outside of the hulls, but recently purchased the rigid as the deck on the floats is flat. Does the rigid board have any role for this curved surface?

----> I have found that the flat board works better than the flexi board when contouring large-area curved surfaces. Here's why - the flexi board is TOO flexible. The rigid board allows for better control. This is especially true if you overlap your strokes by an inch or two. They say the human eye can only detect imperfections in contour to about 3 feet. Even most large-area surface contours are effectively "flat" at that length.

I can't overstate how switching to the rigid long board improved my end result. I would also say that reading 'Chapter 25' is a must for all builders embarking on the fairing challenge.

The first new idea I wanted to try was the "skim coat before raw epoxy finishing technique" for scratches and pin holes (see step 5 in chapter 25). After all the effort of fairing the sides, I was anxious about slathering on some mixture with a squeegee, so decided to try it on the deck UNDER the QF. I had a couple of areas that I could experiment on, that had something similar to pin holes where the area between the glass tows were light on epoxy (and I had previously found that QF was difficult to get into very small imperfections). I did it just like he described, and it gave me enough confidence to plan on doing the whole hull. Here is a close up.

Before it cured I did a skim coat of QF over the deck and transom. I'm still sold on the candy-stripe method for curved surfaces, but I know what Menno was saying about the extra work, so tried without.

It was four days before I got back to it. QF gets harder! Still, the end result worked out well. I played with what direction to work the long-board. The usual orientation -- keeping the long-board pointing towards bow/stern -- could leave a camber in the deck. I countered this by also adding a bow-to-stern motion with the board oriented at 45°, then back the other 45°. The skim coat method worked well, given that I was sure the deck was flat by rebating the foam in areas with two layers of glass and having it flat against the strong back while attaching it.

QuickFair requires several coats of epoxy sealer. Originally I figured that any scratches from sanding would be taken care of by these layers. After reading Wayne's advice, I modified this to include a skim coat of low-density filler with a soft, creamy consistency. I applied this with a squeegee, just as I had experimented with on the deck.

I worked it into pin holes etc., smoothed uniformly, then removed with a forceful drag of the squeegee. What I removed was pushed forward to a new area. I actually used very little material. I did discover that the QF (I assume) slowly absorbed the epoxy, causing the mixture to become thicker. So after moving it forward several sections, I could discard it for new mixture. Using a resin mixture with a long work time was so pleasant after the 10 minute work time of QF.

What a great idea! I was pleased with the results. It seems that my 36 grit has some clumps of abrasive which I'll try to watch for in the future, but I think this technique is a great way to take care of these and other small imperfections.

When this got tacky, I rolled and tipped a layer of epoxy. Seeing that first, shiny coat was pretty fun. When cured enough I put on a second layer. After this layer, all the imperfections were hidden.

Off to storage, switching places with the other float.

The second float seemed so far behind. Luckily I had done all the interior work. Having done this once before, I was able to be more efficient with certain steps. Here I filled screw holes.

Below: fitting my three deck pieces together and cutting out the deck flange for the stiffeners. (I joined the pieces while the hull was upright, but I did the interior taping while the deck was flat on the strong back, just before joining to the hull. I had previously rebated the pieces so as to not make high spots on the finished deck.)

Between steps -- while things cured -- I began to think about other small parts. Below is the mold for the rudder gudgeons.

Since I made two foam bows earlier, it was a simple matter to attach the second before the deck bond had even cured.