In constructing Justine’s molds, I followed established procedure and cut the bottom of each mold in the plane of the construction baseline shown on the plans. With the baseline positioned near the barn floor, this placed the molds at a convenient working height for construction. As a first step in setting up the molds I needed to lay out the construction baseline in our barn.
I decided to make the baseline using long lengths of 1 x 2″ lumber that could be screwed to the barn floor and shimmed into position using opposing pieces of cedar shingles. My laser level was the ideal tool for positioning the baseline. The two photos below show what I did. I used about 21′ of 1 x 2 for my construction baseline as it would be needed over Justine’s entire length.
I set up my laser on a camera tripod on the ground just outside the barn door and adjusted its height and position to serve as a temporary construction baseline. I knew that the laser would make a perfectly straight and level beam of red light. I then began to secure a 1 x 2″ pine board to the barn floor, with its upper right edge (as seen here) positioned very carefully along the laser beam.I used two cedar shingles with their wedges opposed underneath the 1 x 2 to adjust the height so that the laser beam just kissed the corner of the 1 x 2. Once in position, I used dry-wall screws to fasten through the 1 x 2 and shingles into the barn floor. The top of my laser level is visible just beyond the barn door.
Each of Justine’s molds needs to lie perpendicular to the construction baseline. I drew a very accurate perpendicular line about midway along the baseline, using the method I was taught in my high-school plane geometry class. Then I’d used a ruler and a compass. Now, I’d need a long compass and I made one up using a small lumber off-cut fitted with trammel points. One of the trammels has a pointed end, and the other can hold a pencil. So I had an adjustable compass about 5′ long with which to draw the arcs to erect a perpendicular to the baseline.
The beam and trammel points served as a large compass.Starting with a tick mark on the construction baseline, arcs are drawn on the baseline a few feet on each side, then longer arcs are drawn from centers where the first arcs intersect the baseline. These arcs intersect along the perpendicular to the baseline, drawn from the initial tick mark. I put a small pad approximately where the intersection would like, before drawing the intersecting arcs shown here.
With a straight and level baseline, and an accurate perpendicular reference line, I was ready to start fastening the molds in place. I began near the center of the boat with the mold at station 14. This mold was supported with diagonal bracing to keep it exactly vertical. All the other molds were secured in the vertical plane with battens that tied in to the mold at station 14.
Here molds 13-23 are set up in position on the barn floor. The diagonal bracing holds mold 14 in the proper vertical plane, and the other molds are secured to it with battens to achieve a uniform 9 3/4″ spacing between stations. Shingle shims under the butts of each mold bring the mold to the correct height and with its cross spall level .With the construction molds in position, you can really visualize the there-dimensional shape of the hull! The long battens hold each of the molds in the proper vertical plane. You can see that Justine did a good job of taking up much of the floor space in our barn…
The photo above was taken on September 15, 2008. I began construction of the molds in the fall of 2006. The project’s pace was slow. I was working full-time in Boston, and building Justine in Georgetown, ME, a three-hour drive away. So much of the construction work was done on long weekends, holidays, and vacations. I didn’t retire and move to Maine until 2013.
The lofting I’ve described so far produced 22 half-sections that represent the shape of the planked hull at stations 2-23. Two factors arise in determining the actual shape of the construction molds on which Justine would be built. I’ve mentioned the first in my previous post: The dimensions of the molds need to be reduced from the lofted sections to allow for the thickness of Justine’s planks (5/8″) and the thickness of her frames (7/8″). One could simply reduce the dimension of the half-sections by the sum, 1 1/2″, at every point. This would only work well amidships, at points where the plane of the construction mold is perpendicular to the hull’s surface. Because the hull shape narrows toward the bow and the stern, the hull lies at an acute angle to the plane of the molds. By taking angles off the lofting it’s possible to calculate a more accurate dimension by which to correct for the difference between the lofted dimension (outside surface of the hull) and the mold dimension. (I did this using vector algebra. Feel free to contact me if you’d like details.)
The illustration below shows an example of how I altered the shapes of the lofted sections to produce drawings of the construction molds near Justine’s bow. Because of the narrowing of the hull forward, the molds in this part of the hull need to be smaller than the sections by about 1.65″, rather than the 1.5″ combined thickness of the frames and planking.
Lofted sections at stations 2-5 (light blue lines). Red circles have diameter equal to calculated reduction for frame and planking thickness. Darker lines are drawn tangent to the corresponding circles to give the shape of the construction molds. The baseline corresponds to the common plane on which all 22 construction molds will rest.
The second factor concerns the hull’s curvature: the stations at which the sections were lofted are planes at precise distances along the hull, but the construction molds have a specific thickness (mine would be built from pine boards that were planed on one side and about 7/8″ in thickness). Because the hull is curved, and the frames necessarily have to lie fair against the hull’s planks, the construction molds have to be beveled along their edges. Furthermore, it will be a “winding bevel” whose angle changes along the molds’ edges. The construction molds generally have a wider side and a narrower side because of the bevels. The wider side is placed exactly on the plan’s stations (where the mold shape was drawn). Consequently, the molds are positioned so that the beveled edges face forward on the bow end of the boat, and face aft on the stern end. (This is shown clearly on the Flatfish plans.)
