Taking Al Breed's Shell Carving Class

Al Breed showing how to handle the grain when carving the various angled lobes of the shell. Some cuts you have to push the gouge, some you have to pull it. He said 99% of carving is following the grain and knowing how the tools respond to it.

This weekend, Al Breed offered a shell carving class for the Guild of New Hampshire Woodworkers members participating in the Townsend document chest project. He had a full house of 12 students. Lunches were provided, excellent wraps and cookies from the Black Bean Cafe.

The chest has three shell carvings, two convex and one concave. The convex shells are carved out of blanks, then glued to the chest doors. That gives you a chance to try again if you make a mistake. The concave shell is carved in the solid out of a door panel, so a mistake can be more costly.

The pattern Al used for the class is a classic Newport shell. The shells on the chest are an earlier, smaller version. Al gave everyone a mahogany blank glued to a backer, a copy of the pattern, a sheet of carbon paper, and a reprint of his Master Class article from Fine Woodworking. He also set out castings of the pattern to follow (he sells the castings on his website; in addition to being a 3-D model, they show the step-by-step sequence of cuts, so they're like a self-contained class). The carbon paper is used to transfer the pattern to the blank.


Al demonstrates the initial rounding of the shell's top edge in a mahogany blank. It's glued to a backing board with hide glue and brown paper. The backing board provides a solid support that's secured to the bench. The paper makes it easy to release the carving once it's finished.


My blank after rounding, setting out of the lobes with a V-tool, and initial shaping of some of the concave lobes. The lobes are alternately convex and concave.


The class hard at work.


John Short working on his portable bench raiser. Oh yeah, I built a bench-on-bench for this kind of work...and didn't bring it!


Continued progress shaping the concave lobes to establish the fillet angles on the their sides. From here I scooped the lobes out.

By the end of the day, we all had creaking backs from concentrated stooping over the work. For the second day, I brought my bench-on-bench so I could work at a more comfortable height.

I also brought my Queen Anne foot stool to ask Al about some fine shaping details (getting the most out of my time with a museum-quality expert). The transitions from the curved tops of the knees to the straight sides of the post blocks weren't as fine as I'd like. He recommended using the square edges of metal-working files to get a crisp inside corner and smooth surface, as well as using them on the straight grain in the rabbet at the top of the post blocks. He also said I should bevel back the undersides of the knee blocks. That maximizes the visibility of their curves.


My workpiece secured to the bench-on-bench. I've completed rounding over the convex lobes.


Al demonstrates a pull stroke with a broad gouge to excavate the ellipse at the center of the shell.


Outlining the petals with a V-tool.


Scooping out the petals.


Emulating Al's technique on my ellipse.


My final piece. The lighting and stray guide lines make it a little hard to see. I did have to "erase" the ellipse and petals with a broad gouge and re-do them. Al said you can use custom scrapers and square-edge and round files to smooth everything down once the piece has been removed from the backer.

Al ended the class with a demonstration of carving a concave shell. The ellipse and petals are the same, but all the lobes are reversed, and the rounded field is sunken in.


Al roughly excavating the field.


Carving out the reversed lobes.


The other unique detail on this shell is a fine scrolling outline that follows the lobe ends. Controlling this requires a specific technique: steering with the handle, the thumb is rolled against the small back-bent gouge to advance it through the cut. This allows tracing out any curve in a continuous line.

I'll practice this a few times before it's time to make the shells for the chest. In some ways it's surprisingly easy, though it does take a lot of patience and concentration. It's also good to be bold. That gives the work a dynamic flow. Timid work looks flat. With over 30 years of practice, Al was able to do in minutes what took the rest of us all morning.

SAPFM Demo at Whitehorne House, Newport, RI

The e-mail flyer sent out by the Newport Restoration Foundation.

This past Sunday I had the opportunity to participate in a truly unique event.

The Whitehorne House, in Newport, RI, was celebrating its 200th anniversary with an open house. It's one of several properties owned by the Newport Restoration Foundation, founded in 1968 by Doris Duke. Built in 1811, it now serves as a museum of fine Newport and Rhode Island furniture from the late 18th century, featuring Townsend and Goddard pieces, among others.

As part of the celebration, the Foundation had contacted the Society of American Period Furniture Makers about doing a demonstration of period furniture-making techniques. Founded in 1999, SAPFM is a national organization, though its membership is primarily in the eastern half of the country. At the SAPFM New England chapter May meeting, Freddy Roman had asked me if I would be interested in participating. Of course! Once again, I feel blessed to live in an area so rich in history, with so many exceptional woodworkers willing to share their passion for the craft.

There were four of us in attendance: Mickey Callahan, SAPFM co-founder; Freddy Roman, SAPFM NE chapter coordinator; Paul Williard, professor of chemistry at Brown University; and myself. Together, we provided a broad range of experience, from the 25-year professional down to the weekend hobbyist.

You can see in the flyer above they mention "expert craftsmen"! Well, three experts and an apprentice, me! My skills are 10 to 20 years away from the museum-quality work of people like Al Breed and Phil Lowe. I mention them because there was a tea table much like the one I saw at Al's in the Whitehorne's drawing room, and two chairs much like the one I saw at Phil's in the dining room.

Mickey Callahan working on a ball-and-claw carving in mahogany.

Freddy Roman veneering a serpentine card table with banding.

Paul Williard carving a fan in mahogany. Note his use of a bench-on-bench to elevate the work.

I'm chopping a mortise in poplar, banging away noisily with mallet and pig-sticker chisel.

The original plan was to have us working outside in the beautiful urban garden, but the weather was uncooperative. It rained the day and night before, and in the morning. So as a backup plan, we laid a dropcloth on the floor and set up in the house's kitchen. It was a bit tight, but Mickey, Freddy, and Paul shared Mickey's bench, and I worked on my portable Underhill bench.

Mickey had brought a beautiful, nearly-completed mahogany corner chair with ball-and-claw feet, and worked on the carving for a matching chair. Freddy worked on a serpentine card table similar to the ones in the house's dining room. Paul had a mahogany fan carving and a tall clock case.

I chose to demonstrate basic skills. I made a mortise-and-tenon joint and a dovetail joint as people watched. Since this was intended to be interactive and hands-on, I also brought some pieces of walnut in raw and dressed condition to let people try out different bench and molding planes. There's nothing that brings a smile to someone's face like making shavings!

I ended up making two mortises, demonstrating both a failed joint and a successful one (unintended, since I neglected to leave a horn on the first one and it blew out the end-grain). The dovetail was truly ugly, but nevertheless demonstrated the fine mechanical properties of the joint. They're all teaching moments!

None of us completed much actual woodworking, because we had many questions from numerous interested visitors, but of course engagement was the real goal for the day, not projects. In fact, after closing the doors, Lisa Dady, the NRF Director of Education and Public Programs, said it was a record day, with 175 visitors. For a small specialty museum, that was good to hear.

Mickey gave several impromptu lessons in blade sharpening, and I even got as far as showing one group the differences between rip and crosscut saws and how they are sharpened. People were fascinated by the results that could be achieved with hand tools, and the variety and specialization of tools. To most people, a saw is just a saw, and about all they expect to do with it is cut off the end of a 2x4.

Several people asked if I had run my dressed walnut board through a planer or sanded it, and I explained that it was all hand-planed, and I wouldn't want to ruin that satiny surface with sandpaper. With a little oil finish, it would flash as beautifully as the mahogany antiques behind the ropes in the other rooms.

I had people scrubbing the rough board with a #5, edge jointing the dressed board with a #6, taking gossamer face curls with a #4, and rabbeting and chamfering edges with a wooden skew rabbet molding plane. Some quick lessons in tool handling and body mechanics produced satisfying results.

These two pieces of the same walnut board show rough and partially dressed condition in back, and fully planed to thickness and smoothness in front. It's always hard to capture grain, but you can just make out some flash in the figure at the right end of the smooth board. This wood had been sitting in a basement in Durham, NH for nearly 30 years.

