In reality, I have been doing some smaller bits of work on the project as time has progressed.  I just have not had the time to write about it here.  Now, with some distractions out of the way and more time to send in the shop, I’m back at it with a vengeance!

The next step in the process was to tackle the headrest.  If you’ll recall after coopering the headrest it had been in hibernation for a while now.  There actually is a fair amount of work to be done with it before permanently attaching it to the chair.  However, as you can probably guess before doing that work I needed to make a jig to help facilitate the process!

The jig is actually a very useful one for cutting circles and/or arcs at the band saw.  You can see from the pictures that the jig consists of a base with a movable arm that holds a dowel pivot point.  In addition, for this task of cutting an arc along the front and rear faces of the headstock there is also a temporary carrier board with some scrap L-shaped supports to hold the piece while cutting.

The headstock arcs are cut using a 29 ½” radius on the jig.  I attached the rough piece to the supports with double stick tape and cut the rear arc first.  After that cut was made I shifted the carrier board so that the finished thickness of the headstock would be 1 ¼”.  I then cut the front face of the headstock again with the same 29 ½” radius.  You can see that on the second cut the band saw blade busted out from the piece and left me with a little extra clean-up to do.  I cleaned up the band saw marks with a sanding disk on the angle grinder for the coarser areas and then followed up with the random orbit sander at 120-grit.

With the front and rear arcs completed the next step was to trim the ends of the headstock to fit between the rear legs on the chair.  This was done with the same six-degree angle jig I made earlier to cut the rear legs.  This time it was used along with the sled on the table saw.  This was actually a bit of a tricky cut in that the front of the headstock wants to drop forward as you complete the cut.  I did this cutting in stages and crept up on the fit.  The six-degrees was close but not exact and that, coupled with some variance in the rear leg angles and flexing of the legs themselves, rendered a bit of fine tuning with a hand plane before a good fit was obtained.  To be honest, this was probably one of the more tricky parts of the build so far.  In the pictures you can see this process in various stages of completion.  What’s not shown is the drilling of holes in the headstock through the pre-existing holes in the rear legs to hold the headstock.  This was done with a long 1/8” drill bit (because I could not find a long 5/32” bit locally).  I then had to remove the headstock and enlarge the holes to 5/32” with a shorter bit.  While the headstock was mounted in the chair, I made marks to indicate where the tops of the rear legs and transitions into the legs were located.

The marks that I made helped with the next step of drawing arcs along the top and bottom of the headstock.  These were done with a shop-made trammel set at 41”.  I adjusted the far trammel point (with a fixed radius) until I was hitting the marks on each side of the headstock as I swung the arcs and drew them once it was hitting both marks.  These arcs marked the location of the next two cuts on the band saw.

With the headstock arcs cut I moved on to marking and drilling the headstock for the back braces.  I made a jig to locate each hole using those already routed into the seat.  These were evenly spaced across the headstock.  The holes were drilled in a two-step process at the drill press using a 25/64” drill bit.  The first hole for each ws drilled straight and the second hole was drilled at six degrees toward the front of the headstock.  This extra clearance allows the flexible back braces to move freely when the sitter leans against them.

Because there is a large radius along the top of the headrest I then made one last cut at the band saw.  This was to chamfer the top edge of the headrest so that I did not have as much material to grind away when shaping.  I did this by tilting the band saw table to 45 degrees and using a single point fence to help support the headstock while cutting.

In the final picture you can see the headstock after some rough shaping with the angle grinder and random orbit sander.  The headstock is not yet glued to the chair – it’s just held in place temporarily with screws.  Once all of the back brace processing is competed the headstock can be glued to the chair.

Before diving in to the sculpting efforts I performed one more operation on the arms.  This was to prepare the arm to rear-leg joint.  To do this, I first cleaned up each of the rear legs at the arm joint location.  These were still rough from the original band sawing.  I just did this with a few swipes of a sharp block plane.  Next, I clamped each arm at the front leg transition area and adjusted it so that it was fairly tight against the rear leg at the joint location.  Because the angles were no perfectly matched, these joints needed to be trued to one-another.  This was done with some strips of 100-grit sand paper between the two pieces at the joint.  I pulled the sand paper through the joint, pulling away from the side that the grit was on so that I did not round over the edges of the pieces.  This took a while with the grit alternating from the arm-side to the rear leg side.  After a bit of work I had good tight joints.  After I had done this and ripped several strips of sand paper in the process, I thought about putting some strapping/packing tape on the back side of the sand paper to strengthen it – I’ll file that idea away for the next time I have to do this which will be when I fit the rockers to the chair.

With all of the possibilities of stalling exhausted, I arrived at the point where I needed to again take the angle grinder to the chair.  After all of the work so far, I was both excited to start the sculpting as well as a bit concerned over this step.  This work was done with the angle grinder and a 36-grit disk so material was going to be removed very quickly.

I started with the relatively simple task of leveling the joints between the rear legs and the seat.  This got me again used to the motions necessary to smoothly move the grinder and the aggressiveness of the stock removal. Next I moved to the front legs for the more complex of the sculpting tasks.  The difficulty here was two-fold: first, sculpt the leg to reveal a smooth curve between the leg and the seat and leaving a continuous line of the front leg; second in doing so, work to move the transition of the joint between the leg and the seat away from the corner.  The second point was necessary in order to achieve a smooth curve between the legs and seat without having an abrupt 90-degree corner.  Surprisingly, this was accomplished by grinding further into the side of the leg – effectively moving the joint line away from the corner!

In the pictures you can see the progression of the sculpting on the front legs.  The first picture shows the original joint.  Each subsequent picture shows the steps of removing material to sculpt the leg into the seat.  Note how the joint line moves from the corner outward.  This was helped on the front side of the leg by relieving the front corner of the seat to allow better access for the grinder.

The last thing to do before putting the grinder on the shelf for a while was to form a round-over along the top and bottom of the seat sides in between the legs.  To do this I first marked a line along the edge about ½” in from the corner.  I then chamfered between these lines with the grinder trying to keep a smooth line.  After that I made smaller chamfers along each edge to create the round-over.

Of course, all of this sculpting will still require more work with the rasp, files and sand paper to complete the job.  But, for now the chair is starting to look more sculpted.