Once I had the mold shapes drawn on my computer, I printed them out full-size on an inexpensive HP ink-jet printer. I’d added registration marks on my computer drawing, and the my software was able to print the full-size drawing on separate 8 1/2 x 11″ sheets that I then taped together. I was very careful to position the sheets so that the registration marks were aligned and the grid on which I’d positioned the marks remained perpendicular.
I fastened some sheets of plywood to the barn floor on which to position and connect the numerous sheets on which I’d printed out my mold drawings. Here I’m in the process of fastening the 8 1/2 x 11″ sheets together. Each sheet was carefully positioned and tacked to the plywood before joining adjacent sheets with tape. I started with the smallest mold at station 2 and worked aft. Completed molds for stations 2-5 are visible in the background.To economize on printing and aligning 8 1/2 x 11″ sheets, I overlaid four mold drawings in a single computer file before printing them. This is the drawing for molds 17-20 after aligning the sheets and tacking them to the plywood building surface.
With the mold drawings assembled and in position on the plywood, it was time to cut into some lumber! I’d bought pine boards that were planed on one side and joined on one edge from a local supplier. There’s a clever technique for transferring the line defining the mold’s shape on the drawing to the wood, using some common nails. A nail is placed on the drawing, with its head exactly on the line and its shank perpendicular to the line. Nails are placed at suitable intervals, about 6″ in my case. A hammer is used to tap the edge of the head into the drawing and plywood, so that the entire nail shank contacts the drawing. Then a piece of mold stock is placed in position over the drawing and on top of the nails (planed side of the stock facing down). Next, you walk along the mold stock so that the exposed portions of the nail heads are pressed into the mold stock. When you pick up the board you have a series of impressions that reproduce a series of points on the line on the drawing. Use a batten to connect these impressions with a smooth curve, and you’re ready to make your cut!
The newly-cut curved piece is laid in place on the drawing and marked across the construction baseline and centerline, and those cuts are made. The pieces are fastened temporarily in place to the drawing and plywood, then joined together with a separate piece across the top and at the sheer line.
Molds 2-6 are made of a single thickness of pine board, and the remaining molds, which are larger, more highly curved, and support more loads, are made of a double thickness.
The construction mold at station 6 is being assembled directly on the full-size drawing. Because the port and starboard sides of the mold have the same shape, the boards for each side can be fastened together temporarily, marked, and sawn to shape simultaneously. The pieces are carefully aligned with the construction baseline at the bottom, and about the centerline, and temporarily fastened to the plywood surface I’d attached to the barn floor. A cleat fastens the two halves together, and the top edge of the cleat is positioned at the position of the sheerline for that station. A separate piece will be added at the top of the mold to secure the pieces there.Larger molds with more curvature are made with several pieces. Here the mold for station 19 being assembled. A second layer of planks will be added above the long cleat. Molds for stations 2-18 are already assembled and stacked nearby.Here is a larger mold close to completion. The upper part of the mold has two layers to add strength, and it’s made from 8 pieces of pine. The second layer of planks serves as cleats that back up joints in the first layer. Brackets made from pieces of angle iron fastened to the bottom of the mold will serve as attachments to the shop floor.
At this juncture the molds are sawn to shape but not beveled. Beveling is left until the molds are in place, attached to the shop floor. Also left for later is making some modifications to the molds amidships (stations 13-17) to accommodate the centerboard trunk. I’ll describe both procedures in future posts.
As I was gaining experience lofting Justine’s lines on the plywood panels attached to the barn floor, I began to realize that unevenness in the floor surface was making it difficult to judge the fairness of curves that I needed to draw. I had shimmed the plywood lofting surface as I was laying it, nevertheless irregularities remained. These were a result of the barn being on a deteriorating foundation (I began lofting in 2003 and didn’t rebuild the foundation until 2006), the advanced age of the barn (c. 1900 we think), and the fact that the floor was comprised of a mix of boards of varying thickness.
Might I be able to do the lofting on my computer? As part of my work life I had coauthored two textbooks and become very proficient doing technical illustrations using a software program called Canvas. The program is loaded with features for architectural drawings, among other things, so making very accurate scale drawings of large objects would be possible. There were tools for drawing and making careful adjustments to curves, so producing fair lines through a set of points seemed will within reach.
Could I produce the full-size drawings that I would need? I knew I wouldn’t likely need to print out the whole lofting. But I would need to print full-size half sections of the hull, and drawings to make templates for parts like the stem and transom. I would need to print things as large as 4′ x 6′. I was confident I would be able to find a way to get what I needed printed.
A full-size profile drawing of Justine’s profile would be about 21′ long. My laptop (on which I’d be doing the lofting) screen was about 13″ wide. The Canvas program allows for producing scale drawings over a huge range of scales very conveniently. You specify the scale you want. Let’s say you want a 12″ line on the screen to represent 24′ on the boat. Then you set the scale at 1:24. What you see on the screen are “rulers” at the edges of the viewing window marked to the true size on the boat. If the window was 8″ x 12″ on the screen, the rulers would be 16′ x 24′. Now let’s say you use your mouse or trackpad to draw a horizontal line on the screen by clicking and dragging. Small panes in Canvas’ menu bar will indicate the starting location, the ending location, and the length of the line in the full-size drawing. This is incredibly convenient. Furthermore, by “zooming” the screen view, you enlarge the view, but the original scale of the drawing does not change: zooming the view also “zooms” the rulers. When you need to, you can zoom the view to full-size, allowing very precise positioning of features in the drawing.
A variety of methods can be used to draw curves in Canvas. One drawing tool allows you to click on any number of successive points on the screen, and Canvas will fit a smooth curve that connects the points. This is a good way to begin to develop a fair curve. You plot the appropriate points from the table of offsets (there might be ten or more), then you enable the curve tool, and click near four of five of the points, including the first and last. After the last, Canvas will connect the points with a smooth (but rarely fair) curve. Then zooming to full size, you can carefully set the points on the curve to coincide with the corresponding points you plotted from the table of offsets. To further refine the curve, each of the points on the generated curve has a pair of “handles” that control the slope of the curve at that point. This gives you real control on the fairness of the curve. By iterating the locations of the points and the position of the handles, you can come quite close to matching the points from the table of offsets with a fair curve. (The table of offsets gives points to the nearest 1/8″, so it’s not necessary to have the curve exactly on each of those points.)
The Herreshoff Manufacturing Company built most of their boats upside down over a set of construction molds that determine the hull’s shape. To illustrate, here’s a picture of Justine’s construction molds in position on the barn floor.
Justine’s construction molds in place on the barn floor. The shape of each of the 22 molds is determined from measurements taken off the plan-view and profile-view loftings. The three battens you see help position the molds with respect to their neighbors. Justine’s stem is also placed in position in this photo.
Justine’s table of offsets specifies coordinates at 11 “stations” positioned 19 1/2″ inches apart along the hull’s length. These correspond to station numbers 2, 4, 6, …, 22. The lofting is done based on this set of points, and there is a construction mold at each station. Molds are also needed halfway between each of these stations, corresponding to station numbers 3, 5, 7, …, 23. (There is no need for a mold at station 1 because of the very small space at that position in the hull.) In all there are 22 molds, each spaced 9 3/4″ apart. Each station mold needs to be drawn full-size.
My computer lofting of Justine’s plan-view and profile-view lines is shown below. Using information provided with the plan set, I’ve added full-size lofting of the transom, as well as the profiles of the stem and keel plank, and ballast keel. “LWL” is the load waterline of the boat—this is the designed location of the waterline when the boat has all its equipment and crew aboard. The curved lines in the upper drawing represent the hull’s waterlines which include the LWL and lines in parallel planes at 6″ intervals above and below the LWL. The curved lines in the lower drawing are buttock lines that represent the hull’s contours in vertical planes along the hull’s mid plane and at 10″ intervals athwartships.
My computer lofting of Justine’s lines from points given in the table of offsets. You may be able to see light gray circles at every station. These are the points I plotted before adding the smooth curves connecting the points. Also shown is the “construction baseline” that serves as a key reference line for setting up the construction molds. I mounted a long 1 x 2 on the barn floor and one corner of it served as Justine’s construction baseline.
Note that in my lofting, I only show the even numbered stations, as these were the locations at which the points in the table of offsets were given. Before lofting Justine’s sections, I added lines representing the locations of all the odd-numbered stations as I would need to draw those sections in addition to the even-numbered ones.
My lofted sections of Justine’s hull are shown below.
Lofting of Justine’s sections. Because the hull is symmetric about the mid plane, only half-sections are shown. Half-sections on the left are for the forward stations 2-14 and on the right are for the after stations 15-24. Gray circles are the points I took off the plan-view and profile-view loftings, then I used Canvas to draw fair curves connecting the corresponding points. Also included are sections of Justine’s ballast keel and deadwood.
There’s an additional wrinkle in drawing the shapes of the construction molds from the lofting of the sections. The table of offsets actually gives points on the outside of the hull’s surface. Between the outside of the hull and the construction molds there are 5/8″ planks as well as 7/8″ frames (steam bent “ribs”). Furthermore, the edges of the construction molds must be beveled to match the curvature of the hull at each point. So I used a two-step process to make full-size drawings of the construction molds: First, I produced lofted sections representing the outside of the hull, then I made an allowance for the reduction in mold size to account for the thickness of the planking, the dimensions of the frames, and the bevel of the molds. I drew a fair curve through the corrected points and then I was nearly done with the lofting. I just needed a way to print out my results!
I first explored options to do large-format printing directly from my laptop. My brother-in-law Win Fowler had a sail loft with a computer-c0ntrolled fabric cutter that could be fitted with a pen and draw on paper or mylar. He made some test prints on mylar sheet for me, but I was not satisfied with the accuracy of the printout.
I knew from experience that it would be very expensive to have large-format prints made commercially, so I developed a “work-around”: add a set of registration marks on a grid to my computer-generated mold shapes, and simply have Canvas print my drawing on a number of overlapping 8 1/2 x 11 sheets of paper. I found that if I was very careful in positioning the registration marks as I assembled the printed sheets, I could generate suitably accurate full-size drawings comprised of 8 1/2 x 11 sheets attached to each other with two-sided tape.
