Preamble
I had heard of my vintage of E38's (1986) having poor keel-stub adhesion due to the factory not removing the mold release wax before fitting the keel. For this reason, and also having experienced crevice corrosion in my previous boat, I decided I would drop the keel and rebed it properly. In retrospect, that was NOT necessary in my case, and I'd advise anyone thinking about doing the same to really think hard about the cost/benefit. If you do it, a good plan will come in handy too.
I found that my boat had a smile developing at the front edge of the keel-hull joint, as well as some delamination of a fiberglass strip covering the joint along the aft edge of the boat. Using an angle grinder and a flathead screwdriver, I ground exposed the keel-hull joint in the problem areas. I could see gelcoat on the bottom of the keel stub, and figured this was an indication that I had one of those boats that never got the wax taken off out of the mold (I was wrong).
This would have been an excellent time to dig into the exposed sealant with an angle grinder, if I had taken it back ½" I would have found excellent adhesion back there, and I could have filled the gap with fresh sealant, retorqued the keel bolts, and been on my way. Note that there are no rust stains that would indicate a problem with the bolts. If you find yourself in this position, here is what I would advise:
Keel Bolt Access
E38-200s have 13 keel bolts. At least mine does. The forward-most one is easy to access. The four behind it are difficult to get at without removing the mast plate, which of course necessitates removing the mast. The next four are easy to access under the aft dinette settee. The three behind that live under the engine, assuming you have a under-sink engine as I do. They are not easy to access, I did not need to remove the engine, but if you ran into complications, I could see that becoming necessary. The 13th, and aft-most keel-bolt is well hidden just ahead of the transmission. Very hard to access.
Before getting started with this project, I wanted to make sure I could put adequate torque on each bolt. A shopping trip to Harbour Freight procured the following necessary tools:
Next, I removed and re-installed each nut one-by-one. I wanted to make sure that I wouldn't be battling a stuck nut while sitting in (and paying for) a lift. I'm glad I did, several nuts wouldn't budge, and required an air impact wrench that a boat-neighbour luckily had on hand. Some of my nuts were in pretty ugly shape, and the washers looked even worse. Some were quite badly corroded, which I hoped wouldn't be a sign of more rust to come...
Preparing for Keel Removal
I ground back the paint, fiberglass, and filler that was covering my keel-hull joint. The adhesion between these layers was excellent along most of the seam. Only the front and rear areas had delaminated and were easy to remove. This would have been another great sign that perhaps the keel should not come off. I ignored this, and happily exposed the entire seam. If you find yourself doing this job, I would consider at this stage to use a thin disc on an angle grinder, and cut into the seam about ¼" in. Just enough to get the point of a chisel in. More on that later.
I had to plan for where the keel would drop, and not cause mayhem. The marina I'm at (Marina Seca, San Carlos, Mexico) had built a custom keel-dolly for one of their customers the year prior. The dimensions were a bit small for my keel, but would be acceptable provided I took care to strap the keel down to the dolly so it couldn't tip forwards. Your yard may have something similar, or you may need to build a keel catcher. Make it strong, the E38 keel weighs over 3 tons, that's nearly 1000 lbs heavier than a Cadillac Escalade. Plan ahead.
Drop It
This marina doesn't have any travel-lifts or slings, all boat movement is done with what they call a 'low boy', effectively, a low-slung hydraulic trailer. Luckily my keel cart had already been built with the geometry of the low-boy in mind. If your marina uses slings, the geometry of your keel catcher should not be much of an issue.
The plan for me was something like this:
The low-boy is a bit more complicated to get out of than slings. The hydraulic legs come down off the hull, and then the whole trailer is pulled back. As the trailer backs off, jackstands can be placed below the boat. Critically - you can only get jackstands in behind the trailer, or on the centerline, before the trailer is removed. That is fine when the boat is resting on its keel, as that takes all the weight and the jackstands are only there for lateral stability. When the keel is hanging, all that weight goes through the hull and into the jackstands. Two stands on the stern and one at the bow is not even close to sufficient. Luckily, I caught the error before the operators tried to back the trailer out. We slid a 4-ton car jack under the keel, jacked it up to take the weight of the keel/boat, and then were able to safely remove the trailer.
