Saturday, January 31, 2009


I took a couple weeks off around the holidays. This marks a year of work -- I'm still having fun, a fact I attribute to a generous and understanding wife and family. And a well-thought-out design and set of plans. The only times that I haven't enjoyed myself so far are when I find myself stressed about the 'schedule'. My New Years resolution is to continue making goals, but not worry when I fall behind. When I started, people would ask, "how long?" and I would say, "a couple years." I'm still responding, "a couple years." Life happens, and I don't want to miss it. 
I did a wet lay-up for the wingnet rail -- peel ply, cloth, peel ply. With a little help I moved the whole lot onto the form, then used my usual plastic and squeegee to compact the layers and remove bubbles. Worked great.

I mapped out what size I wanted the cut-outs to be. The important areas seemed to be the rigging and hatch. One thing to remember is that the deck/float is tilted, so the clearance for the rigging is better than it would first appear. Like severals others, I chose the Tempress 1317 Marine Hatch. It fits the space, has two cams, and a gasket around the edge. It actually looked better (stronger) than I thought it might. Hopefully it will hold up to the UV and not leak too much. It's much lighter (including on the pocket book) than the aluminum/glass options that I saw, and I didn't really want to make my own.

Using double stick tape to hold sandpaper around the circular template makes a great sander.

(There was a reason that I stopped and made the wingnet rails before glassing the deck. This is not in the plans and I won't know if the effort is worth it, or what trouble I've caused myself until later. I evaluated where the feet of the wingnet rail would fall -- they happened to be over stiffeners and doubled glass, or bulkheads -- so I decided to rebate the deck and potentially make the fairing easier. This isn't something that I'm advocating, I'm only admitting to it since it's visible in some of the pictures.)

My plan had been to vacuum bag the deck. It would be a change of pace and seemed easy enough. I placed PP around all the edges to make the seal good and the tape easy to remove. I cut all the supplies, mixed some resin and had at it. Luckily I started at the stern, where I saw the air vent hole which I had forgotten about (remember Jay's experience with the bloated float and Tor's with the collapsing float . . .). I did a quick switch-a-roo and started using the cut bag material to do my usual PMVB. My first thought was that this was going to be the 'premium' material, but remember how my stretchy VB material affected my vacuum infusion results? Well, the force of the squeegeeing action stretched the material (which was not evident at the time). Then, since the edges are not controlled, it gradually pulled back, making enormous wrinkles! But what seemed like a nightmare turned out like any other wrinkles I've had which are caused by the plastic -- a collection of resin above the PP which just pulls off.

The lesson is that thin or stretchy plastic, which at first might seem good because it can conform to curves, actually is not what you want if you are going to work the surface with a squeegee. Thicker plastic cut into appropriate sized, overlapping strips is best.

I insulated the carbon of the chain plate cut-out at the same time.

I feathered the glass over the radiused deck edge with the long board. Watching how fibers with different orientations show up during sanding helps in deciding how much has been removed.

A great moment -- removing the PP from the side of the hull. It's hard to capture the shape in a picture, but I'll tell you it looks nice.

My second float is stored outside with the interior finished but the deck/bow not yet attached. I plan to fair and seal this one, then switch them out.

I'll admit to having a sense of dread about the fairing. It seemed that everyone talked about how it was a nuisance, a lot of work, etc . . . but I hadn't picked up enough facts or tips to feel like I could just jump in.

First, most people end up trying a couple methods, then settling on something that matches their personality and material. Some, the 'sanders', tend to put on most or all of the material first and then sand. Others, the 'fillers', will sand then fill, sand then fill . . . I thought of myself as a little more of a sander than a filler.

The general advice is to put on some limited material which is faired -- and acts as a guide when applying the bulk of material. This is to limit effort and waste (QuikFair - QF - and I'm sure all premixed materials tend to be very expensive).

Then of course, you have specific limitations from your fairing material. QF has a limited work window -- 10 minutes -- being hand-sandable at 3 hours @ 70º F. It has to be sanded between applications for adhesion purposes. When I first contacted System Three they said, "there is no window of time which allows a second coat, and it MUST BE SANDED" (capitals are theirs).