Coming next: building Justine’s construction molds.
To build a boat to a specific design you need a set of plans. (I bought mine from the WoodenBoat Store.) Plans vary significantly in level of detail. The plan set for my Flatfish consists of six sheets, the largest of which is about 3×4′: a table of offsets; a lines plan, a construction plan, two sail plans, and a plan with drawings of the spars and important pieces of hardware.
The table of offsets quantifies the location of a collection of x,y,z coordinates for points on the surface of the hull using a standardized format. These points are used by the boatbuilder to generate the designed three-dimensional shape of the hull.
The lines plan is a scale drawing of the hull shape as projected when viewed from the top, side and front/rear. The lines plan is produced by plotting points from the table of offsets and connecting them with fair (smooth, “eye-sweet”) curves.
The construction plan gives details about the various parts that make up the boat, such as shape, thickness, suggested materials (there are six different wood species in Justine), and some of the fasteners that are needed. The construction plan contains top, front/rear, and side views, with detailed information about how the various parts fit together. To me, this is the most interesting of the plans set because it gives some sense of complexity of the building project and it takes a lot of study to figure out some of the finer details.
The sail plan shows a side-view drawing of the boat with its sails raised and gives sail dimensions, area, center of effort, as well as details about the running and standing rigging. The Flatfish plan set includes two sail plans, one for the traditional gaff rig (quadrilateral mainsail), and one for the more modern Marconi rig (triangular mainsail).
The spars and hardware plan provides dimensioned drawings of the wooden mast and boom(s), locations of various hardware items on the spars, and three-view drawings of many of the bronze hardware items that were standard on the boats produced at the Herreshoff Manufacturing Company. For the Flatfish, the spars and hardware plan also provides dimensioned drawings and a perspective illustration for constructing a mold for the boat’s 1200+ lb. ballast keel.
It would be impossible to a good job building a boat of Justine’s size and complexity without first producing full-size drawings of the boat in top, side, and front/rear projections. The process of taking the data from the table of offsets and rendering full-size drawings is called lofting or laying down the boat’s lines. Careful lofting is of critical importance: mistakes in lofting, if not caught, result in a hull that deviates from the designer’s intended shape. Once the boat’s lines are lofted, numerous important parts of the boat can be drawn on the lofting, in full size. This generally includes the stem, transom, and keel plank/timber which comprise the “backbone” of most wooden boats.
When I built my 14’5″ Biscayne Bay Sailing Skiff, I prepared a surface for lofting by attaching two 4×8′ plywood sheets together end-to-end to form a 4×16′ panel. I painted it with a coat of flat white paint, and did the lofting on the floor of a study in our house. Justine required a larger surface for the lofting and she was to be built in our barn in Georgetown, Maine. So I began by fastening painted plywood sheets to the upstairs floor of our barn, forming a surface 6×20′ on which to begin the lofting. I started this lofting in 2004.
Here’s the surface on which I began the lofting. The photo doesn’t reveal many of the lines on the plywood, but the two pieces of unpainted plywood you see are templates for the stem and the transom knee that I made from the lofting.
So now that you know something about what a lofting is, how do you actually make one? I began by doing a lot of reading. Most boatbuilding books include a chapter or two on the topic, and I read several before I began lofting my Biscayne Bay Sailing Skiff. [A favorite book of mine that has a good treatment of lofting is Bud Macintosh’s How to Build a Wooden Boat.]Then I dove right in and learned by doing.
Lofting begins by drawing a grid of perpendicular lines spaced at convenient intervals consistent with the locations of points in the table of offsets. I started drawing the profile of the hull, as viewed from the side. By taking appropriate values from the table of offsets, I plotted a series of points that mark the full-size profile on the drawing. In lofting the boat’s profile, you plot out the front of the stem, the bottom of the keel plank, the transom, and the top of the uppermost plank that forms the side of the hull (known as the boat’s sheer line).
Once you have a set of points that represent a smoothly curved part of the hull, you need to connect the points to form a very fair curve. This is accomplished with the aid of a long relatively thin piece of wood, called a batten, fixed to the lofting at several places along the curve so as to intersect the plotted points. It requires some fussing with the locations at which you fix the battens to the lofting, so that the batten takes a very fair shape while hitting (or nearly so) all the plotted points. It helps to have someone experienced (I had Scot) to give you a hand when you are starting out using battens. You need a collection of battens with different cross sections so you can draw lines of varying curvatures. The battens need to have an even grain, with no knots, so they naturally form smooth curves when bent.
Laying out a batten to draw one of Justine’s sections. The batten is relatively thin to accommodate the abrupt curve at the “turn of the bilge.” You can see a few of the nails that hold the batten in place. By using a relatively small number of nails placed strategically, the resulting curve will be very fair. Once you’ve achieved that, the line is drawn on the lofting using the batten as a guide.