The shaping process for the arms involves two steps.  The first step utilizes a jig that facilitates cove-cutting a grove diagonally along the length of each arm blank.  This cove cut removes stock that will eventually become the area that your arms rest in when sitting in the chair.  The second step of shaping the arms will occur when they have been attached to the chair and more stock can be removed and the curves can be faired into the legs.  Before doing any cutting, I first marked the outlines of the arms on the blanks so that I could see their eventual orientation and not get confused while performing the following steps.

The jig for the cove-cutting at the table saw is a rectangular box with an angled portion that rides against a fence while transporting the arm blank over the table saw blade.  The angled piece allows the blank to travel, top face down, across the blade so that the resulting cove is cut diagonally through the length of the arm blank.  These cuts were done in a slow and methodical manner, raising the blade about 1/32″ – 1/16″ for each pass over the blade.  The jig was moved across the blade slowly to allow the blade to cut the cove as smoothly as possible and to also allow the significant dust to be cleared.  Even with this technique I stopped every few passes and used the shop-vac to clean the dust remaining on the table top.  The jig has handles with threaded rod through t-nuts to clamp the blanks in place as well as a handle to help push it through the blade.  I also used a Gr-ripper push block for added control and security.  The jig flips over and the handle reverses for the alternate arm to be cut with a reverse orientation of the diagonal cove.  The first two pictures show the right arm being cut and the last picture shows the left arm.

Once the cove cuts were completed I again traced the outlines of the arms onto the blanks (albeit a carefully due to the now undulating surface of the arms).  Then the arms were band sawed to shape and sanded with the oscillating spindle sander to remove the band saw marks along their edges.  In the picture you can see the rough shaped arms after cutting and sanding.

The other step that I completed at this time was to glue up stacks to the rockers.  These stacks consist of rocker strips and are located at the points where the front and rear legs will contact the rockers.  The stacks serve to elevate the chair above the rockers an additional distance as well as to provide material for shaping and fairing curves from the rockers into the legs.  This was a bit of a tricky operation as the stacks wanted to squirm away while they were being glued.  I again used the same rocker clamping caul to distribute the clamping pressure to the stacks.

Some of the shaping of the legs could be done with a round-over bit in a router table however, because of the irregular curves of the legs some round-overs required a unique fixture to be used with the router.  This fixture is the donut-shaped piece that you can see in the picture.  The donut allows for the tight curves near the leg joints to be rounded over reasonably well.

I first completed the round-overs on the flat faces of the rear-legs with a 5/8″ RO bit at the router table.  Next I moved to the router fixture with the same bit and proceeded to round over the other edges (all except the inside long edge).  This operation was a bit tricky and required a good deal of attention to stay safe and to achieve good results.   Noting the direction of rotation of the router bit (as it related to the curves) was very important so that areas would not tear out during the routing operation.  I needed to always make sure I was routing down-hill or I stopped short of the bit exiting the piece.  I kept my hands a good bit away from the spinning bit and in some instances avoided routing all the way to the ends of the stock to avoid any chances of kickback or over cutting.

Because I did not route to the ends of the work pieces on all edges, there was some cleanup to do with a spokeshave , rasp and file.  The ends of the legs will ultimately be shaped to final dimension when they are attached and faired to the rockers and arms at a later stage in the chair build.  However, there were also other areas in which the round-over bit was not able to do a complete job and/or where areas needed to be blended to create a flowing curve.  This hand shaping was actually quite enjoyable.  There’s something very organic about shaping wood by hand with these kinds of tools – it establishes a real connection with the material and form that you are creating.

Once the shaping of the rear legs was completed I moved on to the front legs.  The first step was to trim the bottoms of the legs at an angle (~8 degrees).  Then, I was able to lay out the leg outlines from a template for band sawing.  The front legs were band sawed to follow this outline and then a secondary curve was free-handed to define the final leg thickness.  After the legs were marked, they were band sawed again being careful to maintain a square cut while accounting for the irregular surfaces of the legs.

I followed a similar process to shape the front legs, using the same router fixture and 5/8″ RO bit.  However, due to the smaller size of the front legs I used a quick clamp (minus the soft rubber pad) to hold each leg from the center at the leg joint in addition to holding one end with my hand so that I could stay safely away from the router bit.  A similar bit of hand shaping was required after the rough router work.  With all of the initial shaping done on the legs I put them into position for a test fit.  It is starting to look like a chair!

Next up: Creating and rough shaping the arms

When I last left the project, I had completed sculpting the seat.  With that task completed my attention turned back to more processing on the rear legs.

The rear legs needed to have a 20-degree angled cut from the top of the legs down to the arm rest area along their inside edges.  Also, a parallel cut to this one needed to be made at the top of the legs (to match the height of the headrest) along the outside edges.  The result will form a parallelogram shape at the top pf the rear legs which will match the angle of the headrest.

These angled cuts needed to be made at the band saw.  However, because my band saw table (like most others out there) will only tilt greater than 20 degrees in one direction (i.e. it is limited to about 10-degrees of tilt when angled back toward the saw) I needed to either cut one leg in the reverse direction or to find a way around the tilt limitation.

I chose to create a simple 20-degree jig that could be clamped to the saw table and reversed for each leg cut.  This allowed me to make the cuts from the tops of both legs downward toward the arm rest location which is the most desirable direction.  You can see the jig in the pictures.  The riser blocks beneath the jig are just there so that I had more clearance when rotating the legs during the cut so that they did not run into the saw table.

As, I mentioned, there were two cuts to do on each leg.  The first cut was from the top down to the arm rest area.  Then the leg was rotated and the second cut was made from the top downward to match the headrest height.  In the pictures you can see the cuts as well as the parallelogram shapes that resulted at the tops of the legs.

With these shaping cuts completed I next moved to the seat joint areas of the rear legs.  As you will recall, these curves have only been roughed shaped and marked along the seat top and bottom for later shaping.  To do this shaping, I first penciled in the curves to meet up with the lines that were scribed earlier with a shallow saw kerf.  These curves were based on a 2” diameter circle – the reason for this will become clear shortly.

Next, it was back to the band saw to cut along these curves.  The important thing here was to keep the leg balanced on the seat joint area during the cuts.  The picture shows the results. With the band saw work completed.  I moved to my oscillating spindle sander with a 2” diameter drum and I sanded these rough curves to shape until I just barely met the scribe lines marking the top and bottom of the seat. This resulted in legs that will require only minor shaping where they meet the seat.

With this shaping completed, it was a good time to drill the holes to allow attachment of the headrest.  This was done in two steps.  The first hole was drilled with a 3/8″ Forstner bit to counter sink for the the screw head.  With the leg in the same position I switched out to a 3/16″ bit to drill through the leg for the screw shank.  Doing it this way assured that this hole would be perfectly centered on the countersunk hole.

Well, maybe after a quick commercial interruption…you’ll recall that in the last post I had glued the two perfectly good front legs together with a piece of two inch stock in between them.  Well, of course they were not going to stay that way.  So, I traced a curve onto the adder piece between the legs and then cut them apart on the band saw.  The front legs will be cut and sculpted further at a later stage.  Now that that was out of the way, I could turn my attention to the first efforts of sculpting on the chair – shaping the seat.

To start this process, I first layed out the outline of the seat area from my templates.  Next, I carefully layed out the locations for some depth holes.  I drilled these holes into the seat to help me judge when the sculpting had reached the desired depth.  The rear holes were drilled at 3/4″, the center holes at 5/8″ and the two sets of front holes at 1/2″.

To do the rough shaping I used a coarse Kutzall donut-shaped disk in an angle grinder.  This disk removed material very quickly and produced copious amounts of dust!  However, if careful I was capable of fairly fine work with it.   The key was to use slow, steady movements of the grinder in the direction against its rotation.  I started the shaping at the rear of the seat, staying about 1/4″ away from my layout lines and gradually working toward the finished depth.  When I was about 1/8″ away from the finished depth I moved to the front to work it to a similar level.  The center keel of the seat had to be shaped manually as I went along.  After I got the entire seat to within about 1/16″ of finished depth, I carefully used the grinder to shape up to my layout lines.  The sequence of pictures below shows the process from start to finish.

The next step was to move on to sanding with the random orbit sander.  I started this with 60 grit to even out the coarse surface left by the grinder and then moved on to 120 grit.  I also used folded sandpaper and my thumb to ride the curved surface along the edge of the shaped seat leaving a crisp line along the top.  In the picture, you can see how rough the surface was prior to sanding. A little more layout on the front of the seat, some filing and some sanding and I had contoured the front of the seat to allow the user’s legs  to wrap nicely over the front edge of the seat.

In the final pictures you can see the final shaped seat ready to be joined with the legs.

Next up: Shaping the rear legs and arm rests

As you will recall, in my last post I created the joints in the seat that will receive the legs.  These joints essentially have a tongue that must mate with a corresponding grove in the legs. There’s a bit more too it, as you will see, but that’s the basic premise.

Before I could create this joinery on the legs, I needed to do two things: for the rear legs, I needed to create profile on the inside faces and visible from the front and on both the front and rear legs I needed to add some material to make parts of the legs wider.

In the pictures you can (barely) see the outlined profile toward inside of the rear legs which was to be cut on the band saw.  Because the legs had already been contoured, these cuts were a bit tricky.  I was careful to be sure that there were always two points of the leg in contact with the saw table – sometimes one of those points was at the edge of the table as the leg was either entering the table or leaving the table. This allowed the cut to be completed safely even though it was sometimes happening at a point elevated above the table. In the second picture you can see the profiles cut on the inside of the rear legs.

After profiling the rear legs I milled adder pieces for both the front and rear legs.  The front adder piece runs the length of the legs. The rear adder piece runs just between the sections that will eventually contact the seat.  For the front legs, the adder piece will be profiled to become the outside of the legs. For the rear legs the adder piece was ripped at the band saw after gluing in preparation for more detailed operations to follow.

The rear legs of the chair need to cant outward at the top at an angle of six degrees.  In order to accomplish this, the adder pieces on each rear leg must be ripped at six degrees from the outside faces of the legs.  In order to accomplish this, I created a jig to use at the table saw.  The jig is designed to ride against the fence on the saw and it has a fence of it’s own that is fixed at six degrees to the table saw blade.  With a leg clamped to the jig the piece was pushed through the blade cutting off a six degree wedge. For the other leg, the jig was rotated 180 degrees and the other leg was cut in a similar fashion.

I took this opportunity to clamp the legs together in the vise using the angled offcuts from the previous operation so that I could smooth the seat joint area and square it to the angled faces which were previously cut on the table saw. After achieving smooth and square faces, I laid out for the joinery that will join the legs to the seat.

The notches to accept the seat joinery were cut on two faces of the rear legs using a sled on the table saw.  Because each leg had only a small bearing surface resting on the sled, I clamped a block to the outboard side of the legs to prop them up at the correct angle.  One leg was cut in the orientation as shown in the picture and the other leg was done at 180 degrees from this orientation. The notches were nibbled away with repeated passes of the leg over the table saw blade until the joint would just fit the seat. Once the rough notches were cut, I cleaned up the saw marks with a shoulder plane.  The last step for this joint was to round over the inside corner of this joint to match the corresponding rebate that was cut into the seat.  I did this with a ¾” round over bit in the plunge router.  Because of the small size of the area where I could rest the router, this did not give optimal results. So, I subsequently repeated this step with the same bit in the router table.  In the picture, you can see the general fit of the rear leg into the seat.

The front leg joinery was done in a similar manner to the rear legs.  However, the notches were made on three faces of the legs. Also, this joinery was done prior to gluing the adder piece to the two front legs.  In the pictures you can see the competed front leg joinery and the adder piece being glued between the outside surfaces of the two front legs after the joinery was completed.

Next up: Sculpting the seat

The joinery starts on the seat blank.  So, I first used a sled on the table saw to trim the seat to finished length and width. The rear legs will be joined to the seat via a 3 inch by 3 inch cutout that is rabetted top and bottom on each back corner.  The front legs have a  similar configuration except they are set back from the front corners into a 2 inch wide notch in the seat.

These joints need to be very precise and with the thickness and size of the seat blank, that’s a challenge. In order to create the notches to exact size, I first laid out the joints with a pencil and then rough cut them at the band saw.  This allowed me to get very close to the lines without worrying about the precision of the joints. The rear leg notches were cut out as single blocks.  However, because the front legs are attached set back from the front corner, I used the band saw to create a series of kerfs to rough out the notches for them.