A little about the Whitehorne House and its contents: it was built as a symbol of prosperity by Samuel Whitehorne, Jr. in 1811. He had accumulated his wealth through shipping and other enterprises following the American Revolution. But as readers of Patrick O'Brian's sea stories know, fortunes can turn in an instant on the blue water. Whitehorne was bankrupted by the loss of two ships at sea. The house was sold at auction in 1843, and over the next 125 years, through use as shops and apartments, fell into disrepair.

In 1969, Doris Duke, heiress to tobacco and energy tycoon James Buchanan Duke, bought the house and had it restored. It was in shockingly hideous condition inside and out. You can see before and after photos of the exterior here, and another pre-restoration photo here. The fact that she could discern anything of value in it is amazing. But much like a rusted hulk of a tool a woodworker finds in the back in the dark in some old barn, she had a vision of what it could be.

Once it was restored, she set about acquiring authentic period pieces to furnish it. The house now contains a wonderful collection. As one of the guides pointed out the secret drawers in one of the desks, I was again reminded of a scene from O'Brian, where a British warship is docked for repairs after being mauled in battle by a French frigate. The shipwrights and carpenters having completed their tasks, the "confidential joiners" move in to build secret compartments into the captain's cabin for hiding gold and admiralty documents.

This was my first demonstration, and it was instructive. I brought far too much stuff. I was prepared to do every possible operation (9 years as a Scoutmaster does that to you), but in the end a couple of interesting joints and a variety of planing operations suffice for a general demo. Maybe a little sawing, since the humble saw is the least intimidating to most people.

Many of the same questions were repeated as one group of visitors left and another came in. We all offered to let people try their hand at things, and the responses were either that they were afraid to damage our work, or they were eager to jump in and try it.

My gift of appreciation for Freddy: a Fray and Pigg Spofford brace that I found in an antique store in Maine, made sometime between 1859 and 1866. A little rusty, I tested it and found it to be dead-on in alignment. A few drops of oil in the handle restored it to perfect operation. It should serve him and his descendents for another 150 years.

This morning my toolbox went straight from Newport to my daughter's school, where for her end-of-year session, her group is building the 12' dory Daisy from the 1995 Wooden Boat issues 126 and 127. They only have a week, so they won't complete it, but they hope to put her under sail next year. I'll be taking two days vacation to help out, so hopefully that will be my next post. They got the first station mold up today.

Building An Al Breed Carving Vise and Bow Lathe

My version of Al Breed's carving vise.

This is something I've been itching to make for months, ever since I took a ball-and-claw carving class from Al Breed. Each student's workbench included one of these carving vises, which he had built based on an iron vise someone had given him. He very kindly gave me permission to do a blog post on them, so I took plenty of pictures. I included my practice carving as a reference piece in the photos, so I could scale general dimensions from it.

This vise allows you to hold a long spindle for carving between centers and manually rotate it as you work. Al just screwed his down to his benchtops as needed, but it can be easily held down with clamps or holdfasts.

My version is almost identical to his. The only real difference is that I made the base a little wider. I used 1/2" Baltic birch plywood for the base and triangles, and 3/4" for the uprights and crosspieces.

The overall dimensions I used are 3" x 5 1/4" x 36" for the base, and 5" x 10" for the upright sides, with crosspieces and support triangles fitted to size. I actually think I should have made it at least 6" longer (ok, yeah, I should have taken some measurements!). If I ever need to, I can make an extension for the base and attach it with a couple of pieces inside acting as gusset plates.

As you can see, the upright with the veneer press screw (the headstock) is fixed to the end. The one with the thumbscrew (the tailstock) slides along the base to be held in fixed positions spaced at 2" intervals via a rod through the base. One consequence of building it a little wider was that I couldn't find a long-enough 3/8" bolt at the hardware store to go all the way through, so I used a length of all-thread with a cap-nut.

First sub-assembly was the base.

Cross-section diagram of the end of the base, 3" tall by 5 1/4" wide, showing how the sides are rabbeted for the top and bottom.

