Aka Fairing Reinforcements

Applied reinforcements last week.  There are three layers of unidirectional tape down the spine of the curve.

300 gm/sqm cloth draped over beam.

The uni is visible under the eglass.  It will provide lateral stiffness.  I had to fair the edges in a bit as the stack of uni stood proud a couple of millimeters.

The glass covering physically locks all of the layers together to the main beam at the bottom.   More importantly, the glass provides a measure of toughness.  The foam with only a bit of fairing compound over it was quite fragile.  It would have been easy to poke a hole or ding it.  The fairing is now firm, but may still be vulnerable.

The second beam is now mostly shaped and the surface being faired.

Task time: 8 hours

Total project time: 854 hours

Aka Fairings

The outer edges of the aka crossbeams are covered with a foam fairing.  This will allow waves to flow past them easier as well as providing a great deal of lateral stiffening.

The foam is lightweight polystyrene building insulation that was CNC cut by Turn Point Design.

After looking around the internet, I decided to use Gorilla Glue to join the pieces of foam.  The glue is gap-filling, strong and has a similar sanding characteristic as the foam.  Above, is a test glue up.

The foam is carved into sections that are stacked.  Every few feet there is a half ellipse template for a fairing guide.

At the windward end of  the beam a solid piece of wood will be attached and shaped. This strengthens the end where is will be attached inside the ama.

Once all of the sections are fastened in place you start sanding.  I have been using a homemade longboard that is about 3 feet long.   The foam sands best with 40 to 80 grit paper.  It is very easy to clog up sandpaper with the blue dust, so 40 is used for all of the initial work.

The solid wood end is a piece of laminated black walnut.  The initial shaping of the wood was done with a belt sander and a plane.  After establishing the basic shape it is blended into the foam with the long board.

I also built a 6 foot sanding board for leveling long sections.

There can be tearout of the foam while sanding along seams.  The picture above shows the first beam after multiple cycles of applying fairing compound and then removing it slowly.  The beam is starting to take on a pleasing shape.   The next step will be to reinforce the leading edge with unidirectional carbon.

Task time: 60 hours

Total project time: 830 hours

Aka Brackets

The plans call for four hat-shaped brackets on the back of the main bulkheads to carry the akas. These are simple e-glass structures.  Further structural steps on the akas will keep them from laying flat, so these need to be moulded now.

The green tape shows the designed width of the brackets.  The blue tape is a target area for me to layup.  I made them oversize so that alignment issues and rough edges can be removed.

This shows nine layers of wet glass. I used some 4 oz. plain weave left over from another boat.  There was a small amount of colloidal silica in the epoxy to help fill the weave.  The dark area on the right is where epoxy seeped through the packing tape covering the beam below the layup.

This is a rough trimmed part.  This will be properly trimmed when it goes into the boat.  It would be easier to glue them to the bulkhead on a flat table before going into the hull.   The bracket would block access for filleting the bulkhead to the hull, so it will have to wait.

Task time: 8 hours

Total project time: 490 hours

Big Bag

I have been working on the last layups for the main structural element of the beams.  The akas required a lot of fairing due to the stair-step shape of the pultrusion laid on the tops and bottoms. Unfortunately this added a bit of weight and lost some of the advantage of pultrusion over regular uni.  The layers did not quite span the width of the beam.  Some of this may have been caused by rounding over the edges of the wood frame prematurely.  We had not decided to use the pultrusion until after this step was done.  I have not done all the math, but I might have been able to get layers of six rods across if I had had the full flat width of the beam.  But, that might have required rounding of the edges of pultrusion causing a loss of strength.  Future builders should definitely look carefully at the options to cut weight and fairing effort 

I ended up laying battens on the sides to dam up fairing compound.  I then sanded the tops and sides to square corners.  The corners were then true enough to run a roundover router bit down the edges getting an even, eased corner and removing some of the excess weight.