Once I’d drawn the hull’s profile, I began to plot more points from the table of offsets to create an additional set of lines that represent sections of the hull at 10″ intervals out from the hull’s mid plane (fore and aft + vertical), again as viewed from the side. This set of lines (called buttock lines) produces something akin to a topographical map representing the hull’s shape as viewed from the side.
Next, the “plan view” of the hull shape, that is, the shape looking down from above, is plotted out. Because the hull is symmetric (the port side is a mirror image of the starboard side), it’s only necessary to loft one half of the hull shape. The profile lofting and the plan view lofting are drawn with a common axis, specified by the table of offsets, and both are superposed on the lofting. The only difference between the “lines plan” provided in the set of plans and the curves described so far in the lofting is that that the lofting is full-size. In theory, if you had an enlarger that could simply scale up and reproduce the lines plan, you would not have to loft those. But I am sure that with current technology you’d get a much more accurate result by doing the lofting yourself. The table of offsets for the Flatfish includes plan-view coordinates at the boat’s waterline, and additional sets of coordinates at 6″ intervals both below and above the waterline. These lines, called water lines, are also drawn on the lofting and the set of waterlines can be visualized as a topographic map depicting the hull shape.
Finally, a third set of lines is drawn on the lofting giving a set of “sections” that represent the cross section of at fixed intervals (in the case of the Flatfish, 19 1/2″) from stem to stern. It’s as if the hull is sliced like a loaf of bread, and the shape of each slice is drawn. If you ever studied mechanical drawing, you know that the shapes of these sections can be derived from measurements that you make on the profile- and plan-view drawings. Lofting books explain how to do this, and by doing it this way the section lines will be consistent with the faired lines in the profile and plan-view lofting. This set of curves is that comprise the sections is critically important for constructing the Flatfish, because these curves determine the shape of forms onto which the boat’s frames (“ribs”) are steam-bent and held in shape while the boat is constructed.
For mathematically-knowledgable readers I can make the relative orientations of the profile-, plan-view, and sections a little more explicit. Imagine an x axis running from the boat’s mid plane out to the port (left) side, the y axis as pointing upward, and the z axis running forward toward the boat’s stem. Then the profile curves are in the y-z plane, the plan-view curves are in the x-z plane, and the section curves are in the x-y plane.
So a complete lofting gives three sets of curves related to side, top, and fore/aft views of the hull shape. Each set can be thought of as contour-lines on a map, and with some practice your mind can visualize the shape of the hull directly from the lofting.
The flatfish lines plan looks like this:
Flatfish lines drawing showing plan view with “waterlines” (top); section views (middle, note forward sections on left and aft sections on right); and profile view with “buttock lines” (bottom). A drawing of the transom is shown on the profile view at the right (stern). [From “Forty Wooden Boats,” Woodenboat Publications, Brooklyn, Maine, p. 45 (1995).]
Our barn had suffered some neglect over the years before we bought it. The foundation needed major work. The floors were not level and in places they were quite squishy underfoot. One corner of the barn’s sill lay directly on ledge, but the other supports were tenuous. At one time, one corner of the barn had been attached to a tree by a steel cable. I think that was done to keep the barn from toppling off its supports. More recently three posts resting on stones had been added along the long downhill side, but the stones were lying on dirt. Two of the posts showed signs of rot.
The barn on its original foundation in 2005.
Ideally, a boat building floor should be level. But most importantly it should be stable over time. I anticipated that my boat would take several years to build, and if the building settled during construction, the shape of the boat would be changing as it’s being built. So before doing any building the barn’s foundation needed some serious work.
My neighbor Scot Smith gave me lots of advice about how to proceed, helped me at crucial times, and loaned me some tools I needed to do the work. I spent most of my August vacation in 2006 renewing the foundation.
The barn needed to be raised by about 5″ at one corner and lesser amounts at two others, and I needed to take the weight off the old foundation so old parts could be removed and replaced as needed. Scot provided me with a hydraulic jack and a supply of cribbing to raise and support the barn. Cribbing consists of stout pieces of lumber about 24″ long that can be stacked Lincoln-log style to provide stable temporary support while the new foundation pieces are set in place.
I have found a laser level to be an indispensable tool in boatbuilding. I used a laser level inside the barn to gauge how much I’d need to raise the corners of the barn. Not surprisingly, the highest corner was the one supported directly by ledge. The lowest corner was the opposite one. The long side farthest from the ledge was the place to start lifting the barn.
Because the jack I used has a relatively short throw, we built a crib below the foundation on which to place the jack, then started lifting. Another crib was built nearby to take the load and support the building after the jack was lowered from the first crib. Cribbing and jacking points were shifted as needed.
Beginning to raise the barn. Here I’m lifting the barn off one of the original posts (just beyond the first crib) so I can remove it.
Scot advised me to add concrete supports under the new posts that I would install. We assumed that I could dig down until I hit ledge, then pour concrete supports that would be anchored to the ledge. I’d make the concrete forms using 10″ diameter cardboard “Quik-tube,” and before pouring in the cement I’d drill holes in the ledge and put in short pieces of rebar to help lock the concrete to the ledge.
So I dug down about 18″ for my first hole and indeed hit ledge. The ledge was uneven, so I had to scribe the Quik-tube and cut the bottom edge so it closely fit against the ledge to prevent the concrete from oozing out.