Ian's building manual describes one way to produce guides. Put on the first layer with a notched trowel, sand the tops to achieve a fair surface, then fill the grooves before the final sanding. While this might work if done within a cure window for some systems, QF did not seem to be one of them. Since this technique was mentioned in the System Three Epoxy Book also, I contacted S3 about my concerns of the interior grooves not being sanded before the second application. Their response:

"You raise an excellent point. But there is a way around it. Before sanding, scrub the QuikFair with vinegar then rinse with water. Allow drying. This will make the unsanded QuikFair surface bondable with additional QuikFair. Vacuum the dust before applying more QuikFair."

Two things made me stay away from this. One was this comment by Oliver Blanc (a Scarab 22 builder, made on a nonF-boat site), "Be careful with that method. The second application of bog does not bond well inside the grooves of the first application. I have seen old cedar strip F-9 floats done that way that had to be redone after a few years." And second, I have a house that has peeling paint no matter what you do . . .

So I decided that if I used guides, I would do a candy bag method similar to Henny and others.

I reread several blogs and decided to contact Menno for some specific advice since he had just finished final fairing. Here's his email reply:


As you can probably see reading my blog I'm not an expert building with fiberglass, so I can't guarantee I did it the right way. Anyway, here are some thoughts I have on the fairing.

I stopped using the candy-bag method because it was extra work and I didn't really notice the benefit of it. Also the guide lines tend to cause a lot of voids and cracks in the putty along the lines because when 'filling up' air tends to get trapped alongside the lines. Hope it is clear to you what I'm trying to explain. It's not fun backfilling all the small voids after sanding. I guess using the 'notched spreader' method will also give some problems with voids.

If you want to try the candy-bag method, make sure to put the lines not too far apart(when I would do it again I would go for about 10 cm apart maximum) and make sure they run all the way from the top to the bottom; lines with a gap will cause the trowel to 'hang' and give nasty depressions which you'll have to back-fill later on.

Just putting on a fairly thick layer of compound (using a big sturdy trowel, dry-wall style; you'll get better at it after some practice) and then sanding back as far as possible with the longboard (until you're almost at the glass) worked fine for me. It's probably a bit more wasteful then the candy-bag method. I wouldn't worry about the possible extra sanding that's involved. The first rought passes don't take up that much time. With a longboard and grit 40 you'll remove the bulk of the compound in (relatively) no time anyway. I found the last part of the sanding most time consuming.

With both methods (with or without candy-bag) I felt I was using quite a lot of compound per square meter, but a lot is sanded off eventually. On the cutout sections of the hull (windows) it's clear to see there's only a minimum of faring compound left.

An idea I read about but didn't use - maybe I should have - is to put some dark color on the glass before putting on the fairing compound. That way it's better to see when you are almost through the compound and sanding glass. When sanding with the rough grit sometimes you're there before you know it, and it doesn't take much to sand through the glass.

Sanding. Of course you'll have to use a longboard. The float sides have little curve, in fact they are almost straight for long stretches. You can use a big longboard for those. I did, but I'm not sure it's necesarry. On the main hull I mainly used a smallish longboard (about 50 cm long and about 6 cm wide) because the bigger boards would just not fit in all the curved sections. Final paint is not on yet, but I'm quite sure it's all fair. Because of the vertical planking the hulls should be fairly fair anyway, and with the 'fairing' you can only get rid of the last small imperfections.

I did all the sanding diagonally. Start at one end with the longboard just pointing straight ahead, make a diagonal sanding movement (keeping the longboard pointing forward), put the longboard about 2 cm further forward, repeat till you're at the end, then do the same thing with diagonal strokes in the opposite directions (thus sanding an 'x' pattern), and then again the other way around, etc., etc.

I hope this is of any help. If it is let me know, maybe I can put this text on my blog or the builders group....

Good luck building

(As you can see, since it was a great help, I took the last line as his permission to publish this.)

So I decided to try a small area. I knew that the area where the foam bow and hull attach was slightly low, and this seemed like a place to start.

I hadn't gotten far before I had my first feelings of foreboding . . . how thick was was this material going on? Even with a wide trowel it had grooves and transitions. And although I didn't notice them at first, bubbles. Needless to say I was glad that I was only trying a small spot. Then I sanded . . . and sanded. I started with 36 on a flexible 3M Hookit board. An early mistake was changing to 60 too soon. And bubbles! I extrapolated this experience to the whole boat, and was quickly depressed.

I was telling someone at work the next day about my experience [I say 'telling' because I know I wasn't complaining ;-) ] and the response was, "You knew when you started the project that you were going to have to do this, and it's just part of the process, right?" 