This is 6 inch diameter 14.2 oz sleeve (481 gm/sq m) rather than 11.7 oz called for in the plans.  I had my choice of 7 oz or 14.2 oz cloth, so I went with the heavier.  This again is extra weight, but it will add strength and stiffness.  The sleeve came from Soller Composites. ACP Composites is another source.  They seem to have pretty much the same product line.

The python in action.

The 6 inch sleeve was able to conform well to the variance in circumference on the beam, from 13 to 22 inches.

A shot of the 18 foot vacuum bag.  It took two of us about two hours to wet out the sleeve and bag up the aka.  I bought some extra slow hardener to give us some more time.  The shop was about 65 degrees.  Once the vacuum was on, I raised it to 78 degrees and left it overnight.  The wet beam was quite unwieldy.  It had to be flipped several times.  As with the mainsheet beams, I felt I could not get pre-wetted carbon over the beam.  I got a little paranoid with the epoxy and used plenty to ensure it was wet all the way through.  I figured the vacuum bag would remove the excess.

Overall, I am fairly pleased with the result.  The sleeve is well adhered and the weave is well filled.

I had some annoying surface wrinkles.  I think this is excess epoxy that pooled in some places.  I am able to flatten these with just a few strokes of sharp 220 paper on a flat board.  The dust is white rather than black, so I'm not sanding carbon.  Wrinkles in cloth this heavy would have been much larger. Getting peel ply and perforated release plastic on the compound curves was a pain, especially having to shift the aka multiple times.  I wish I knew more about these processes, so I could get nicer results.

Foam cowlings now have to go on one side before the last layer.  I will probably use carbon or fiberglass sleeves over the foam.  It is smoother and maybe stronger that patches of regular cloth.  The shape with the foam will be even more challenging.  Perhaps it should be infused?

With all the fairing, these steps on the two akas are taking about 80 hours.

Task time: 80 hours

Total project time: 482 hours

More of the Same

There has been slow steady progress.  Most of the work has been on the akas and mainsheet beams and I have posted about those before.

This is the 16th and final pultrusion glue up on the beams.  Nice to have that done.



Here is a quick vid of the outer sleeve going on the second mainsheet beam.

The wet-out went on for a while to force epoxy through the sleeve.

Task time: 40 hours

Total project time: 402 hours 

First Double Bias Layup on Second Aka

Here are a few shots of my evolving technique for laying heavy 12 ounce double bias carbon on an aka over wood and pultrusion.

Cut the cloth.  I was using about 4 foot sections.  On the first aka I used longer ones, but they were difficult to handle.  They also did not want to conform to the curve of the beam well.  They would trail off over the side make for uneven overlap points.

Prime the raw plywood on the beam.

Start wetting out the cloth.  This stuff soaks up crazy amounts of epoxy.  You need to use plenty and be very patient about getting it thoroughly wet and even.

I spread a wave of epoxy back and forth making sure to get all edges.  I have been using a very liberal amount.  Excess epoxy rolls off the cloth, but can be scavenged back to the middle.  After one side is wet I flip the cloth and start over.  Usually the other side of the cloth is only about 10% wet when turned.

Once both sides are quite wet, I fold the plastic over the cloth to create an envelope.

The envelope is squeegeed with heavy pressure.

The squeegee creates a wave of air and epoxy.  It is very difficult to see how wet the cloth is prior to this.  I pull the epoxy in about six directions (+-90 and +-45).  I flip the envelope and do it again on the other side. At this point, I'm pretty sure that it is wet all the way through.  I carefully peel back the envelope and the peel off the cloth.  This needs to be done with even pull as one does when removing tape.  Uneven peeling will distort the cloth and fray the edges.

The cloth is transported gingerly to the aka.  I have been overlapping about a half inch.  The wet cloth is firmly milked  across the beam with the gloved hand  I Want it tight and smooth.  

Later, the rise in the overlap will need to be carefully ground back smooth. I only want disturb the second layer and feather it in.

Peel ply is stretched across the surface.  It gets wetter than this picture as it is worked to get the air out smoothly.  The peel ply is a lot less compliant on the curves than the carbon and usually needs a few darts along the sides.