I created a couple of jigs to help with the process of precisely finishing these notches.  In the picture you can see the front and rear leg jigs that I made to use with the router with a top bearing pattern bit.  The openings on the jigs are exactly the size of the notches to be cut and the bearing on the bit rides along the edges of the jig to trim away excess material from the joint leaving a perfect notch.  Because of the thickness of the seat, my pattern bit would not cut the entire rear leg joint in one pass. So, I first used the rear leg jig from the top and then again from the bottom to finish the joint.  After routing the rear notches, I used a sharp chisel to remove the radius left by the router bit and to square up the inside corners of the notches.

The second operation for this joint was to create a rebate along both the top and bottom of each notch.  This effectively leaves a tongue in the center of each notch that will later mate with a corresoponding groove in the legs. This was done with a series of shallow passes with a 1 1/2″ by 1/2″ rabetting bit.  In the pictures you can see that I used a scrap block clamped to the outside of the seat to eliminate any tearout. The resulting rebate leaves a 3/4″ radius on the inside corners of each notch.  Later the corners of the legs will be rounded over with a corresponding 3/4″ rounding over bit to create complementary profile on the legs.

With the joints for the front and rear legs completed I setup another jig to rout holes for the back braces where they enter the seat.  These were done with a template that I got from Hal Taylor.   The holes were routed with a 1/4″ straight bit and a 5/8″ OD collar on the router.  These holes are actually slightly oval in order to allow the back braces to flex slightly as you sit in the chair.

Next up: Leg Joinery

There are a bunch of potential ways to do laminations of this type – everything from a using a vacuum bag to building a form and using veneer screws to create a press. I don’t have any vacuum press equipment or veneer screws so I elected to make a couple of forms to be used with clamps for this process.

The forms were built out of construction lumber that I laminated together to double thickness and squared up with the jointer and planer. I then band sawed the curves into each form and carefully smoothed the curves to eliminate any flat spots  This is especially important for the rocker form because a flat spot on a rocker will be easily felt when rocking in the chair.  Next, I drilled a series of holes in each form to accept the heads of C-clamps. Along one edge of each form are metal stops (I used some scrap aluminum angle and steel straps that I had lying around). These stops allow the parts to be aligned evenly along one edge while they are being laminated.

As a first step, I laminated a backer strip on each form with the extra strips that I had previously cut.  This allowed be to get acquainted with the clamping process and to determine how much glue to use and open time I would have.  After the backers were dried I cleaned up the glue squeeze-out and trimmed them to be slightly narrower than the finished parts will be.

The moment of truth arrived and it was time to laminate the first rocker. I went about the process of spreading the glue as quickly as I could. I am using Titebond III for the gluing and it does not allow a lot of extra open time – the rockers have 9 strips so I had to move fast.  With the wet stack on the form I went to work clamping – progressing from front to back on the rocker with the stack lightly clamped to the metal alignment strips.  In the end it worked out OK, but let’s just say I had no time to spare when tightening that last clamp and let me tell you, there was a lot of glue squeeze-out! I left the rocker to dry in the form for 24 hours.  Laminating the second rocker went a bit smoother as I was more prepared and familiar with the sequence of operations.

In the picture you can see the stacks of strips for the back braces.  I am using Ash for the two inner strips for flexibility (I am also using original Titebond for this to avoid the dark glue lines that Titebond III would give).  The Walnut strips on the front and back of the stacks were cut in sequence from the same piece of stock to give a visual repetition to the braces when viewed from the front and back of the chair.

The process for laminating the back braces is similar to the rockers except that they are done two at a time – one on top of the other. This process does alter the curves of the top brace somewhat when compared to the bottom brace. However, they are laminated in an order such that the bottom braces (with the greatest curves) are placed toward the outside of the chair and the top braces are placed toward the inside.  This results in a change in the curves that is consistent across the back of the chair. With seven back braces, this arrangement requires four gluing operations with a 24 hour drying period in between each.  While that’s going on, I can move on to some joinery for the chair.

Next up: Joinery for the seat and legs

The headrest (as well as the back braces of the chair) curves substantially to provide a more comfortable position while sitting. To obtain the necessary curve for the headrest you either need a very thick piece of stock (with a lot of resulting waste) or the pieces must be put together using a process known as coopering. For this headrest I am coopering individual pieces of stock with the grain oriented vertically.

The coopering process is much like the process that would be used to build a wooden barrel. The individual pieces are beveled on along their edges so that when glued together they form a segmented curve. The bevel angle on each edge can vary depending on how much of an overall angle or curve is desired and also how many pieces are being assembled. In my case, I needed an overall width of about 26″ for the headstock with an overall included angle of 40 degrees. I started with six pieces at 8″ high by about 4 to 5 inches wide.  The six pieces provided 5 interior glue joints (the ends of the headstock are not beveled). Each of these joints has two edges to bevel.  Put that all together and you have 10 bevels to provide a total of 40 degrees which equates to 4 degrees per bevel.

To create the bevels I elected to make multiple passes over my jointer with the fence set at 4 degrees from vertical. With a setup for a very light cut (less than 1/32″ per pass) this operation was much safer than using the table saw with the relatively small size of these pieces. After the beveling was done on the jointer I elected to take a few swipes with a hand plane over each edge to eliminate the mill marks and to assure a pristine glue surface.  In the picture you can see the pieces dry fit after beveling the edges.

Because of the bevels on these pieces, the glue up operation was very tricky.  The glue up of the six pieces was initially done in three pairs. Next, two pairs were glued to each another with the use of angled blocks faced with sandpaper adhered to the clamps. Because the headrest now exceeded the depth of the heads on my clamps, the final glue up required the addition of some temporary blocks glued to the headrest to act as clamping points. At a later stage, these glue blocks will be cut off and the curve will cut into the headstock at the band saw.  In the pictures you can see the sequence of this series of gluing operations.

Setting the glued up headstock aside, I moved on to cutting the thin strips for the laminated rockers and back slats. This was a repetitive operation that required precision and safety.  So, I created a simple jig to allow me to rip thin strips at the table saw to a very tight dimensional tolerance. The jig also allows the strips to be ripped on the waste side of the blade so that strips can be ripped from a larger piece of stock and avoiding the dangers of trying to do this between the blade and the fence.

Each rocker requires 9 strips at 1/8″ thickness and each back slat requires four strips at 90 thousandths of an inch thickness. For the rockers, I made additional strips because some will be needed to create stacks under each leg where they join the rockers.  I also made some extra strips to create a curved backer for use when these strips are later laminated with a clamping form. There are seven back slats required but I made an extra for safety as well as some additional strips to create a backer. for their glue up as well.  In the picture you can see the resulting strips for the rockers.