"Then it sounds like you need to suck-it-up-and-drive-on!"

I'm not sure where that comes from, but I've felt better ever since she gave me that advice.

Here's the section. I was pretty happy with the shape, but didn't know how much further I was going to have to sand -- and bubbles! I decided to try something different and come back to it.

I might be a 'sander', but I needed guides. I heeded Menno's advice about straight, close, and continuous, using a Zip-Lock bag to place beads of QF.

And then I sanded with the long board. I knew about a couple problems near the keel where it was a little concave, and nothing shows that like sanded candy stripes. And there were other areas, nothing drastic, but the matte surface left by the PP hides the imperfections that need to be taken care of.

Then I filled between them with a narrow trowel. The process seems to have two parts: getting it on and giving it a finished surface. Applying it at 45º to the beads minimizes the air trapping along the edge. Working the material helps remove any bubbles (all the product instructions say, "mix so that no bubbles form," -- good luck -- when the clock starts ticking and you have ten minutes to get it in the bag and on the surface . . .). A final pass with minimal material on the trowel gives a nice finished surface.

Other blogs well describe how the material will be low between the guides. This has several causes:

1) Flexing of the trowel (and the angle of attack).

2) The curved hull areas are convex, so by definition the area between the stripes will tend to be too low if you put a straight line between them.

3) And the viscosity of the material affects how it stays behind or gets carried along. If you're moving too much material, it will begin to 'roll' under the trowel, and the mass of material effectively leaves the surface depressed.

Some material left behind on the trailing side of the trowel is OK, but don't accept low spots -- like tipping paint, if the material is still workable, keep at it. You want it to be right. Obviously it took many batches because of the short work time.

Here you can see large areas where the guides were sanded too low and I couldn't reliably leave material behind. It was tempting to let the trowel 'float', leaving more material behind, but unsystematic high spots don't help.

After fairing, it seemed so close. But I had many low spots and the bubbles near the bow to fill (I had sanded that down till I could see glass in areas, then opened and enlarged the bubbles with a small burr on a Dremel). I decided the best way to solve all these problems was to prep the low spots, then skim coat the whole side.

I learned several lessons here. This material likes to be worked! A soft touch doesn't give the best result. This is when I remembered, "thixotropic: the property of becoming less viscous when subjected to applied stress, shown for example by some gels that become temporarily fluid when shaken or stirred." Previously, I had really only thought of this in the 'can be applied, but doesn't sag' version. Now it was, 'push harder so it flows better'.

I found that putting parallel beads of QF controlled for an even volume. Adding a 45º crisscross motion to the trowel, vs. just deck to keel, also increased the uniformity of the skim coat. I finished with a continuous, firm-pressured swipe from deck to keel.
Sanding: 36 grit, board parallel to the deck edge, with 45º motion. Start at one end, advancing ~1/2 inch with each 45° swipe until you get to the other end, then return the other direction with the alternate 45º.

It's taking shape. It goes slowly at first (a low actual surface area of board touching because of high spots?), then the material starts to disappear faster. It still takes hours. If I felt discouraged or felt like stopping, I found that changing to a new piece of sanding paper would improve the speed and my outlook.

I decided that, in order to get a uniform finish, I needed to take care of any low spots early. You can see some below.

It's easy to get focused on the low spots when sanding. But remember, you can't sand a low spot -- you can only sand the high spots. Truly low spots have to be filled. Since the last step is sanding, not filling, I think stopping to fill early is likely better than hoping the low spots will disappear before you've reached glass. You have to stop when you see glass, though if low spots remain it can be so tempting to try just a little more.

I probably filled some that were going to disappear before I stopped sanding, but I was becoming determined that this was it, no more rounds of QF!

Tired sander, but happy to have sucked-it-up-and-sanded-on. (And yes, it seems I tried a couple of different methods too.)

Several interesting observations from my patch-work of filling.

1) By the time I had spot-sanded the 'low' spots for adhesion purposes, they were now low for sure.

2) When sanding at this stage, more than just the high spots (which are now the slightly proud filled areas) get removed. The grit also moves across the previously faired material, making it go down also. The high spots are removed, but not at the same rate as during the initial sanding. Said another way -- after filling, the new faired shape/depth will be lower than where you stopped before filling. So, fill early.

No comments: