September 2020 OES Beacon

COVID-19, Keel Bolts, Engine Mounts & 474 Blows

For all of us dealing with the effects of the corona virus, not all aspects of the social lock-downs are bad, as explained in Sandy’s article on keel bolt replacements on his sail boat. As the former VP for Technical Activities, Sandy provides quite an educational discussion for those society members who enjoy sailing their boats on the world’s oceans.

Albert J. Williams 3rd, IEEE Life Fellow
Woods Hole Oceanographic Institution
Woods Hole, MA 02543 USA
awilliams@whoi.edu

Abstract— Social Distancing and closure of Woods Hole Oceanographic Institution (WHOI) imposed restrictions on many of us professionally but I had a project that kept me fully engaged at home.  This is a tale of the keel bolt replacement on my sailboat, Shadowfax, in 2020.

I.     Introduction

When my wife, Izzie, and I returned from Borneo on March 21, a pre OCEANS Singapore trip cut short by Malaysian flight restrictions, we immediately went into self-quarantine at home in Woods Hole.  But I was engaged in preparing my 30’ sailboat for launch on April 30, the latest date permitted at our boatyard for a boat not kept there.  Being trapped at home was just what I needed.  This was the year for replacing my keel bolts; did them in 1980, 1997, and now it was a little overdue in 2020.

II.    Keel Bolts

A.    Old Bolts

Fig. 1. Fig. 1. Two new hot dipped galvanized 7/8″ keel bolts with old (23 year) keel bolts driven out, cut off when they reached the ground beneath the supported boat, and stacked next to the new bolts ready to replace the next two old bolts when they are driven out. The hex nut is on the top (left) of bolt #4 to illustrate its length compared to the 38” bolt similar to the one that has replaced it. Bolt #3 is 42” and was replaced by a similar one to the longer bolt ready to replace bolt #2. It was expected that the shorter one would replace bolt #1.

Two of the four old keel bolts were driven out.  I had made replacements and had them hot dip galvanized before leaving for Borneo.  Two bolts are 42” long and the other two are 38” long, 7/8” in diameter, and each went through wooden floors, about 12” of dead wood, and about 18” of cast iron ballast keel.  Fig. 1 shows two new ones and two old ones that did come out.  The lower end (right) has been upset by mushrooming them while red hot after excess was cut off and then the bolt was drawn up into a socket in the cast iron with a nut at the top.  The socket had been filled with epoxy before bottom paint had been applied.

  • Two new hot dipped galvanized 7/8″ keel bolts with old (23 year) keel bolts driven out, cut off when they reached the ground beneath the supported boat, and stacked next to the new bolts ready to replace the next two old bolts when they are driven out. The hex nut is on the top (left) of bolt #4 to illustrate its length compared to the 38” bolt similar to the one that has replaced it.  Bolt #3 is 42” and was replaced by a similar one to the longer bolt ready to replace bolt #2.  It was expected that the shorter one would replace bolt #1.

B.    Pile Driving

Fig. 2. Fig. 2. The driving rod extending up from the floor is being driven by the steel 40 lb. weight (right) dropped through the PVC pipe which is placed over the rod. The weight is picked up and dropped through the pipe. When the driving rod is about to disappear below the top of the floor, a longer driving rod is substituted until at the end, after the driven keel bolt has been cut off to permit continued driving, the last piece of bolt falls out and the new bolt, well-greased, is driven into the hole.

I knew that it took a heavy hammer to drive the old keel bolts out and in 1997 I set up a pile driver with 5” diameter PVC pipe, 5’ long, into which I initially dropped a 20 lb. steel weight. The eventual arrangement is shown in Fig. 2.  But bolt #1 and bolt #2 stopped moving after only about 1 3/8” when driven down by pile driving with a 20 lb. weight dropped 5’ repeatedly onto a driver resting on the bolt.  Thinking the lower end might be hitting the blocking under the keel, new blocking was acquired along with a 6 ton hydraulic jack and the boat was lifted and the original blocking replaced by new blocking farther aft shown in Fig. 3.  But even with the mushroomed heads exposed up in the socket, there was no movement.  Over the next six weeks various approaches were tried.  On the principal that if it doesn’t work, use a bigger hammer, I acquired a second 20 lb. weight and taped the two together to make a 40 lb. driving weight, the one shown in Fig. 2.  No improvement.  The characteristic was that the bounce of the weight off the rod was elastic until movement started and then became inelastic and the bounce stopped.  But it didn’t stop.  A long 1/2” bolt was tapped into the bottom of keel bolt #1 and with a jack and 4’ lever an attempt was made to pull it out. Oil was run into the bolt hole.  It was clear that something must be resting on top of the cast iron keel, possibly a splinter of iron from the keel bolt.  There was no progress.

  • The driving rod extending up from the floor is being driven by the steel 40 lb. weight (right) dropped through the PVC pipe which is placed over the rod. The weight is picked up and dropped through the pipe.  When the driving rod is about to disappear below the top of the floor, a longer driving rod is substituted until at the end, after the driven keel bolt has been cut off to permit continued driving, the last piece of bolt falls out and the new bolt, well-greased, is driven into the hole.