Once the trailer was out, I inserted about a dozen jackstands, so that the hull could be supported like the roof of the acropolis. The next day, I let the boat hang in the stands by lowering the jack, and managed to get the keel unglued from the hull. The key for me was several sharp chisels, and a handful of hardwood wedges. Pound the chisel in first to make some space, then insert hardwood wedges on either side, pound those in, remove the chisel, move it over, and repeat. Once you have wedges all along the seam, you can go about in a circle gently tapping them in further and further until the keel releases. Digging into the seam with an angle grinder and oscillating multi tool was useless, skip that step and go straight to the wedges. If that doesn't work, an idea that I considered would be to buy an extra-long hole saw that just fits over the keel bolt, and drill down into each of the keel bolt holes where the sealant is gooped in thick. Luckily I did not need to do this.
This was also the moment where I discovered the aft-most keel bolt. Up until now I thought there were only 12. Using a combination of the car jack under the keel, and torquing up the front three bolts, I was able to get the keel back up into position, remove lucky bolt #13, and then drop the keel into the cart. The nuts remained threaded onto the tops of the studs. That allowed me to bring the keel back up into position using the jack + torque on the nuts. Completely separating the hull and keel would have made the prep work much easier, but because my yard has no travel-lift, I needed the keel in place to take the boat's weight in order to get the lift in under the hull (the jackstands come out as the trailer is slid under the boat).
Mating Surface
In my misguided attempts at removing the keel using a reciprocating tool, I had unintentionally gouged a deep scar into the laminate just above the keel stub. Instead of attempting a backfill into the narrow gap, I instead chose to make the laminates work for me, and got a chisel into the gap. It was pretty easy work to get the laminate to peel off in several large chunks. That saved me the effort of trying to get old 5200 off the keel stub, and also meant I wouldn't need to try and backfill the gouge I'd created.
The goal was to create a near-perfect mating surface between the keel and stub. The procedure I followed in preparation went something like this:
At this point I felt I was ready to lay down some thickened epoxy and create the mating surface. I'll detail the procedure, then I'll tell you about all the mistakes I made along the way. The process itself is pretty simple:
Insufficient/wrong materials: My first attempt, I underestimated how much gap filling would be required. I ran out of thickener, and to make matters worse, the West 206 (slow) hardener was curing much faster than anticipated in the Mexican summer heat. This half-assed job was a complete waste of time and resources, and I ended up having to completely grind off what few places did adhere to the top surface, and do it again. For attempt #2 I procured some West System 209 hardener (extra extra slow) and a lot of the 403 Microfiber filler. Whatever amount you think you need - double it.
The gap between the keel and stub must only get smaller: During my second attempt, as in the first, I used the jack underneath the keel to bring it up to the hull. Once I had the keel in position, I tightened up the keel bolts, since I knew this jack would slowly lose pressure over time. What I did not anticipate, is that the jack preferentially raised the aft end of the keel, and after tightening the keel bolts the aft end would settle a bit. This led to air pockets in the aft third of the thickened epoxy layer. Luckily, the front ⅔ displayed excellent adhesion. I vowed to remember this lesson when it came time to apply sealant, the keel cannot be allowed to settle!
Did not level the hull: This is probably the most important lesson, it was the hardest-earned. My hull was not 100% level in the stands, and I took 0 measures to ensure that the mating surface would be level to the hull. What I should have done is this:
Seal the Deal
The yard would not allow me to sit in the hydraulic trailer off the keel for the night, which meant that all my final preparations for applying sealant would need to occur while paying for lift time ($100 - $130 / hr). On the big day I had my tools ready and a plan of attack:
I chose to use Sika 291 for the sealing job, which was a mistake. It is an excellent product, and I would use it again, but nobody else in Mexico uses the stuff. So when I found myself two tubes short of finishing the job, I was in a big panic. All the shops around here only carry 3M products. I managed to find an extra tube of 291 LOT kicking around in my process materials box, and was just able to squeak out the job. But it was certainly a panic getting to the finish line.