Cutting the base plywood to length from some narrow panels I had.

Marking out the width of a base piece.

Ripping it down. This stance is a little easier on the back than leaning all the way over with knee down.

Jointing the top and bottom base pieces, ganged together. They need to match precisely, since the base is a long rectangular box.

Rabbeting one edge of a base side.

All glued up. I should have used cauls to get more even pressure.

The hand tools all worked fine with the plywood. The only thing that gave me any trouble was rabbeting one of the side pieces. One inner ply had a knot and some punky grain that caught and tore out. However, since I was going deeper than that ply, it didn't cause any permanent problems.

Next sub-assembly was the uprights. I cut them out as above, then ganged them all up to notch them for the crosspieces.

Gang-cutting the base of the notch. You can see how the end of the base came out as well.

Down the length of the notch.

Cutting one pair of support triangles out of a rectangle. I scored down the line with the chisel to get the cut started.

All the parts of the uprights ready for assembly. Those are 12" and 24" folding rules for scale.

I drilled the holes in the crosspieces for the veneer press screw nut and a tee nut for the thumbscrew. However, for more precise alignment of the holes, it would have been better to wait until everything was assembled before drilling the tee nut hole. Push the tailstock up against the headstock, and crank the press screw up to it to mark the hole position, then drill it. If necessary, alignment can be fine-tuned by fidgeting with the press screw nut in its hole.

Then I clamped things up dry to the base for a test assembly. The first thing I noticed was that I couldn't run the press screw in all the way, because the handle bumped the headstock sides. Aha, that's why Al had sawn curves into the headstock!

Cutting a curved profile into the ganged headstock sides to provide clearance for the handle of the veneer press screw. This is the bow saw I built with the Gramercy Tools hardware kit.

Sharpening the thumbscrew to a point. It serves as the center in the tailstock.

After assembling the uprights with glue and screws, I noticed a couple of things that needed tuning. First, the width of the base was not quite uniform down its length. There were a couple spots where the tailstock would bind and not slide. Second, when the uprights were in place, they actually extended past the bottom of the base by a hair. This meant they would either hold the base up off the benchtop, or the support triangles would not rest flush on the base.

With the base on its side, smoothing the high spots with a #4. The result was smooth travel of the tailstock along the length of the base.

Shaving the bottom of an upright to provide some clearance. From each end of the piece working toward the center, I chamfered the sides, then did the center. This avoided tearout of the plies.

Once everything was tuned up, I completed the assembly, securing the headstock and drilling placement holes for the tailstock.

Before assembly, I had ganged the two sides of the tailstock and drilled the hole for the rod. Here I'm setting the tailstock at each position and using those holes as the guides to drill the sides of the base.

This detail shows the clearances I want. The bottom edge of the triangle rests flat on the base to provide support, and the bottom edge of the tailstock is clear of the bench surface, so that the base rests flat.

Detail of the face of the tailstock before lining it with adhesive-backed sandpaper. What are those mysterious holes to the left of the thumbscrew?

Detail of the headstock. I also ground the tip of the press screw to a sharper point. Again, those mysterious holes.

Here you can see the fit of the headstock to the base.

The finished clamp secured to the bench with holdfasts, holding one of my ball-and-claw practice carvings. But see the paragraph and photo below, the left end of the carving needs to be butted up tight against sandpaper lining the face of the tailstock.

One detail I forgot to show above, Al lines the face of the tailstock with a half-circle of 60-grit adhesive-backed sanding disc, and backs off the thumbscrew to just barely poke through to the carving. When you tighten the press screw, the sandpaper grips the end of the carving to keep it from rotating while you work on it. When you loosen the press screw so you can turn the carving, the tip of the thumbscrew keeps it from dropping down. Note that this also means the tee nut for the thumbscrew needs to be flush with the face under the sandpaper, so you need to carve out a small recess for it with a gouge and epoxy it in place.

Sandpaper lining the tailstock face to grip the end of the workpiece and keep it from rotating while carving. I didn't have a 60-grit disc, so this is a rectangle of 80-grit adhesive-backed paper.