C.    A Window into the Deadwood

Finally a window was cut into the deadwood where the distortion of bolts #1 and #2 was revealed.  The sections of distorted bolt were removed by sawing as shown in Fig. 4 with the bright cut at the top and a cut at the bottom.

  • Because the blocking supporting the boat covered the sockets for keel bolts #1 and #2, it was necessary to move the supports farther aft. A six ton hydraulic jack lifted the boat so that a new stack of blocking could be installed and the original stack removed.  However, there had been no movement of the lower ends of bolts #1 and #2.
  • Deadwood was plunge-cut away to reveal the region where the keel bolt was distorted. Then the bolt was cut at the top and bottom of the window with the plunge cutter and removed.
Fig. 3. Fig. 3. Because the blocking supporting the boat covered the sockets for keel bolts #1 and #2, it was necessary to move the supports farther aft. A six ton hydraulic jack lifted the boat so that a new stack of blocking could be installed and the original stack removed. However, there had been no movement of the lower ends of bolts #1 and #2.
Fig. 4. Fig. 4. Deadwood was plunge-cut away to reveal the region where the keel bolt was distorted. Then the bolt was cut at the top and bottom of the window with the plunge cutter and removed.

D.   Extracting the Top Section

The top end of the bolt was tapped for a 1/2” bolt and the keel bolt was drawn up and out with the jack and lever as shown in Fig. 5.

E.    Driving out the Bottom Section

A high strength steel 3/4” rod was used to drive the lower piece out the bottom.

F.    Filling the Window in the Deadwood

Finally the window was filled with pieces of fir cut to fit and epoxied in around the replacement bolts as shown in fig. 6. .

  • A ½”.bolt was drilled and tapped into the top of the #2 keel bolt and drawn up and out with a 4’ steel channel and hydraulic jack.
  • Pieces of fir were cut and shaped to fill the window in the deadwood and after the new bolts were in place, these pieces were bedded in place with thickened epoxy.
Fig. 5.
Fig. 5. A ½”.bolt was drilled and tapped into the top of the #2 keel bolt and drawn up and out with a 4’ steel channel and hydraulic jack.
Fig. 6. Fig. 6. Pieces of fir were cut and shaped to fill the window in the deadwood and after the new bolts were in place, these pieces were bedded in place with thickened epoxy.

III.   Secondary Consequences of Keel Bolt Replacement

A.    Floor Damage

Driving the keel bolts with short rods, and then longer rods, using 20 lb. weight and then 40 lb. weight took a toll on the wooden floors supporting the keel bolts. During keel bolt replacement in 1997 two of the floors showed damage on their aft surfaces and 3”x3”x1/4” steel angles were hot dip galvanized and bolted to the floors to strengthen them and provide a surface for the nuts of the keel bolts to rest upon.  In this new replacement, the floor for keel bolt #2 became so damaged that it required replacement as well.  I had a large piece of American white oak remaining from a mast step replacement project five years ago and this provided material for a new floor as shown in Fig 7.

  • White oak replacement floor above and damaged floor below that it was copied from. Width exceeded available opening so the wings were drilled for carriage bolts and then cut for assembly after insertion in the bilge.

Access under the interior finished work was limited by fresh water tanks on either side so that the wings of the new floor had to be cut and assembled in parts under the finished work.  Carriage bolt holes were drilled before cutting so that the wings could be bolted on under edges of the sole.  This is shown dry assembled in Fig. 8.

Fig. 7. Fig. 7. White oak replacement floor above and damaged floor below that it was copied from. Width exceeded available opening so the wings were drilled for carriage bolts and then cut for assembly after insertion in the bilge.
Fig. 8. Fig. 8. New floor for keel bolt #2 before the old steel angle was attached. Diagonal cuts to the wings permitted insertion beneath the sides of the sole and under the coverings of the fresh water tanks. The floor was bedded in 3M5200 High Strength Adhesive Sealant and screwed from the outside with #12 silicon bronze wood screws. The external screw holes were plugged with ½” mahogany bungs.
  • New floor for keel bolt #2 before the old steel angle was attached. Diagonal cuts to the wings permitted insertion beneath the sides of the sole and under the coverings of the fresh water tanks.  The floor was bedded in 3M5200 High Strength Adhesive Sealant and screwed from the outside with #12 silicon bronze wood screws.  The external screw holes were plugged with ½” mahogany bungs.

B.    Floor for Keel Bolt #4

Keel bolt #4 passes through the floor that also supports the engine bearers and this floor also saw extensive damage from the keel bolt driving.  The extent of the damage was only realized as the nut was tightened after the lower end of the bolt was upset and drawn back into the socket in the ballast keel.  A remainder of the 3”x3”x1/4” steel angle provided a 12” piece that was machined for the keel bolt and placed under the nut.  This has allowed the nut to be torqued down to 300 ft. lb. as for the others without continuing to bury itself into the floor.  Replacing this floor will be a major job because it holds the engine bearers and the engine will need to be removed to replace it.  Perhaps it will be on the agenda for the next keel bolt replacement in 2040.