For future reference: I used 2 tubes coating the keel bolts, and 3 tubes on the keel/stub mating surface. I would have liked to have 4 tubes available for the keel/stub surface.
When I pulled the keel and hull back together, I did not torque the bolts very hard. I left a 2 mm gap (approximately) between the keel and stub. My reasoning is that I did not want to create a glue-starved joint between the keel and stub. I wanted the sika to cure, and then torque the bolts down a week later. This way the entire joint will always be in compression, and significant movement between the keel and stub will be necessary to get the sika to release. I expect that if you torque down the bolts with wet sika below, you will create a glue-starved joint that is neither in tension nor compression once it dries. Then you go sailing, or torque up the rigging, and parts of that joint end up in tension. If you have a ½ mm layer move 1 mm, that represents 200% elongation. For a 2 mm layer that would have only been 50%, and likely less if you had it in compression after it cured. My opinion on this matter is not authoritative, you'll have to decide for yourself what is right.
It was only after the sikaflex fully cured that I realized that I had taken no steps to ensure that the keel was sitting vertical to the hull. I spent 2 days in a slump, ruminating over making such a silly mistake, despite taking so many precautions to do the job right. In the end, I've accepted that my keel will be a bit off vertical. I estimate that if my 6300 lb keel is ½ in off-center, that's equivalent to a smaller crew member (130 lbs) sitting 2 feet off of centerline. Could have been worse.
Backing Plates and Cost Breakdown
Between bouts of sanding, epoxying, and cursing - I drew up some CAD files for keel bolt backing plates. The fender washers that came out of my bilge were all bent into various cone shapes, which seems like a clear indication that they were not up to the task of maintaining the required keel bolt tension. The process was surprisingly simple:
Cost breakdown for the job was (approximately):
I had heard of my vintage of E38's (1986) having poor keel-stub adhesion due to the factory not removing the mold release wax before fitting the keel. For this reason, and also having experienced crevice corrosion in my previous boat, I decided I would drop the keel and rebed it properly. In retrospect, that was NOT necessary in my case, and I'd advise anyone thinking about doing the same to really think hard about the cost/benefit. If you do it, a good plan will come in handy too.
I found that my boat had a smile developing at the front edge of the keel-hull joint, as well as some delamination of a fiberglass strip covering the joint along the aft edge of the boat. Using an angle grinder and a flathead screwdriver, I ground exposed the keel-hull joint in the problem areas. I could see gelcoat on the bottom of the keel stub, and figured this was an indication that I had one of those boats that never got the wax taken off out of the mold (I was wrong).
This would have been an excellent time to dig into the exposed sealant with an angle grinder, if I had taken it back ½" I would have found excellent adhesion back there, and I could have filled the gap with fresh sealant, retorqued the keel bolts, and been on my way. Note that there are no rust stains that would indicate a problem with the bolts. If you find yourself in this position, here is what I would advise:
- Clean up any separating or unadhered sealant.
- Remove and replace your nuts, lock rings, and washers.
- Clean up the exposed ends of the stud, re-grease, and retorque.
- Consider fabricating backing plates, or using thick, oversized washers (more on that later).
Keel Bolt Access
E38-200s have 13 keel bolts. At least mine does. The forward-most one is easy to access. The four behind it are difficult to get at without removing the mast plate, which of course necessitates removing the mast. The next four are easy to access under the aft dinette settee. The three behind that live under the engine, assuming you have a under-sink engine as I do. They are not easy to access, I did not need to remove the engine, but if you ran into complications, I could see that becoming necessary. The 13th, and aft-most keel-bolt is well hidden just ahead of the transmission. Very hard to access.