Ready to pour the concrete. The rotting post that will be replaced is visible at the bottom left.
Before pouring cement, I made sure the top of the Quik-tube was level. Because I planned to secure the post to the support with steel brackets made for that purpose, I embedded an appropriate size stud in the wet concrete.
After adding concrete to the form, a stud is embedded to secure a steel bracket to which the new wood post will attach. I suspended an old spade-drill bit to act as a plumb bob to ensure that the bracket is located properly in relation to the barn’s sill. After the concrete sets, the cardboard Quik-tube is peeled off.
I made posts of the correct length from pressure-treated 6×6 lumber. The sills were 4×6’s, so I cut notches at the top of the posts so I could secure the overlapped parts with strong fasteners.
A new foundation post in place!
This process was repeated five times, for a total of six posts: three along the front of the barn and three along the back. Working at the back was more awkward because the back is close to a ledge and the sill is close to ground level. Sometimes I had to work under the barn for better access.
The rear of the barn sits on temporary stone supports. All three concrete post supports are in place. The rear posts will be much shorter than those in front.
The end of the barn under the door has a stone foundation with no mortar. Although it was in pretty good shape, it needed to be higher to match the new level of the foundation. I decided to rebuild much of it and in the process get it in contact with ledge along most of its length. I have limited experience laying stones. And some of this work needed to be done from under the barn. Nevertheless, it came out looking and functioning very well.
Rebuilding the stone part of the foundation to match the new height of the barn.Dry wall foundation below the barn door after rebuilding.
The barn looked great on its new foundation, but for further stability the posts I’d installed needed diagonal bracing. (Scot’s chief concern was the ability of the barn to withstand strong wind gusts.) So I cut suitable pieces of 4×4 pressure-treated lumber to make braces for the six new posts.
One of the diagonal bracing pieces positioned and ready to be permanently fastened with lag screws (in place but not yet driven).The rear posts and bracing in place on the new concrete footings.
The barn’s floor joists run between the long sides of the barn so they are about 11′ 6″ long and had no support except at their ends. Scot recommended that I add a central beam that would run parallel to the long sides of the barn and support the joists at their midpoints. We made up two 12′ sections for the beam by fastening 2x8s face to face to make two 12′ 4×8 beams that we’d support with a center post. One end of the beam would rest on the stone wall that supports the door end of the barn, and the opposite end would be supported by a short 6×6 post that rests on ledge.
A 4×8 support beam was added under the center of the barn’s floor to stiffen the floor and remove some sagging that had developed over the years. A 6×6 center support for the beam rests directly on ledge. Shims were added between the beam and the joists as needed to make the floor firm and level.
The barn is very solid on its new foundation. The floor feels very stiff underfoot. I took careful measurements inside the barn and determined that the floor level varied by no more than 5/8″ around its perimeter—a big improvement over the 5″ variation before my work commenced!
By the end of my 2006 vacation, all that remained to be done was to install lattice work over the opening under the barn. That was accomplished in 2007. In 2008, I had the barn doors replaced and that end of the barn re-sided. In the meantime, construction of Justine was underway. More about that in future posts!
Minerva’s first few seasons in the water were spent in Muscongus Bay, in Maine’s midcoast. We were in the habit of renting a cottage in Waldoboro or Friendship each August, so vacations began with launching and ended with haul out onto a trailer. Minerva spent her off-season months under a cover at my mother- and father-in-law’s yard in Cumberland Foreside. Muscongus Bay is somewhat protected and offers many possible routes around islands, through channels, in addition to some open water. It’s a great place for a small daysailer!
In 2002 we purchased a second home in Georgetown, Maine near the village of Five Islands and with a limited view of Harmon’s Harbor. Through the generosity of two very good friends, we’ve been able to use their dock for water access to the harbor. We put in a mooring that we can see from our front porch, and Minerva had a new home port and a season in the water that typically went from July through August.
Harmon’s Harbor is fairly narrow and about 1 mile long, with a narrow entrance protected by ledge. The harbor exit takes you right out into the mouth of the Sheepscot River, which is virtually open ocean with depths approaching 200′ in places. It’s not unusual to encounter big swells outside of the harbor. At 14+ feet, Minerva is a small boat for significant swells. She also has no cockpit seats, so you are generally sitting on the rail if there is a good breeze. It can be a rough, wet ride! So once we had a taste of these more demanding conditions, I realized that having a larger boat would be prudent, allowing us to sail more safely and comfortably.
In my reading about boats and boatbuilding, I began to learn about America’s celebrated yacht designer/builder, Nathaniel G. Herreshoff. His career spanned more nearly 60 years, beginning in 1878 up to his death in 1938, and included design of five successful defenders of the America’s Cup between 1893 and 1920. The Herreshoff Manufacturing Company produced upwards of 1000 boats during Herreshoff’s lifetime. One particular model, designed in 1914 and known as the “12 1/2 footer” (that was its length on the waterline), became extremely popular and over 400 were manufactured. Naval architect Joel White, designer of the Nutshell pram and many other boats, drew up plans for a centerboard version of the 12 1/2 footer, known as the Haven 12 1/2, and this became a very popular boat for home builders.
