Summer Stop-the-Leaks Refit Project

The old phrase "You can never be too thin or too rich." shouldn't apply to the deck under load-bearing fittings. For example, I was curious about the best place to attach a cruising spinnaker's tack. The answer back was to use the D-rings located on the bows. I was originally a little hesitant because the fiberglass looked a bit thin. I didn't know how thin.

We did a summer refit on LUX, addressing a number of projects, including fixing the small leaks from the hardware of the bow-mounted D-rings that are used to attach the snatch blocks that we use for the spinnaker tack lines. The D-rings came off easily; no problem there. Just be careful to not drop the small rubber piece that keeps the ring from vibrating against the hull. However, we were surprised at the resulting fiberglass thickness--about 3/8 inch! We also noticed that there were some gelcoat cracks on deck around the fittings. We don't recall those being there when we purchased LUX. Inspection of the aft D-rings showed similar cracking of the gelcoat. Removing those fittings showed even more cracks and thin fiberglass. A significant amount of the thickness of the deck at the aft fittings was the layer of gelcoat. The resulting fiberglass was only about 1/4 inch thick! No wonder it was cracking!

Cracking around D-Ring Aft

Cracking around D-Ring at Bow

English Wheel

We had some 1/8 inch thick aluminum plate left from making fuel tank inspection plates, so we cut pieces for the bow and stern. The bow plates were cut as trapezoids 6 inches long, 2.5 inches wide at the forward end and 4 inches wide at the aft end. Next problem: how can we get the plates properly shaped to fit into the tight curve of the bow? Well, we don't have an English Wheel.

Shaping the Bow D-Ring Backing Plate

But we do have a ball peen hammer, vice grips, and a tree stump handy. Just grab the aluminum plate with the vice grips and start hammering. We started hammering a line down the middle and expanded to either side. Sure enough the plate started to take shape. Four trips between the stump and LUX resulted in a good fit.

We cleaned up the inside of the hull with a rotary tool and a sanding drum, cut two layers of fiberglass mat and epoxied the mat and plate into place inside the hull. Thickened epoxy was used to help fill the gaps between the hull and the mat and the plate, providing a good distribution of load to the hull. The plate was pressed into place with a stick, forcing excess epoxy out and making sure that the plate was well supported.

The deck cracks were repaired with new gelcoat. (We've still not determined the mix of colors used in the off-white deck gelcoat.) The D-rings and screws were cleaned up on a wire wheel. It turns out that the metric screws had to be replaced with longer screws, which we were able to obtain from Fawcett's in Annapolis.

The D-rings were reinstalled using butyl tape as described by Compass Marine. http://www.pbase.com/mainecruising/rebedding_hardware

We had to bed the bow D-rings twice because we didn't use enough butyl tape the first time. The hull curves under the D-rings, which makes sealing them a challenge. One of the bow D-rings needed the gelcoat sanded a bit to provide a flat surface against which the D-ring would properly seal.

Starboard Aft D-Ring Backing Plate

We used the same process on the aft D-rings, making differently shaped backing plates and cutting a little of the old fiberglass around the fitting to allow the backing plate to fit. Some shaping of these backing plates was also needed to adapt to the curve of the balsa core in the decking. (We had to remove the head liner in the aft cabins to perform this work.)

Cabin Top Hardware

While doing the D-ring work, we had a nice rain storm come through and we found some more water dripping from the headliner in the main salon. When we removed the headliner, we found that it had been leaking for quite some time. It looks like most of the time the headliner absorbed the dripping water and no leak was apparent. Over time, the water started to rot the headliner plywood. We used Git-Rot epoxy to seal and stiffen the damaged headliner plywood.

Water Stains on Plywood Headliner

The water was coming from the hard-top support struts where they bolted to the salon deck. The struts have backing plates, which we thought was a good idea. We removed the struts and found that the holes through the hard-top had also been leaking, so we removed the wet balsa and dried out the holes with a small computer fan that's powered from a 12v wall wart. It takes about 24 hours to completely dry. Note the wood blocks that supported the hard-top while the repairs are being done. When the holes were dry, we filled them with epoxy thickened with colloidal silica.

Computer Fan and Power Brick for Drying Balsa Core

Hard-top Strut Backing Plate

The backing plate for the port hard-top strut didn't sit flat, relative to the bolts, adding strain to the setup. We cut a piece of oak flooring to fit under the backing plate so that it would sit square to the bolts.