Before getting started with this project, I wanted to make sure I could put adequate torque on each bolt. A shopping trip to Harbour Freight procured the following necessary tools:
- 1-½ (1.5) inch deep-drive and regular socket
- 1-⅛ (1.125) inch deep-drive and regular socket
- Universal-joint adaptors
- Drive extensions totalling 2' length total
- 250 ft-lb torque wrench with 15º flexible head
Next, I removed and re-installed each nut one-by-one. I wanted to make sure that I wouldn't be battling a stuck nut while sitting in (and paying for) a lift. I'm glad I did, several nuts wouldn't budge, and required an air impact wrench that a boat-neighbour luckily had on hand. Some of my nuts were in pretty ugly shape, and the washers looked even worse. Some were quite badly corroded, which I hoped wouldn't be a sign of more rust to come...
Preparing for Keel Removal
I ground back the paint, fiberglass, and filler that was covering my keel-hull joint. The adhesion between these layers was excellent along most of the seam. Only the front and rear areas had delaminated and were easy to remove. This would have been another great sign that perhaps the keel should not come off. I ignored this, and happily exposed the entire seam. If you find yourself doing this job, I would consider at this stage to use a thin disc on an angle grinder, and cut into the seam about ¼" in. Just enough to get the point of a chisel in. More on that later.
I had to plan for where the keel would drop, and not cause mayhem. The marina I'm at (Marina Seca, San Carlos, Mexico) had built a custom keel-dolly for one of their customers the year prior. The dimensions were a bit small for my keel, but would be acceptable provided I took care to strap the keel down to the dolly so it couldn't tip forwards. Your yard may have something similar, or you may need to build a keel catcher. Make it strong, the E38 keel weighs over 3 tons, that's nearly 1000 lbs heavier than a Cadillac Escalade. Plan ahead.
Drop It
This marina doesn't have any travel-lifts or slings, all boat movement is done with what they call a 'low boy', effectively, a low-slung hydraulic trailer. Luckily my keel cart had already been built with the geometry of the low-boy in mind. If your marina uses slings, the geometry of your keel catcher should not be much of an issue.
The plan for me was something like this:
- Lift boat off stands
- Place keel cart below
- Lower boat down gently so keel is resting on cart
- Loosen all nuts so they sit flush with the tops of the studs
- Hoist boat up 1.5 inches, keel falls off, lock it into keel cart
- Remove nuts completely, and lift boat up so studs are completely clear of hull
- Sit hull down on stands
The low-boy is a bit more complicated to get out of than slings. The hydraulic legs come down off the hull, and then the whole trailer is pulled back. As the trailer backs off, jackstands can be placed below the boat. Critically - you can only get jackstands in behind the trailer, or on the centerline, before the trailer is removed. That is fine when the boat is resting on its keel, as that takes all the weight and the jackstands are only there for lateral stability. When the keel is hanging, all that weight goes through the hull and into the jackstands. Two stands on the stern and one at the bow is not even close to sufficient. Luckily, I caught the error before the operators tried to back the trailer out. We slid a 4-ton car jack under the keel, jacked it up to take the weight of the keel/boat, and then were able to safely remove the trailer.
Once the trailer was out, I inserted about a dozen jackstands, so that the hull could be supported like the roof of the acropolis. The next day, I let the boat hang in the stands by lowering the jack, and managed to get the keel unglued from the hull. The key for me was several sharp chisels, and a handful of hardwood wedges. Pound the chisel in first to make some space, then insert hardwood wedges on either side, pound those in, remove the chisel, move it over, and repeat. Once you have wedges all along the seam, you can go about in a circle gently tapping them in further and further until the keel releases. Digging into the seam with an angle grinder and oscillating multi tool was useless, skip that step and go straight to the wedges. If that doesn't work, an idea that I considered would be to buy an extra-long hole saw that just fits over the keel bolt, and drill down into each of the keel bolt holes where the sealant is gooped in thick. Luckily I did not need to do this.
