Episode 10 The keelbox reconstruction

I have procrastinated on this post long enough. I tried to put it off with stories about my Cousin Mary and testicles. The reason for my tardiness is that I need a drawing to illustrate what I'd like to describe. Having failed to draw something meaningful, I now have to soldier on with words alone, helped by the photos at the end of this post.

Pulling, and sometimes smashing, the boat apart has revealed much about the way it was built. The engineer in me set about this adventure with the innocent belief that the boat-builders who built it knew what they were doing and, happily, this belief has been mostly justified. I say mostly, because in some aspects of the boat's construction the exigencies of selling the thing for more than it cost have manifested in some surprising corner-cutting.

I'll try to describe the keel box construction starting at the bottom.

The keel itself is a cast iron plate, about 38mm (1.5") thick, 1.78m (5' 10") long and 0.45m (18") wide. It weights about 210kg (463lbs) and pivots on a keel pin at the bottom of the front of the keel box. When in the up position, the keel lies on its leading edge and slots fully into the keel box.

The keel box is a long, thin box, about 40mm (1.6") wide and up to 750mm (30") high. It is made with a 3mm (1/8") smooth GRP inner surface, with a 12mm (1/2") balsa wood core and a 2mm (1/10") outer GRP surface (see the cores in the photos below). The sides of the keel were probably cast onto a flat surface before being turned upright and fitted into the boat.

At the front of the keel box is a small pine bulkhead, which joins the pine table on top of the keel box. This was the pine table that I smashed earlier. There are two things that fit under the table in the slot of the keel box. One is the sheave assembly at the front of the keel box and the other is a pine timber spacer, 40mm wide that runs from the sheave assembly to the rear of the keel box. The sheave assembly comprises some stainless steel plates in an L-shape that hold two sheaves (pulley wheels) over which the keel line passes that raises the keel. The sheave assembly was held by three bolts that passed horizontally through the keel box, two 6mm (1/4") and one 8mm (1/3"). The table was screwed onto the pine spacer, and this was held in place by about six screws, screwed horizontally through the sides of the keel box

On top of the table is the compression post, which is a vertical piece of pine between the table and cabin ceiling, just under the mast. As the name suggests, it takes the compression forces from the mast and transfers them down to the keel box. Without it, the mast would punch through the cabin ceiling. The upper end of the compression post is held in place by a pine socket, which is in turn held in place on the ceiling by the four bolts that hold the alloy mast step, which holds the bottom of the mast in position on top of the roof.

Starting at the top, the forces are transferred downwards as follows;
Mast > mast step > cabin roof/ceiling > socket > compression post > table > keel box.

When the boat was constructed, it appears that the quickest way to put it together was to put the table on top of the keel box, put the compression post on the table, get the compression post to hold the socket in place on the ceiling, then to drill through the holes cast into the mast step, though the cabin roof, through the socket, and then to bolt the socket in place. The reason I think this is that when I took the socket down, I could see that it was about 5mm off-centre and about 10mm to the rear of the centre of the four bolts coming down from the mast step, which is what you'd expect from two blokes trying to get the job finished on a Friday afternoon.

Although the keel box assembly was nice and stiff, it had some inherent problems.

The first was that there was no obvious way to take it apart, which led me to smash the table earlier. There was also no way to take off the table without removing the compression post, and the compression post could not be removed without taking off the mast step. In all cases, none of this was possible with the mast up, making the maintenance of the keel impossible without disassembling a large part of the boat.

The second was in the bolts that held the sheave assembly in place. When the keel was being lifted, the sheave assembly took some very heavy loads, maybe up to a tonne. These loads were carried by the three bolts passing through the keel box. The bolts themselves were OK, but they were supported by the GRP-balsa-GRP sandwich of the keel box, which was stiff but relatively soft. Over time, the bolts had enlarged their holes (see photos below), allowing the sheave assembly to wiggle a little, which could have led to asymmetric loading and warping of the keel box.

I had also smashed the table, and needed to make a new one.

It seemed like a good idea at the time, so I made a new table out of merbau hardwood, rather than replace the pine of the original. I thought this would give a more durable finish (you have to step onto the table to get into the cabin) with a stronger material. Further, merbau decking comes in a size that almost matched the pine table, and it had pre-made chamfers. Being unemployed at the time, I bought the shortest length I could find, and must have used it all, except for a handful of shavings. Having started with merbau, however, I had to use it for all of the timber around the keelbox, including a new compression post, socket and bulkhead. The latter was a shame because it was the only piece to remain in tact after I had pulled everything apart.

My new design allowed for the removal of the table with the mast up. The new table comprised two pieces, the larger rear piece extending from the compression post to the rear of the cabin, and the smaller front piece covering the sheave assembly. A rectangular hole in the front piece held the compression post in place, and the compression post sat not on the table, but the sheave assembly underneath. I kept the old pine spacer, because it was the right size and would be hidden from sight. As a finesse, I fixed the sheave assembly and pine spacer to the table tops by screwing them in from underneath, which hid the screw heads from sight, too. All new screws and bolts were in stainless steel, except for the four screws that held the front bulkhead, where I used brass instead.

The merbau hardwood was harder to work than pine softwood, mainly because every screw hole had to be pre-drilled. I also varnished everything. I need to improve my varnishing skills because, no matter what I did, I still ended up with drips on the undersides of the pieces. I had learned to apply the varnish in layers, and to sand each layer with 240 sandpaper, which I did. I also found I had to wait at least three days between each coat before it became strong enough to sand, which lengthened the process considerably.

To reinforce the bolt-holes, I decided to make grommets out of the merbau hardwood. I had previously acquired some core drills, useful for putting latches on doors and I used them to core-drill out the old bolt-holes. I then used the same drill to drill a grommets from the offcuts from my merbau table, which fitted neatly into the holes in the keel box, where they could be epoxied into place. These grommets spread the loads from the sheave assembly from the old arrangement that has nothing but the (e.g. 6mm) bearing onto the GRP-balsa sandwich, to the width of the grommet (25 to 40mm).

Finally, I needed a waterproof seal between the keelbox and table. Previously, the seal had been made up and patched with a mixture of sikaflex, wood filler and bathroom silicone. The latter drew much bad language from me because it was so difficult to remove. My solution was to use a single strip of neoprene around the perimeter of the keel box, held to the keel box by a sikaflex bead. This meant that the table was not chemically bonded to the keelbox, thus easing its removal for maintenance.

The final result was a snug assembly that allowed me to remove the table over the keel without having to rip out most of the guts of the boat. I've still got some lumps where the varnish has dripped, but the finish isn't too shabby.

Austral 20 keel box reconstruction: Old bolt holes after being core-drilled out. The holes have been enlarged by the abrasion from the old bolts under keel-lifting loads

Austral 20 keel box: Bolt hole cores, showing the cross section through keel box walls, which comprise 3mm GRP on the inner face, balsa core and 1 to 2mm GRP on outer face.

Austral 20 new ceiling socket for compression post. The timber infill was needed because the first attempt at the hole was in the same location as the old hole, which not centred under the mast on top of the roof.

Austral 20 keelbox reconstruction: new neoprene sealing strip on top of keelbox looking aft

Austral 20 keelbox reconstruction: Forward part of new table, which holds the new compression post in place, which sits directly on the stainless steel sheave assembly on top of the keelbox.

Austral 20 keelbox reconstruction: completed assembly, with the exception of the side-leaves for the table.