So what's this business about a bow lathe? Someone had posted a link on Sawmill Creek Neanderthal Haven forum to this awesome video of a street vendor in Marrakech. He uses a bow lathe to make chess sets, guiding his chisel with his foot. The general consensus was that he embodied the true meaning of craftmanship, because using the simplest of tools, he was able to produce wonderful turnings with great skill.

I was fascinated by this, and I realized Al's vise, with it's two spindle centers, ought to make a fine bow lathe. That's why I made the base just a bit wider, for more lateral support. And now we see what those mysterious holes are for. I can run a length of steel rod through them, and that serves as a fixed tool rest for the lathe chisel. The upper hole is actually a mistake; it's on center with the spindle centers, but that means the upper edge of the tool rest is above the axis of rotation, so I added the lower hole for better position.

Shaving down the bow with a spokeshave. It's from an oak log I had rived down to small spindles a few years ago, so it has good strong continuous grain down its entire length.

Bow lathe mode. You can just see the steel rod tool rest. The clamp on the press screw handle keeps it from turning. There's also an extra nut cranked down against the tee nut on the thumbscrew to keep it from turning.

I'm not going to show you my practice turnings. They look like they were gnawed on by rats. My skill with a lathe is poor enough to begin with; take away one arm to power the bow, and it's clear this will take some practice to learn!

I'm also not too happy with the tool rest solution. There are some other nice bow lathe videos that show up YouTube along with the one from Marrakech, so I'll have to study these all a bit more closely to develop the technique. Otherwise, functionally this all worked fine. The vise stayed solidly in place on the bench under the forces of the lathe action.

I do see that I need to build a spring pole lathe. Roy Underhill has just the one. But that's for another day!

Peter Follansbee, SAPFM New England Chapter May Meeting

Left to right, Peter Follansbee, Freddy Roman, and Phil Lowe.

Freddy Roman organized another excellent monthly meeting for the New England chapter of the Society of American Period Furniture Makers this past Saturday. Hosted by Phil Lowe at his Furniture Institute of Massachusetts, it featured Peter Follansbee demonstrating 17th century green oak joinery and carving. Videos of the day appear at the end of this post.

Peter has an excellent website and blog packed full of information. A joiner at the Plimoth Plantation living history museum in Plymouth, MA, for nearly 20 years, he's a protege of Jennie (formerly John) Alexander and Drew Langsner, doing green woodworking (woodworking using undried, recently felled wood); he also teaches at Drew's.

He has an article from the Spring 2009 issue of Woodwork Magazine in PDF form here, covering much of what he showed us. He also has a new instructional DVD available from Lie-Nielsen. He's an excellent presenter, so if you ever get a chance to see him in person or buy his DVD, I highly recommend it. He's used to working in front of inquisitive audiences at Plimoth Plantation.

He works almost exclusively in green oak, building carved and joined chests and stools in 17th century New England style. Many of these items came over from England at the time. His primary method (he might say only!) of joinery is drawbored mortise and tenon, often in frame and panel configuration, without glue. The fact that existing pieces have survived four centuries attests to the durability of this method.

One of Peter's reproduction chairs built using these techniques.

Peter's primary stock preparation planes: top, a smoother with iron ground to scrub profile; bottom, a shop-made jointer. I love the jointer, because it's literally made from a piece of firewood. Crude as it is, it still does great work.

His simple travel toolbox, in the style of Toshio Odate's book "Japanese Woodworking Tools". This kind of work requires only a small set of tools. He carries around the wooden plane that's resting on top to show the indentation on the side at the front from the user's thumb. How many years of use did it take to wear that into the dense beech wood?

For the morning, Peter demonstrated stock preparation and joinery. He had several pieces in various stages of readiness, including a turned stool leg ready to receive its rails. He actually works very quickly, but takes lots of time to explain details and answer questions.

Out in the parking lot, Peter shows his hewing hatchet with cranked handle as he prepares to take the twist out of the piece of riven oak in front of him.