We also noted water stains from some of the turning blocks on deck, so they came off too. Our set of balsa removal tools received a workout on all the deck holes and hard-top holes. One of the tools is made from a coat hanger. The other two are made from steel wire. The round handle on the thicker tool is easier on the hand than the T-handle on the medium tool. The end of each wire is shaped into a chisel wedge by hammering it on the anvil section of our big bench vise at home, then sharpening it with a file. 

Another tool is made by cutting off the head of a nail, bend the end 90 degrees, hammer the end flat, and sharpen it with a file. Chuck it in a drill and spin it around in the bolt hole to cut out the balsa. We have several of these tools with different length cutting arms.

Balsa Extraction Tools

Port Cabin Top Turning Block Backing Plate

Some of the turning blocks only had fender washers and nuts holding them in place. No wonder they started leaking! So we made and shaped backing plates for them too. This plate is on the turning block that's above the electrical panels on the port side of the salon.

There is a whole series of blocks on the starboard cabin top to lead the lines into the set of line clutches and the winches at the helm. We created one big backing plate for all three stacks of blocks. The other backing plate in this photo is for the starboard hard-top strut. We didn't find much wrong with the backing used for the line clutches and they weren't leaking, so we left them alone. We added another clutch for the spinnaker sheet from the port side. 

Starboard Cabin Top Backing Plates

All cabin top and hard-top holes were filled with thickened epoxy.

Toe Rail Leaks

Another storm or two later and we identified leaks along the outside of both hulls in the forward cabins. These took a while longer to diagnose. We rebedded the bow pulpit railings on both sides with butyl tape. That helped, but wasn't the only source of water.  To simulate a rain, we directed a hose set to light spray onto the toe rail. This allowed us to localize the spray and help identify the leak. This time it was the toe rail bolts. We could see corrosion on some of the bolts and nuts. Those were good candidates for replacement. We used butyl tape for them too, wrapping it around the beveled head and the upper body of the bolt before inserting it into the toe rail. This took care of the port side leaks, but the starboard side still leaked. The boat is built by setting the deck on top of the hulls and bolting through the overlap. The overlap is similar to asphalt shingles on a house roof. A low-grade caulk is used for the hull-to-deck joint and sometimes this caulk doesn't extend as far as it should to prevent leaks. The toe rail seems to direct some water back into the overlap, where it eventually finds an opening. Most of the time the opening is back outside, but sometimes it is through a screw hole.

Carbon Screws Holding Hull-to-Deck Joint

The hull-to-deck joint looks like it is held in place by carbon steel screws, which have since rusted. We see them ever few inches along the joint. A better caulk (3M 5200 or 4200) with stainless screws would have been much better and not much more expensive. Of course, when the screws rust, they expand, adding a separating force to the joint.

How do we fix the problem with water ingress along the hull-to-deck joint? We're adverse to filling the space with regular epoxy due to its low elongation characteristics. However, we note that in several areas, such as the lockers and the V-berth, the seam has been covered with fiberglass. We decided to use the G/flex epoxy to fill some of the spaces and that seems to have solved the problem. In some cases, we had to drill out the toe rail bolt hole, fill it with G/flex epoxy, and re-drill in order to create a seal between the aluminum toe rail and the hull-to-deck joint. A better solution is to rebed the entire toe rail, but that's more work (i.e. time) than we could afford this summer.

Hatches and Line Deflectors

Of course, we also had leaks from two hatches, so they had to get rebedded. It seems that someone in the past decided to rebed the hatches with either 3M 4200 or 5200. After persistent attack with thin putty knives, we were able to break the bond between the hatch and the deck. It required going around the inside of the hatch from inside the cabin as well as working along the deck outside. We found that the bolts around the hinge area are particularly susceptible to leaking. We rebedded them with butyl caulk.

The rope guard around the starboard forward hatch looked like it had been rebedded, but with a very thin layer of poor quality caulk. It leaked a lot! We had only seen a little of the water, because it primarily dripped on the back of the headliner, which directed the flow to other areas as well as absorbing some of the water. Only when we removed the Ocean Air screen assembly did we see the full volume of water ingress. These were really easy to rebed with butyl caulk.

The final leaks that we found this summer were around the helm. There are two hand railings outboard of the helm seat and both needed to be rebedded. We could see staining on the head liner plywood and on the nut + washer. So they were unbolted, cleaned, and rebedded.

The result of our work was a drier boat. The bilges have remained dry through several hard rainstorms this winter.

  -Terry