Task time: 4 hours

Total project time: 362 hours 

Aka Wrap

Here are a couple pictures of the first layer of bidirectional carbon wrapped around an aka.  I start laying up more pultrusion tomorrow. Two more layers on each side and then it gets wrapped again.

Task time: 24 hours

Total project time: 382 hours

Boring Beam Mounting Holes

The Akas are attached to the main bulkheads with 18mm bolts.  The bolts are carried through the beam holes with a 20mm (3/4 inch) G10 tube liner.  I needed to bore the holes through the solid interior of the beams before the first layer of carbon cloth.  CNCed holes are marked on the plywood sides and I did not want to lose the reference marks under the cloth.

This is one of those seemingly simple tasks that takes a lot of pondering and worry.  The holes must be perpendicular to the beam all the way through and be accurately located so that they will mate properly with the bulkheads.

All kinds of little issues arise for consideration:

- How do you level an 18 foot beam with one end across a drill press table?

- There is not enough throw on the drill press without shifting the tables and re-leveling in the middle of a hole.

- How do you drill an accurate hole without using a drill press?

- The forstner bit is not long enough.

- The forstner bit extension does not fit a metric forstner bit correctly.

- The collar on the bit extension does not fit down into the  hole.

Finally found a drill stand for a hand drill, but would it be accurate enough?

G10 (Garolite) Tube

Setup that allowed a hole "almost" deep enough.  It would not reach the last millimeter, but the point of the bit went through for reference.

A 3/4 inch hole that has been sanded slightly with a dowel covered with sandpaper.  The 3/4 inch tube fits snugly through most of the hole.  There is a little extra room at the top where the original hole in the plywood is slightly bigger (2mm), but the tube is squared in all directions.

All you can say when you drill that first hole and it is OK is whew.

Task time: 2 hours

Total project time: 311 hours

Pultrusion Dispenser

I have done five pultrusion layups on the aka beams since returning.  Having run out of cut pultrusion, I had to open a new 500 foot roll.  In order to avoid the "pultrusion hairball" that I discussed in a previous post, I quickly hacked together a dispenser.  It would work better if the inner dowels actually turned, but it gets the job done.

Task time: 25 hours

Total project time: 285 hours

After an Absence, A New Clamping System

This Spring and Summer, I spent quite a bit of time refitting our large boat and then I spent some time out on the ocean.  I'm back at work now.

The first order of business was to build a new set of clamps to handle the aka layups.  Russell and others had lent me "just about every clamp in the neighborhood".  The problem is that active builders need clamps back sooner than later.  The akas  need about thirty six clamps and when you multiply that by the price of a Bessey or Jorgensen clamp it begins to hurt.

I decided to make a set of clamps using cheap pine, allthread and wing-nuts.  These clamps are not as versatile, but they will do the job and cost less than three of those store bought types.

Recently, I took a friend by the Northwest School of Wooden Boat Building  where I saw some nice long reach clamps.  Maybe some say I'll have time to make some of these.

Task time: 6 hours

Total project time: 238 hours

The Slows

I have been busy with some other projects and have only spent about 20 hours on the boat the last couple of weeks.  Tasks include more layers of uni carbon protrulsion and fairing corners of the pultrusion layers into the beam corners. The stacks of pultrusion create gaps that have to be faired before layers of bi-directional cloth can be laid over the top and bottom of the aka beams. You get glue squeeze out under the clamps and it makes for inconsistent edges.  I'm doing some test layups to see if I can find a more uniform way of doing these.  Pictures later...

I have calculated there are about 20 more carbon layups for the beams.  About 16 of these require hand clamping every six inches.

There is more information on the Cheers posters at PT Watercraft.

Task time: 20 hours

Total project time: 232 hours

First Layer of Uni Pultrusion

Russell helped me to get the first layer of uni on yesterday.   Most of the task is preparation.

18.5 foot sections of pultrusion are cut off from the roll, cleaned, taped, sanded, vacuumed and wiped.