Next up: Bent lamination and creating the leg joints in the seat

I especially admire Maloof’s rocking chairs. These are the pieces that he is arguably most well known for. The flowing hard and soft lines of his design work together in a truly elegant form to create a work of art in wood. For many years I have wanted to make the time to create a chair patterned after the Sam Maloof designs.

There are several current rocking chair makers that have patterned their work after Maloof.  One, in particular, seems to taken his queues from Sam but also added some unique design and functional elements as well. Hal Taylor was originally inspired by Sam Maloof and has developed a chair design that goes beyond Maloof in both aeastetics and function. Hal has been building his chairs in Virginia and refining his design for many years.  He has also taught many people to build his chairs in classes at his shop. While I don’t have the ability to do a class with Hal right now, I contacted him and I will be building a chair based (at least initially) on plans of his design.

This chair calls for about 35 board feet of 8/4 stock. I will be making mine from Walnut and I purchased about 45 board feet to cover any waste generated during the build process. Most all of the chair components need to be milled close to a full 2 inches thick before any joinery is done on them.

The first step in this effort was to create hardboard templates for most of the chair components. Once I had the templates cut out and their curves smoothed, I took some time to read the stock and to identify where each of the chair parts would come from.  The most important parts here were the layout of the rear legs and the seat. I was lucky enough to have obtained some fairly wide stock so I could get the rear legs and seat components from the same board. This allowed for consistent color and grain characteristics for these parts.

At this stage I just marked the parts on the rough planed stock and cut to rough dimensions.  I wanted to be sure that I had enough stock for all of the components before starting to cut to critical dimension. In the first picture, you can see the rear legs marked and their orientation for best grain match. The next picture shows all of the major pieces rough cut and ready for further work.

I cross-cut the seat on the table saw using my shop-made cross-cut sled. My stock was wide enough for me to make a symmetrical 2-board layout for the 21 1/2 inch wide seat.  However a three, four or even five board seat would also be attractive with the proper layout for grain. You can see the glue-up of the two seat boards in the next picture – the orientation of these boards is with the growth rings arching upward when viewed from the front of the seat. This will yield the most lines of grain to be cut through when sculpting the seat out at a later stage which should produce a very attractive seat.

Satisfied that my layout of the remaining parts would work, I set out to cut the back legs next.  I traced the outline onto the Walnut following the template using a white pencil and went to work at the band saw. I used a 3/8″, 3TPI blade for this task. It could handle most of the curves, but to avoid any binding I chose to make relief cuts into a few of the curved areas before following my lines to free the pieces. I stayed about 1/16″ away from my lines for the most part.  The result was two legs that are very close in shape.

Next up: The coopered headrest and cutting thin strips for the rockers and back-slats

I’m not sure how I liked working with the Hard Maple on this project. Even with careful reading of the grain of each board, the wood was very prone to tear-out with the planer and jointer. This was with wood of only relatively moderate figure. Because of this, the stock required a lot of work with my smoothing plane and a card scraper in order to smooth the areas that showed torn out grain. I usually smooth the surfaces of project parts to eliminate mill marks but this wood required more that that.

Of course, I did all of this smoothing work as I went along creating the various parts for the table so all that was left after assembly was sanding. I always hand sand my projects using a cork-faced sanding block and this was no exception. The wood was generally smooth from the planning and scraping. So, all that was necessary was a bit of blending of the planed and scraped areas – I did this with 180-grit paper.

The finish on the table is a simple one. I did not want to add too much color to the Maple but, I did want to warm it up a bit. I made a test board surfaced with the same processes as the table and tried applying Boiled Linseed Oil, blonde Shellac, and General Finishes Arm-R-Seal in various combinations. In the end, I chose straight Arm-R-Seal. For the first coat, I scrubbed it into the surface with 400 wet/dry paper. After that coat dried over night, I scuff sanded the finish with the same grit paper, wiped the sanding dust with Naptha coated cloth and applied second a third coats of the finish.  After letting the finish cure for a week, I applied paste wax with 0000 steel wool and buffed the surface.

In the pictures, you can see the finished result. Sorry for the quality of the pictures, they were taken in a hurry just before the table was packed up and sent off to its new home in South Carolina with my in laws.

In the end, I think that the arched aprons and top, the inlay and the splayed legs on this table helped to modernize the basic shaker-inspired design. After it arrived in its new home I received a report that it looked great and that the Maple/Cherry combination worked perfectly with the dark wood floors of the house. I’m hoping for a picture of it in place to see for myself.

However, you may recall, the design for this table calls for the legs to splay outward to the left and right by 2 degrees.  In order to accomplish this I needed to accurately cut the ends of the front and back aprons at 2 degrees.  I used my miter gauge at the table saw for this operation.  With the guage  set to 2 degrees, I cross cut one end of each of the aprons.  To bring the aprons to the exact length I next added an extended fence to the miter gauge.  I placed a stop block on the extension so that I could cross cut each of the aprons to the same exact length – again at 2 degrees on the opposite ends.  In the picture you can see the result of these cuts on the font and back aprons.

The aprons for the left and right ends were easier.  With the apron stock already cut to finished width I simply cross cut the pieces square with the miter gauge square at the exact needed dimension.  Again to assure that the parts were dimensioned exactly the same I used a stop block for the second cut on each apron.  The reason that I mention the use of the stop block and the batching of these cuts is that by following this procedure you can machine  parts to their exact dimension without changing setups and possibly introducing error.  The fact that both pairs of these aprons are cut to exactly the same length assures that I will have a much easier time of creating a square assembly when the table is ultimately glued up.

After getting the aprons cut to length I moved on to creating the arches in them.  I laid out a fair cure using a flexible piece of scrap stock and transferring the curve to the stock with a pencil.   I then used double-stick tape to temporarily stick the aprons together and cut the curves on the band saw.  I stayed about 1/32″ away from my lines and after the curves were rough cut I smoothed them with the sander.  This operation was also done while they were still taped in pairs to assure that the curves for each set of aprons would be identical.