This was also the moment where I discovered the aft-most keel bolt. Up until now I thought there were only 12. Using a combination of the car jack under the keel, and torquing up the front three bolts, I was able to get the keel back up into position, remove lucky bolt #13, and then drop the keel into the cart. The nuts remained threaded onto the tops of the studs. That allowed me to bring the keel back up into position using the jack + torque on the nuts. Completely separating the hull and keel would have made the prep work much easier, but because my yard has no travel-lift, I needed the keel in place to take the boat's weight in order to get the lift in under the hull (the jackstands come out as the trailer is slid under the boat).
Mating Surface
In my misguided attempts at removing the keel using a reciprocating tool, I had unintentionally gouged a deep scar into the laminate just above the keel stub. Instead of attempting a backfill into the narrow gap, I instead chose to make the laminates work for me, and got a chisel into the gap. It was pretty easy work to get the laminate to peel off in several large chunks. That saved me the effort of trying to get old 5200 off the keel stub, and also meant I wouldn't need to try and backfill the gouge I'd created.
The goal was to create a near-perfect mating surface between the keel and stub. The procedure I followed in preparation went something like this:
- Sand top surface so it's ready to bond to epoxy. Take care of any oil that might be seeping around the bilge, it will definitely ruin your day if it finds its way down a keel bolt. Another cruiser in the yard loaned me a very useful tool for this job, it's a tiny belt sander that fits into spaces just under an inch wide. I would have suffered a lot without it.
- Clean up the bolt threads, I found tapping a screwdriver into the sealant worked well to get most of it off, followed by solvents + wire brush.
- Sand bottom surface to bare metal, try not to breathe in lead dust. Clean the surface with compressed air, wipe down with acetone, then a thin layer of epoxy onto the bare lead. Obviously this needs time to cure before the next steps.
At this point I felt I was ready to lay down some thickened epoxy and create the mating surface. I'll detail the procedure, then I'll tell you about all the mistakes I made along the way. The process itself is pretty simple:
- Coat the base of the keel and the keel bolts with mold release wax.
- Mix up a big batch up thickened epoxy, and spread it between the keel and hull.
- Bring the keel up to the hull, or drop the hull down into the epoxy so its squeezing out everywhere.
- Clean up the mess, let it cure, then separate.
- Knock the epoxy columns off the bolts and you now have a near perfect mating surface. Don't forget to clean off that mold release wax
Insufficient/wrong materials: My first attempt, I underestimated how much gap filling would be required. I ran out of thickener, and to make matters worse, the West 206 (slow) hardener was curing much faster than anticipated in the Mexican summer heat. This half-assed job was a complete waste of time and resources, and I ended up having to completely grind off what few places did adhere to the top surface, and do it again. For attempt #2 I procured some West System 209 hardener (extra extra slow) and a lot of the 403 Microfiber filler. Whatever amount you think you need - double it.
The gap between the keel and stub must only get smaller: During my second attempt, as in the first, I used the jack underneath the keel to bring it up to the hull. Once I had the keel in position, I tightened up the keel bolts, since I knew this jack would slowly lose pressure over time. What I did not anticipate, is that the jack preferentially raised the aft end of the keel, and after tightening the keel bolts the aft end would settle a bit. This led to air pockets in the aft third of the thickened epoxy layer. Luckily, the front ⅔ displayed excellent adhesion. I vowed to remember this lesson when it came time to apply sealant, the keel cannot be allowed to settle!
Did not level the hull: This is probably the most important lesson, it was the hardest-earned. My hull was not 100% level in the stands, and I took 0 measures to ensure that the mating surface would be level to the hull. What I should have done is this:
- Zip tie a clear tube from stanchion to stanchion across the widest part of the boat. The tube should run under the hull and the ends left open. Then fill the tube with water until the water is at deck level. Check both sides, they should match. If not, level your boat until both sides are right at deck level. You could repeat this process forward and aft if you're worried about hull 'twist', but I wouldn't be. Just getting the boat level is a good start.
- Hang a plumb bob off the leading and trailing edge of your keel. After you spread the thickened epoxy, you can confirm that the keel is hanging vertical as you bring it up to the hull. Small adjustments using the keel bolts would likely be sufficient to correct any small deviations.