The first adventure was cutting the pieces off the 500 foot roll.  I was lulled into feeling safe as I cut off the cable ties that held the roll.  The roll laid there calmly. I had expected a little more reaction from releasing it.  Since it seemed under control, I neglected to add any different ties or restraint mechanism. While pulling out a length, the roll slipped off the table and exploded.  It suddenly expanded into a 15 feet wide and 10 foot high hairball.

After cutting of the first 5 strips, there was still about 400 feet of tangle.  Sorting it out was futile. I would pull out 20-30 feet and find a knot.   I will need to pull all of the aka strips and cut them.  This will consume most of the roll.

I will need to build some sort of reel with a cage for the next roll.  If you use this stuff, fair warning.

The strips were clamped to the table to measure and cut them.   This is a bit of a wrestling match.

The next step was to wipe them down with solvent.  There was dirt and a bit of adhesive from duct tape at the factory.  I used denatured alcohol.  Russell thought acetone would be much better.  Ether of these need to be sanded as they leave residues.  Apparently the gold standard is medical and/or HPLC acetone which is pure and leaves no residue.  Big time Grand Prix builds use this at over $200 a gallon.

The next step is to tape the back of the strips with packing tape to hold them together.  The tape is placed perpendicular to the strips.

Sanding was done with 150 grit paper.  Russell warned me that different paper brands and grades can leave residues on composites. I went with Festool Granat which is designed for composites.

An area that has not been sanded.  The sanding should remove all of the shiny epoxy surface without digging to far into the carbon.

Sanded Strips.

The next step is to wipe, wipe and wipe again with clean cotton until the carbon dust is gone.  I found that dust gathers between the taped strips and the unit needs to be flipped to get it out.

Why all of this cleaning?  The optimal bond is very sensitive to dust, oil (off hands) and other debris. The akas are one of the most highly loaded areas of the boat and therefore sensitive to failure from a poor bond.  Believe it or not, the wooden box of the beam could still flex along it's length before the carbon went on.  We supported the curve where the beam rose off the table with legs.  The weight of the clamps could have deflected the curve of the beam.

The beam and carbon are primed with un-thickened epoxy. Then a normal bonding batch is placed on the beam with a notched trowel.

The beam after glue up.

Clamps were placed at 6 inch intervals.  Aligning the strips in the center of the beam is an issue.  We taped the carbon down temporarily to keep it aligned.  Clamping was done from one end to the other to remove any slack and prevent bridging on the curve.

I made an attempt to make guide blocks to keep the strips aligned, but they did not turn out very well. See photo above. I did not make them big enough and they had to be glued on the rounded over corners of the box. They were not very accurate.  It would be best if there were some sort of fool proof jig for this, so you would not have to rely on aligning by eye.  The buttered pieces are also very slippery.  You have to watch that tightening clamps does not  slide the strips one way or another.

Task time: 5 hours

Total project time: 212 hours

Bulkhead Reinforcement Templates

The bulkheads at both ends of the main cabin get reinforced with 19mm timber.  The bolts that hold the akas to the vaka go through these points.  I will be using yellow cedar for these.  I made some templates out of MDF today to insure a good fit before cutting expensive wood.  These will get screwed to the cedar and will guide a flush trim bit on the router table. I should get exact copies in cedar if all goes well.

These bulkheads get assembled and finished before inserting them into the hull.

Here is the interior of Jzerro again.  You can see similar bulkheads at the end of the cabin.

Task time: 3 hours

Total project time: 197 hours

Slow Progress

Life got in the way and between a trip to get materials and other responsibilities not much got done for four days.

I did do some prep for the next steps on the akas.  Surface sanding and rounding over the edges.  I have also been pondering the best methods to apply layers of strips of pultrusion and 400 gsm carbon double bias.  These have to be clamped and/or vacuum bagged over the eighteen foot span.

A layer of five strips of pultrusion will look something like this only longer.

Current state of the beams.

Also got a stack of douglas fir ripped for the mainsheet beam laminations.  These were a bit wet and will need to dry with the heaters and a fan for a while.  Think I need a moisture gauge.

Task time: 8 hours

Total project time: 194 hours