After shapiing the curves on the aprons I moved on to creating the mortises in both sets.  Usually I would do this with the stock still square to allow for easier clamping and squaring.  However, for the loose tenon joinery I’m using to work properly on the front and back aprons, the mortises need to be cut perpendicular to the angled ends of the stock at the as they will be installed.  This allows me to use the simple mortises already cut perpendicular into the legs (recall that the top and bottom of the legs were previously  trimmed to 2 degrees to make everything flush at the top after the aprons and legs are assembled).  I used the same router mortising jig for this task.  The picture shows an apron in place and one of the two mortises already cut.

Next I’ll tackle adding the cherry inlay into the aprons and the top.  Then it’s on to assembly.

In this case I used two rough boards about 6 inches wide to create the top.  I followed my normal milling process to machine the boards square and true and cut to rough length and width.  I took a good look at these boards to determine their best top sides and also which edges to glue together as the center of the table top.

With Hard Maple like this, the sap wood (the wood that grew toward the outside of the tree) usually has the nicest creamy color.  So, I first tried to orient the boards with the growth rings curving downward when viewed from the ends (i.e. with the sap wood facing up).  I then tried different orientations of the boards shifting them slightly until I found the most pleasing transition of grain at the center seam.  I settled on an orientation and glued the boards together.  In the pictures you can see the steps for the glue up.  I some used Parallel Jaw clamps to apply even pressure to the boards along their edges and also some Quick-Grip clamps at the ends across the center seam to keep them as even as possible while the glue dried.

After a few hours drying I unclamped the boards and scraped the glue squeeze-out away with a sharp cabinet scraper.  After the dried drops of glue were removed I applied some mineral spirits to the top to see where I may have missed some glue – it’s a little difficult to see the glue on the creamy colored Maple.  The mineral spirits helped to highlight any remaining dried glue.  In the picture, you can see that I marked the areas with glue and any tearout from the planer with a pencil.  With that done I went at the top with my #4 smoothing plane and the cabinet scraper until I had removed all traces of glue, mill marks and tearout from the top.

The next step was to mark and cut the arches on the ends of the top.  As you can see in the picture, I made a simple trammel with a long piece of scrap, a pencil and a screw as a trammel point.  Nothing real scientific here, I just varied the location of the screw until I obtained the arc that I was looking for.  Once I had that location I simply marked the length of the top at the center line and placed the pencil point there.  Then I held the  screw point lightly against the top on the center line and swung an arc across each end of the top.  No need for any fancy tools or jigs.

With the arcs marked on the top I set up some auxillary support at the band saw. I first contemplated creating a jig similar to the trammel to cut the arcs but, after thinking about it I felt it would be easy enough to just freehand the cuts and then clean them up on the sander.  So, I went the freehand route.  No problems there. With the extra support helping to hold the top I just swung it in an arc following about 1/16 of an inch outside the lines.  Once the cuts were made I cleaned them up on the edge sander until I reached the marked lines.

Next up will be milling and mortising the arched aprons and assembling the base.

I then broke down the rough stock into more manageable, slightly oversize, pieces. I previously wrote about my method of using the jigsaw and band saw for this process.  To me this is the safest, most convenient and economical way to do this.

For the legs of this table I want a finished 1 3/4″ square leg at the top, tapering to about 3/4″ square at the bottom.  I started with 8/4 rough stock and ripped it to 2″  square by 29 1/2″ long to accommodate a 28 1/2″ finished leg length.  For the legs, I tried to utilize the stock such that the grain lines run diagonal across the ends of the leg.  This yields the straightest grain along the faces of the legs.  In this case I used the more quartersawn faces for the front of the legs – this helps to keep any face grain on the legs from competing with the simple design of the piece.  I mention this here because it is important to realize that it is at this point in the build process where these kinds of decisions must be made.  In my mind, this is just an extension of the design process that started with the initial sketch. Choosing your stock wisely here will pay dividends in the look of the finished piece.

I milled the rough leg blanks to 1 3/4″ square and then cross cut them to length on the table saw using the miter gauge with an extension fence.  However, because the legs on the table will be splayed out to each side by 2 degrees, it was at this point that I needed to account for that detail.  I first made a pencil mark on one face of each leg at both the top and bottom to indicate the general angles to be cut – it’s easy to get these angles turned around in your head and to make a mistake.  So, a quick set of pencil marks to orient you when cutting is helpful.  I then set the miter gauge for a 2 degree angle and cross cut one end of each leg.  Next, I set a stop block on the miter gauge fence at 28 1/2″ and cross cut the other ends of the legs, again at 2 degrees.  Batching these operations to occur at with the same setup assured me that each of the legs would be cut with exactly the same angle and also to the same length.  In the picture you can see the angled cuts on the top ends of two opposing legs.

I am using loose tenons for the joinery on this table.  So, while the legs were still square in cross-section I machined the mortises to accept the loose tenons for both the front and side aprons.  The side aprons have only single tenon but the front and back aprons have double tenons for added strength.  I used the router mortising jig that I wrote about previously to create the three 1/4″ mortises in each leg blank.

With the angled ends and mortises cut on the square leg blanks, I could now move on to the operation of tapering the legs.  To start, on the two faces to be tapered I marked the legs at a point 4 3/4″ down from the top and at the bottom 3/4″ from each of the inside corners.  There are many ways to taper legs.  I contemplated just marking out the tapers and free-handing the cuts on the band saw.  I also thought about just using the jointer and making repeated passes to achieve the tapers.  Then, I noticed a little used jig in the corner of the shop.

This old jig was one that that I once used to cut straight edges on rough stock with the table saw.  I thought that it could be re-purposed for the process of tapering the legs of the table.  The jig is simple – it consists of a piece of melamine coated particle shelf stock from a home center and a couple of De-Sta-Co type clamps.  I re-oriented the clamps on risers to account for the thickness of the legs, added a couple of stops to account for the tapers to be cut and I was ready to go.

The jig rides along the fence on the table saw.  To cut the tapers I aligned the edge of the jig with the blade and positioned the fence against the jig.  Each leg was positioned on the jig aligning the 3/4″ end with the edge of the jig at the leading end and the 4 3/4″ mark at the trailing end.   The stock was pushed through the blade cutting a perfect taper – then the stock was rotated 90 degrees and the second taper was cut.  The important thing here was to cut the faces in the correct order so that after rotating the leg I still had a square face resting on the jig.  Doing so required the addition of a small block under the front clamp for each second taper cut.  In the picture you can see a leg blank in the jig with a taper already cut.

Next up I’ll start the work on the aprons and the top.