While the engine ran fine when the boat was launched and produced expected speed through the water and expected maximum engine rpms at full throttle, there was a certain irregularity to the sound.  But it was only on the third excursion that something serious happened.  After throttling back the engine to await the raising of the drawbridge for our access to Eel Pond, Woods Hole, where we keep the boat, there was a load clattering and inspection showed the starboard engine bearer moving.

C.    Propeller Shaft Coupling

With so much work on the keel bolt replacement and damage to floors, it is advised to check alignment of the propeller shaft coupling but this is a very hard place to see.  There is a hand hole in the inboard side of the aft starboard bunk and a place to get another hand in beneath the aft end of the inboard berth panel, but it is hard to see and even harder to work on the coupling.  However, inspection showed that one of the bolts holding the coupling together had sheared off and the remaining three bolts were only finger tight after removing the wire preventing them from unscrewing.  Separating the coupling revealed a misalignment of about 5° and a small vertical misalignment.  These should be adjusted with the four vibration isolating engine mounts.  My three cylinder diesel engine is expected to produce more vibration than the four cylinder gasoline engine it replaced so the vibration isolating mounts are required.

Fig. 9. Fig. 9. The mount that was hammered out on the 474th blow. The drift is in the 3/8” drilled hole and the top of the stud sawed off is beside the drift on top of the vibration damper.

D.   Engine Mounts

In order to remove the 5° misalignment the forward engine mounts had to be raised about 1” which required moving the aluminum angle brackets and re-drilling the bolt holes connecting them to the wooden engine bearers.  This was done and while at it, a 15” section of 3”x3”x1/4” bridged the space between the port and starboard engine bearers tied to the forward engine mount bolts.  This last removed the possible motion of the starboard wooden engine mount where apparently the bracket connecting it to the floor timber had failed.  But raising the forward end of the engine while making the two halves of the coupling parallel increased the vertical offset and the aft engine mounts needed adjustment.  The jam nuts on these could not be moved.

E.    Driving Out Frozen Engine Mount

Access to the aft end of the engine is obtained by removing the cockpit footwell liner, a moderately difficult task but it was done without damage.  This allowed removal of the starboard vibration isolating mount where at home the jam nut was loosened with a long wrench.  But the port mount could not be removed despite taking off the top jam nut.  And application of Kroil, guaranteed to free stuck metal parts, failed to improve the situation in 9 days of repeated application.  So the top part of the stud was sawed off with my plunge cutter (three days) and then a 3/8” hole was drilled into the 5/8” stud stuck in the unthreaded  clearance hole.  Finally a 3/8” drift was machined from one of the keel bolt driving rods.  Hammering on this drift was commenced with the target of 1000 blows to drive the stud out.  First day had 300 blows with a rest after each 100.  The second day picked up with the 400 series but on blow 474 there was a thunk.  This was the mount falling into the bilge.  Fig. 9 is the mount with the drift and the top of the stud that was sawed off.

  • The mount that was hammered out on the 474th blow. The drift is in the 3/8” drilled hole and the top of the stud sawed off is beside the drift on top of the vibration damper.

The replacement mount I had ordered had been delivered and it easily went in the hole and was adjusted to remove the vertical misalignment.  One more adjustment of all four mounts got the halves both parallel and vertically aligned and new coupling bolts were inserted, torqued down, and wired through their nuts to prevent unscrewing.  Fig. 10 shows the aft end of the engine with limited working space between the exhaust pipe, the heat exchanger, and the transmission.

  • The coupling is open and above the exhaust pipe. The propeller shaft enters the shaft log through a green shaft seal that is lubricated by grease in the copper tubing entering the seal from the left.  The shift cable bracket has been detached to allow more swinging room for the wrench on the jam nut of the starboard mount, which has the top nut removed exposing the 5/8” stud just beneath the heat exchanger at upper right.  Lower right is the starboard drain hose from the cockpit liner.  The diagonal wooden beam supports the cockpit liner and provides a seat for working on the mounts.  The black object at the bottom is the Hydro Hush muffler.

IV.   Conclusion

Shadowfax was launched April 30, all painted and ready to sail as in Fig. 11.  The post launch work on the coupling and the engine mounts has displaced most sailing and allowed growth to somewhat slow the boat.  Full throttle now gives 6.0 knots and maximum of 3300 rpm.  But most important is that leak rates are less than any time in the last 20 years.

  • Painted and ready for launch on April 30, 2020 as required.
Fig. 10. Fig. 10. The coupling is open and above the exhaust pipe. The propeller shaft enters the shaft log through a green shaft seal that is lubricated by grease in the copper tubing entering the seal from the left. The shift cable bracket has been detached to allow more swinging room for the wrench on the jam nut of the starboard mount, which has the top nut removed exposing the 5/8” stud just beneath the heat exchanger at upper right. Lower right is the starboard drain hose from the cockpit liner. The diagonal wooden beam supports the cockpit liner and provides a seat for working on the mounts. The black object at the bottom is the Hydro Hush muffler.
Fig. 11. Fig. 11. Painted and ready for launch on April 30, 2020 as required.