Maynard Bray authored a book on building the Haven, which serves as a very detailed construction manual. The book includes scores of photographs showing details of a Haven being built by professional boatbuilder Eric Dow. I purchased a copy of this book and spent endless hours over several years studying it, wondering if I possessed the skills necessary to build such a complex and potentially beautiful traditional boat.
Like many boat designers in his day, Herreshoff carved wooden “half-models” to refine the shape of his hull designs. By making careful measurements of the half-model, the design could be scaled up to the size of the actual boat. Herreshoff likely used the same half-model to design four different-size boats, including the 12 1/2 footer (designed in 1914) and a larger daysailer, called the Fish (designed in 1916), with a waterline length of 16′.
Because the Haven 12 1/2 was so well received by builders, Joel White decided to draw plans for a centerboard version of Herreshoff’s Fish class, a design he named the Flatfish. Flatfish plans were offered for sale by the WoodenBoat Store. Could I possibly build a Flatfish? I began to think I just might be up to it.
Specifications for the Flatfish and illustrations of the two options for the rig. I chose the gaff. [from: Forty Wooden Boats, WoodenBoat Publications, Brooklyn, Maine (1995), p. 44].Our home in Georgetown has a two-story barn that is 12′ x 30′, with a door on the narrow end. Likely it had formerly been used for boat building. Our house dates to 1852, but I believe the barn was built somewhat later. The stairs to the second level are at the rear, reducing the usable length of the barn by about 3′. I’d need space for a workbench, a band saw, and a planer/joiner. It would be tight, but I concluded that I did have enough space to construct a Flatfish in our barn.
Did I possess the skills required to build this boat? My two prior projects had given me some preparation and I had most of the tools I’d need. But the Flatfish was built “plank-on-frame” with “carvel planking.” The hull would be formed of cedar planks that had to fit closely together, and would be rounded as necessary inside and out to form a continuous smooth hull. More techniques to master! Carvel planking is considered the most advanced of traditional boat construction methods, and I decided that I wanted to challenge myself and build at least one carvel planked boat.
My “secret weapon” was that my neighbor and very good friend in Georgetown, Scot Smith, is a professional cabinetmaker who also had worked building traditional and modern boats. Scot had been very generous sharing his woodworking knowledge and I knew I could count on him for advice and occasional assistance as the project progressed.
I was set to proceed and in the fall of 2003, I ordered a set of Flatfish plans from the Woodenboat Store.
Once I’d built my Nutshell pram Pearl and confirmed that I loved to sail my own boat, I began to think about building a larger craft that would comfortable seat more people. But how to go about selecting a design?
I began to investigate possible designs by getting a copy of The WoodenBoat Store’s Fifty Wooden Boats: A catalog of building plans, which set out specifications for each of the watercraft for which they sold sets of building plans. I later got their follow-on Forty Wooden Boats. These books describe plans for six categories of craft: Tenders and Prams; Sailing Dinghies and Pulling Boats; Performance Rowing Craft; Power Cruisers; Daysailers; and Cruisers. How to limit the choices?
Well, I knew I wanted a boat I could sail, and I’d decided I wanted to build a wooden boat. I also know that I’m especially fond of traditional designs and building techniques. Then, the more practical issues: What skills would be required to build a particular boat? How much time would it take to build? Where would I build it, and how large a space would I have?
I had successfully built Pearl from a kit. The hull was constructed entirely of plywood and the pieces were pre-cut. The hull has a small plywood bow transom in front, a larger plywood transom aft, and seven plywood panels form the remainder of the hull: a bottom panel and three planks to starboard and three to port. The planks overlap each other and required beveling with a hand plane before being clamped to a building jig and gluing the overlapping joints with epoxy. This method is called “glued lapstrake construction” and it is relatively straightforward and results in a very strong, leak-free hull that is easy to maintain (scraping, sanding, and repainting).
The Nutshell pram’s plywood planks overlap and are glued with epoxy.
Pearl was built in the basement of our home in Jamaica Plain, Massachusetts. She is 7′ 7″ long, about 40″ in width, and about 24″ high. The basement had a bulkhead and the size of the opening and positioning of the stairs allowed me to carry the finished boat out through the bulkhead, with little room to spare. The basement itself could accommodate construction of a boat about 20′ in length. As I looked anew at prospect of building a larger boat in the basement, I noticed that there were adjacent basement windows with wooden frames that fit in an opening in the stone foundation that was 60″ wide and 24″ tall, provided that the windows and framing were completely removed. So I was looking for a design that when completed, would be able to fit out this opening.
My criteria were relatively simple: A sailboat design that would take a few passengers, have a traditional look, and be able to be built, and removed, from our basement. From my study of available plans it was apparent that I should focus on the Daysailers, and these ranged from a 14′ Catamaran, Pixie, to a 28′ Camden-class sloop. Some designs were plywood, and some of traditional “plank-on-frame” construction that yielded a continuous, smooth hull.