The dark floors in the house drove me to choose Hard Maple as the main wood for the project.  My first thoughts were to start with traditional Shaker styling due to its simple lines and to modernize the look from there.  Adding some curves was a consideration and with Maple as the primary wood, I also considered some other types of embellishment or accents for the piece without deviating too far from the clean modern lines.

In the picture, you can see the initial sketches for the piece.  I incorporated tapered legs that are traditionally found on Shaker pieces but, I decided to also splay them by 2 degrees to each side to give the piece a more modern and graceful look.  I also added gentle curves to the aprons and replicated those curves on the ends of the top to accentuate the more modern styling for the piece.  As you can see in the sketch, I also explored adding some darker inlay to follow the curves. My thinking was that this detail provided a bit more of a more modern look.as well as potentially providing some embellishment to the bland Maple.

Armed with these rough sketches, I moved on to Sketchup to further develop my ideas.  As you can see in the Sketchup drawing, I changed the side aprons to provide a more open and modern look.  I also simplified the inlay to avoid competing too much with the simple lines of the piece.

As shown, the piece is 28 1/2″ high, 48″ wide and 12″ inches deep.  This sizing should allow it to work either as a hall or a sofa table.  The designed height allows for the surface of the top to sit just below the top of most sofas.  The width fits nicely centered oalong the back of a standard 84″ wide sofa and also works well for a spot along a wall.  The depth is  enough to allow things to be displayed on the table top in either setting while not occupying too much space in a typical hallway.

This design effort is not unlike most where there are certain constraints and desires to be balanced in the resulting piece.  Rarely is a design effort for a functional piece of furniture an open book.  However, with the basic constraints in mind we are free to explore any alternatives for the form.

Next up I’ll work on roughing out the pieces for the table and commiting the design to wood.

The stock for the bowl started at about 8″ x 8″ x 3″ thick. I found the center of the stock and outlined the largest circle I could fit within the square. At the band saw I rough trimmed the stock round to make the truing up at the lathe easier. I mounted the trimmed stock to my faceplate centering it as best as I could. The faceplate was mounted to the side of the stock that will be the inside of the bowl so, the screw holes will eventually be turned away.

Once mounted on the lathe, with a bowl gouge I trued the edge until round and then flattened the face at low speed. With the stock true and balanced I started to rough out the outside shape of the bowl and also created a tenon that will be used to hold the bowl in a scroll chuck when hollowing out the inside.

In the picture you can see the rough shape of the outside of the bowl. With the outside turned I removed the stock from the face plate and mounted the tenon into the scroll chuck. Depending on the type of jaws your chuck has the tenon must be shaped appropriately. My chuck jaws have a slight dovetail recess so, I angles the sides of my tenon slightly to fit the jaws. It’s also important to be sure that the tenon does not bottom out in the chuck jaws because that tends to make the jaws grip a bit unstable.

The inside of the bowl was hollowed with a bowl gouge and finished off with a scraper. This was done slowly starting at the center of the bowl and working toward the outer edge while deepening the bowl. i worked the interior of the bowl until the walls were a fairly uniform 1/4″ thick. Once the hollowing was completed I sanded the interior with 100, 150, 180, and 220 grit sandpaper while turning on the lathe. With the interior sanding completed I applied a coat f Watco Natural Danish Oil to the inside, wet sanding it with 320 grit wet/Dry paper.

Next, I created a friction chuck from a piece of scrap. With some padding on the inside of the bowl I reversed it and applied pressure against the jam chuck with the tail stock. Centering the bowl to get it to run true was not too difficult because I had left the tail stock depression in the tenon from when the outside was turned. With the lathe running a slow speed I carefully turned away the foot leaving a small nub at the tail stock for support. Again, I sanded with the same grits and then applied and wet sanded Danish Oil to the outside of the bowl.

With the bowl largely completed. I removed it from the lathe and carved off the nub from the foot with a carving gouge and sanded the foot smooth. I signed the bottom with Sharpie and with a little oil applied to the foot, the bowl was complete. You’ve gotta’ love being able to complete a project in an afternoon!

I started by securing the body to the bench and aligning the end trim piece along the centerline of the tail stock. I carefully scribed with a knife along the edges of the end trim to define the mortise to be cut. With the initial lines scribed, I removed the trim and deepened the lines with the knife. I pondered whether to just chisel this area of the sides out or to use a router.
In the end I opted for the router to rough out the mortise. I set the depth of a straight cutting bit to equal the depth of the binding channels and made a couple of passes staying clear of the knife lines. With the bulk of the material removed I cleaned things up with a sharp chisel. Gluing the end trim in is not something that you can use a clamp for. So, I just used some tape to hold it in while the glue dried. Once it was dried I spent some time with a file to get the ends of the trim piece perfectly flush with the edges of the binding channels so that when the binding was installed there would be perfect seams with no gaps.

With the end trim installed I moved on to the bindings and purfling. Since there is only binding around the back of the guitar I opted to start with it – hoping that if I made any mistakes they would be less noticable on the back! I first carefully checked the fit of the binding all along the channel an noted any areas that were not smooth or of uneven depth. These areas were kissed with a file until the curves looked good. I also ran the inside bottom edge of the binding along a picece of sandpaper to chamfer it a bit and to allow the binding to seat fully.

I laid out my materials and got ready for the installation. I used binding tape from Stewart MacDonald to hold the bindings in place. Since these bindings are acrylic I also used some Weld-on cement appropriate for the material. Starting at the neck block I aligned the edge of the binding with the center seam of the guitar and applied cement for about 6 inches. I applied the tape by first pressing it against the binding and then stretching it first along the back and then down along the side. Every six inches or so, it was time for more cement. When I reached the tail block I carefully made a trim cut on the binding aligning it with the center seam. Once this half of the back binding was dry I applied a second strip on the opposite side of the back being careful to match th end seams of the binding. I let the bindings dry at least a full 24 hours. To remove the tape I slightly heated it with a heat gun ( a hair dryer would also work) to loosen the adhesive and pulled the tape at a 45 degree angle to avoid lifting any slivers of wood. Things looked good so, I moved on to scraping the bindings down to meet the back and sides. Notice in the picture that I put some tape on the corner of the card scraper to avoid nicking the back and sides as I scraped. (By the way that white piece you see on the card scraper is one of those flat magnet strips that I have on the scraper to avoid burning my thumbs as I scrape).