Seal the Deal
The yard would not allow me to sit in the hydraulic trailer off the keel for the night, which meant that all my final preparations for applying sealant would need to occur while paying for lift time ($100 - $130 / hr). On the big day I had my tools ready and a plan of attack:
- Separate the keel and hull completely.
- Run a hole saw through the keel bolt holes to tear out old sealant and prep the surface for fresh sealant.
- Use a wire wheel brush and solvents to clean up the keel bolts.
- Use an orbital sander to scuff up the top and bottom mating surfaces. I used a lot of paint thinner on the bottom to remove the mold release wax before sanding.
I chose to use Sika 291 for the sealing job, which was a mistake. It is an excellent product, and I would use it again, but nobody else in Mexico uses the stuff. So when I found myself two tubes short of finishing the job, I was in a big panic. All the shops around here only carry 3M products. I managed to find an extra tube of 291 LOT kicking around in my process materials box, and was just able to squeak out the job. But it was certainly a panic getting to the finish line.
For future reference: I used 2 tubes coating the keel bolts, and 3 tubes on the keel/stub mating surface. I would have liked to have 4 tubes available for the keel/stub surface.
When I pulled the keel and hull back together, I did not torque the bolts very hard. I left a 2 mm gap (approximately) between the keel and stub. My reasoning is that I did not want to create a glue-starved joint between the keel and stub. I wanted the sika to cure, and then torque the bolts down a week later. This way the entire joint will always be in compression, and significant movement between the keel and stub will be necessary to get the sika to release. I expect that if you torque down the bolts with wet sika below, you will create a glue-starved joint that is neither in tension nor compression once it dries. Then you go sailing, or torque up the rigging, and parts of that joint end up in tension. If you have a ½ mm layer move 1 mm, that represents 200% elongation. For a 2 mm layer that would have only been 50%, and likely less if you had it in compression after it cured. My opinion on this matter is not authoritative, you'll have to decide for yourself what is right.
It was only after the sikaflex fully cured that I realized that I had taken no steps to ensure that the keel was sitting vertical to the hull. I spent 2 days in a slump, ruminating over making such a silly mistake, despite taking so many precautions to do the job right. In the end, I've accepted that my keel will be a bit off vertical. I estimate that if my 6300 lb keel is ½ in off-center, that's equivalent to a smaller crew member (130 lbs) sitting 2 feet off of centerline. Could have been worse.
Backing Plates and Cost Breakdown
Between bouts of sanding, epoxying, and cursing - I drew up some CAD files for keel bolt backing plates. The fender washers that came out of my bilge were all bent into various cone shapes, which seems like a clear indication that they were not up to the task of maintaining the required keel bolt tension. The process was surprisingly simple:
- Prototype the backing plates using graphing paper. This makes it easy to get exact measurements, and keep things square.
- Download Autodesk Fusion 360 - free for personal use. I went through several 'Free CAD' programs which either didn't work, or weren't free. The Autodesk product wins here.
- If you haven't done CAD before, go watch a couple youtube videos. The main thing to remember is that a good CAD drawing depends on constraints. If one line is supposed to be parallel to another, define it as such. You should be able to drag your drawing around the screen and not have it fall to pieces.
- For fabrication, I typed 'waterjet cutting' into Google Maps and found several contenders. Machine shops seem to either be slammed, or looking for work. Call around, because I found the busy ones don't really want to deal with such a small job. The shop I ended up going with quoted me $95 to cut 12 backing plates out of ⅜ in 304 stainless. The other quote I got was ~$400 for ½ in 316 stainless. Given that the bolts are made of 304, I did not feel compelled to splurge on the 316 stainless.
- Waterjet cutters will often leave rust streaks behind on the material being cut. I sanded my pieces down with 400/600/1200 grit sandpaper to get a nice even shine.
Cost breakdown for the job was (approximately):
- $700 yard/lift costs (life is cheaper in San Carlos, Mexico)
- $300 for raw materials (epoxy, thickener, and sealant)
- $120 new nuts and washers
- $95 backing plates
- $300 tools (torque wrench, XL socket set, drive extensions)