The design I chose is the Biscayne Bay Sailing Skiff, a 14′ 5″ sloop that could be built with a plywood hull, using a method known as “batten-seam” construction. In this method, the hull planks butt against each other, rather than overlapping, and the joints are backed up with a solid wood “batten” that serves to both strengthen and provide sealing for the joints. And, this boat would fit out the 2′ x 5′ opening in our home’s foundation, though again with little room to spare.
Description of the Biscayne Bay Sailing Skiff [Thirty Wooden Boats, WoodenBoat Publications, Brooklyn, ME, p. 24 (1988)].Inside Minerva‘s cockpit, the longitudinal “battens” that back up the plywood plank seams are visible, as are the steam-bent frames (ribs) that help give the hull its designed shape.
This project required learning some new skills which I’ll simply mention here and explain in more detail when we get to building my larger and more complex Flatfish. I needed to learn how to “loft” a boat design: making full-size drawings of the boat from which I could take measurements to make parts that would accurately execute the boat design. The new boat would require steam bending some “frames” (ribs) that would fit inside the plywood planks and provide support to maintain the hull’s shape. And not least of all, I needed to learn where to acquire the wood, fasteners, and hardware that it would take to produce a finished craft that was ready to launch and sail.
I enjoy making things with my hands. I’ve been an amateur woodworker for most of my adulthood. My first major project was to build a traditional woodworking workbench with a front vise, tail vise, wooden bench dogs, and a tool tray at the rear. I built it about 35 years ago, entirely by hand, in my basement, with only a crude 2′ x 3′ work table. In the process, I purchased some good-quality hand tools: saws, planes, clamps, etc.
My wife’s parents, Sandy and Sally Fowler, were life-long sailors with extensive sailboat experience “crusing” the Atlantic Coast, especially in Maine. Sandy also built some small boats. I learned a tremendous amount about sailing and navigation from being on the water with Sandy and Sal, and gradually got to the point of realizing that I could apply my woodworking skills to building my own boat.
So where does one start in boatbuilding? Most books about boatbuilding address this question at some length. I’ll just tell you how I went about it.
I knew I wanted to build a small sailboat so I could learn how to handle a boat on my own. Even at the outset I imagined that I would progress to building a larger boat, so I decided to build a suitable “tender” for a larger boat: a small boat that could be stored at a dock and used to access a larger boat on a mooring. So I sought a design for a boat that could be both rowed and sailed.
I was also a subscriber to WoodenBoat magazine, and I knew that the WoodenBoat Store sold plans for an assortment of wooden boats ranging from about 8′ to 40′ and more, including both sail and powerboats. They even offered a complete kit for a small craft called a Nutshell pram which is a rowboat designed by Joel White that has an optional lug sail rig. Plus, you could buy an inexpensive book with step-by-step instructions and photographs that you could follow to do the actual building. So I bought a kit and did the building in our basement over the winter of 1996.
Pearl has been in the water every season since. She’s constructed of marine plywood, is 7’7″ in length, and is capable of carrying 3 people in relatively calm water if you are rowing. Under sail, she’s pretty cramped with more than one person aboard. I learned to sail in Pearl, and supplemented my hands-on experience with lots of reading. In addition I began to learn about maintaining a wooden boat—specifically annual painting and varnishing.
Pearl has been everything I hoped for: a vessel for learning how to sail, a beautiful design, and a capable tender for a larger boat.
Pearl, a 7’7″ Nutshell Pram on the dock in Harmon’s Harbor, Georgetown, Maine.
I’ve long had the dream of building a traditional, carvel planked (round-bottom) wooden sailboat. I selected a specific design and ordered plans for it in 2004. Since then I have almost completed my project—launch is planned for Memorial Day weekend 2017.
Many friends have encouraged me to document my work, and I’ve responded by sending very occasional “updates” describing small parts of the project. Now I’m going to make a deep dive into documenting most of the important parts of my project.
I’m aiming to update the site with new posts each week. Stay tuned!
I built Minerva, my Biscayne Bay Sailing Skiff, in 1999. She’s seen here under sail in Harmon’s Harbor, Georgetown, Maine.
![Flatfish lines drawing showing plan view with "waterlines" (top); section views (middle, note forward sections on left and aft sections on right); and profile view with "buttock lines" (bottom). A drawing of the transom is shown on the profile view at the right (stern). [From "Forty Wooden Boats," Woodenboat Publications, Brooklyn, Maine, p. 45 (1995).]](https://i0.wp.com/samallensboat.com/wp-content/uploads/2016/12/Flatfish_lines.png?resize=840%2C451)
![Specifications for the Flatfish and illustrations of the two options for the rig. I chose the gaff. [from: Forty Wooden Boats, WoodenBoat Publications, Brooklyn, Maine (1995), p. 44].](https://i2.wp.com/samallensboat.com/wp-content/uploads/2016/11/Flatfish_blurb.jpg?resize=729%2C1024)
![Description of the Biscayne Bay Sailing Skiff [Thirty Wooden Boats, WoodenBoat Publications, Brooklyn, ME, p. 24 (1988)].](https://i1.wp.com/samallensboat.com/wp-content/uploads/2016/11/SCAN0055-1.jpg?resize=717%2C1024)