The procedure for the top was similar however, it was a bit more tricky because the binding and the purfling strips must be installed at the same time. Again I carefully checked the channels to be sure that things fit well, chamfered the inside edges of the strips with sandpaper and went ahead with the installation. The main task here was to assure that both the binding and the purfling were fully seated in their channels when applying the tape. This operation was also done in halves and left to dry for at least 24 hours before scraping down even with the top and sides.

The end result is pretty good. There are no gaps between the binding or purfling and the body and the seams in the strips look very tight. There is one small area where the binding is scraped down a little thinner than elsewhere ( I think that this is a result of an area where I bobbled and routed a bit too deep during the channel routing) but, it is not real noticable unless you are looking for it. I the pictures you can see a closeup of the binding an purfling along the top and a full shot of the body as it stands now. Next time, I’ll start turning my attention to the work on the neck.

I have to say that this was the most nerve-racking part of the build so far. After all of the work that has gone into building the box up to this point, the idea of taking a router to it was a bit daunting! As I will explain, this operation had the added complication of dealing with the arch of the top and back while routing to maintain the router in perfect position so that the channels were at the appropriate depth and width to receive the binding and purfling.

I didn’t take any pictures of the flush trimming operation, but I will walk through the binding/purfling operation because the same procedure and jig were used to first trim the top and back flush with the sides. Of course, this was done with a flush-trimming bit in the router.

What you realize very quickly when examining into how to perform this operation is that because of the arches in the top (1/5 degrees) and especially the back (5 degrees), the router base – even on a small laminate trimmer – will not allow the cutter to remain parallel with the sides while moving around the guitar. Similarly, because of the varying contour of the back and top the cutter may also have a tendency to move up and down during the cut leaving you with channels that vary in size.

There are varying ways to deal with the challenges posed by this situation. Everything from scoring and cutting the channels by hand with a specialized tool to creating an elaborate telescoping and/or vertically movable fixture for the router to mount on while moving around the guitar body. In the end, I opted for a compromise and created a simple jig for my Bosch Colt router. In the pictures you can see the jig clamped into position on the router base. The main section is designed to align with the bearing of a rabateing bit that cuts the correct depth channel. If If this section is held vertical and in contact with the side of the guitar while routing the channel will have a uniform depth. To help with this there is a small piece double-stick taped to the router base, effectively moving the base contact point very close to the edge of the guitar where the arch will not affect the width of the channel being routed. The last picture shows the router in position to rout the channel for the binding.

The operation started with trimming the back and top to be flush with the sides. There are some other issues to deal with here concerning grain direction. Because of the rotation of the bit and moving with and against the grain there was a danger of tearing out pieces of the back and top during the routing operation. As a result, I first made a series of clockwise climb cuts in select places and then followed up with a complete pass around the guitar in the standard counter-clockwise direction. The picture shows the path of the clockwise climb cuts in red and the following counter-clockwise pass in blue.

Routing the channels for the binding was next. Even though the back had more arch to deal with, I started there thinking that if I made a mistake, any fix would be less noticeable on the back. The four climb cuts came first and then the standard cut. This was done with a bearing on the rabateing bit which cut a channel that was bout .60″ deep to match the thickness of the bindings. Things went OK on this, with a couple of bobbles of the router that will require some hand cleanup. So, then it was on to the top following an analogous procedure. Again, things went OK but some hand cleanup will be required. Next, I switched bearings on the bit, reset the cutter depth and adjusted the jig to allow routing of the purfling channel along the top – same sequence of cuts with lots of trepidation! Again, a bit of hand cleanup will be required but, the stepped channels for the top binding and purfling are now in place.

In the pictures you can see the results of the operations along with the dry fit of the binding and purfling. The jig worked well. Probably the only thing I would change the next time is to use a slightly smaller piece for the contact point so it is affected even less by the contour of the back and top. Also, I found that keeping the body stationary while routing was a challenge. It turned out that keeping the router in position was not that hard with the help of the jig but, keeping the body from moving during the operation was an issue. The next time I think I will create a fixture that clamps and raises the body into a stable position for the operation.

In order to do this I first had to repeat the process of fitting the back to the sides and notching the kerfed linings to receive the back braces. I followed the same procedures as for the top described in an earlier post. The important thing here was to assure that the body length was 20″ from the neck block to the tail before cutting the notches. Even with the inside mold and the top applied, the sides still flexed a bit so, I made sure that things were aligned to the previous marks that I made on the underside of the back.

Once I had the notches cut for the back, it was time to finally remove the inside mold. I scored an X-shape into the cardboard with a sharp utility knife and pulled out the triangular pieces. This finally gave me a view of how I did on the top glue-up. I guess I did a reasonable job of estimating the amount of glue to apply because there was no squeeze-out on the inside of the guitar body!

One more thing needed to be completed before attaching the back. For reinforcement, strips of spruce were applied to the sides periodically along the perimeter of the guitar. These were simply cut to size to fit between the kerfed linings and glued. The clamping was a bit cumbersome because of the depth of the sides. As I went along on this I created some cauls with some scrap and added tape to the ends so that one clamp could apply pressure to the entire strip. The only thing I found that was problematic during this process was controlling glue squeeze-out . I tended to wait until the glue gelled and then scraped it off however, working around the clamps in a tight area was a bit tedious.

I made sure that the inside of the guitar was clean and then went about gluing and clamping the back. This operation followed much the same procedure as attaching the top. I used the same spool clamps along with a couple of Quick-grip clamps at the neck and tails blocks. In the first picture you can see the glue applied to the neck and tail blocks and the kerfed linings. The second picture shows the clamping masterpiece with the back in place waiting for the glue to dry!

I waited 24 hours for the glue to set, removed the clamps and inspected the joint all along the back. Things looked good so, there was one more operation to complete before I could move on to routing the channels for the binding and purfling. The last thing to do was to remove the little square of wood on the top that covered the neck mortice. I scored inside the edges of the neck mortice with a utility knife and then cut short saw kerfs with a flush-cut saw (just because that is the smallest saw have no hand, not because I was cutting flush with anything). I worked the end of the piece with the knife until I could snap the piece off and then cleaned things up with a sharp chisel.

As I said, routing the channels for the binding and purfling is next. That will be an interesting operation. It will require some jigging in order to do it properly. We’ll cover that in the next post. If you have any questions or comments, please either leave a comment using the comments link at the end of the post or send me an email at thecraftsmanspath@gmail.com.