Back in the shop, using the scrub plane for fast stock preparation. Where the hatchet was coarse, this is medium.

Fine: final flatness with the jointer.

One of the interesting things he showed us was how joiners didn't bother forming parts into perfectly rectangular cross-sections. They left the inner, hidden face truncated, in what Alexander dubbed a "truncadon" in this article. This allowed them to use narrower stock.

Illustrating the "truncadon" cross-section dressed out of a wedge-shaped billet, leaving one corner off the rectangle.

Chopping a mortise.

Cutting the tenon shoulder on a rail.

Boring a peg hole in a tenon. The hole is slightly offset from the matching hole in the mortise, so the peg draws it up tight when driven in. The resulting drawbored joint requires no glue.

Shaping a drawbore peg. He doesn't use a dowel plate, he just rives off small peg billets, then shaves them down with a large framing chisel.

With the pegs driven in and sawed off flush. That joint is tight and solid.

After lunch, Peter showed us a variety of carving patterns and techniques. The work is nailed to a backing board, which is secured to the bench with clamps or holdfasts. He moves around it to accomodate the grain and the line of the cut, rather than moving the piece. This is used to decorate rails and stiles, as in the photo above, or panels that may be nailed together into small boxes or joined into frames for larger chests.

Peter's modest collection of carving tools.

An example finished panel showing a symmetric S-curve pattern. He said this is about 45 minutes of carving.

One thing Peter emphasized repeatedly was postures for control. The following photos show a variety of grips and postures while making curved cuts, both with and without the mallet.

Left elbow in tight, legs spread fore and aft to lower the body, driving the V-tool with a mallet.

Forearm resting on the work, following the curve in one direction.

Coming at it in the other direction. The workpiece stays in place as he moves around to accomodate grain and carving line.

Pivoting with the feet and torso. Some of this is full-body carving, not just the hands. It's very dynamic.

Rear view of foot placement.

The resulting carving, with another example on the bench.

Once the initial V-tool work has been done, further refinement occurs with bevel up...

...and bevel down. Some of these are sweeping cuts to follow the curves. If a chip is not coming out, wiggle the tool, don't flip it up, because that can cause a bigger chip to tear out.

Nailing down another panel. Many period pieces show the nail holes, they weren't done in the waste.

Closeup of the crisp carving and one of the nails. The dressed panel has seasoned for a week before carving to just dry out the surface.

Layout is done with marking gauge, compass, and square.

Peter disassembles the panel he built for illustrating frame and panel construction when he first started at Plimoth Plantation. This shows the back side, where the panel has been raised with the hatchet and plane to fit in the grooved frame; that also deals with any variations in thickness, presenting a uniform flat face in front.

The front side of the panel is carved. The pegs are omitted from the left-hand mortise and tenon joints for easy disassembly.

Here are a couple of chairs built by our host, Phil Lowe. These are deserving of a SAPFM meeting and blog post all their own, don't you think, Freddy? Hint, hint?

It was quite an inspiring day. I badly wanted to try this, and now I have the opportunity.

In central Massachusetts, there's always an oak down somewhere nearby. One had appeared just down the street from my house a few weeks ago. Since my wife had put her foot down about bringing stray logs home, I had resisted the temptation.

However, after seeing Peter, I just had to get some material to practice with, so I popped on over this evening after work to say hello. Of course, it had all been cut to convenient firewood length by the power company, but I figured there was still enough for some small carved panels and rails and stiles.

I found two marvelously straight-grained sections unsplit. One whack with the sledge on the wedges sent them clear through after they were set in a score line. Even the 1/16th splits came out beautifully. There are just a few minor defects in the wood.

Thanks, Jane, I'll make you a nice little box from some of this!

About 15 minutes of work with the tools shown (though I didn't actually need the froe for these splits; it's one I ordered from Drew Langsner 5 years ago). I didn't quite get the puzzle back together properly when I got it all home, but you get the idea.

Here are two videos of the day. Consider these the motivational programs; get Peter's DVD for the instructional program.