Walnut Dining Table (2004)

I've always wanted to build a piece of furniture. Since we didn't have one, and since it often represents a place where friends and family come together, I decided to build a dining room table. Below is an account of that process.

Because grain and color matching is important, especially when making a large tabletop, I wanted to take extra care when selecting the wood for this project. To that end, and not at all by choice, I spent the better part of a day at the local hardwood dealer locating enough boards to accommodate the rough plans I drew up. I was sure to grab a few extra boards for good measure.

Once I got the boards home, I laid them out in the room in which the table would be placed. Since it's the most noticeable part of any table, I started arranging the boards that would make up the tabletop. A couple of hours later, I had them arranged in a sequence that I thought was pleasing to the eye. The boards were then marked with chalk so their order could be easily recreated.

I was surprised at how long this process took but I think it was well worth it.

While deciding which boards I wanted to use for the top, I began to put boards aside for the rest of the table and mark them based on their intended use. For example, the grain pattern on certain boards seemed to lend itself to particular table parts. I thought the "eye" in third board from the top would look good on a long apron.

I also marked the "waste" areas of the boards. That's not to say those areas will be discarded, only that they won't be used in this project.

Once all the boards had been labeled, I placed them on stickers in the room in which the table would be located and let them air dry for a couple of weeks. Since the boards were KDS3S* that may have been overkill, but it assured me the boards would evenly adjust to the relative humidity of that room.

As an extra measure of caution, shortly after taking the picture to the left I placed a few hundred pounds across the width of the boards. I hoped that would dissuade any stubborn boards from warping as they acclimated to the room.

*Kiln-dried, surfaced-three-sides (planed on both sides with one edge ripped straight on a saw)

Wood used indoors will eventually stabilize at 8-14% moisture content. With that in mind, I randomly checked a few boards before the milling process to ensure (as much as possible) they would remain stable after being dimensioned.

Since all four sides (but not the ends, yet) of the boards would be surfaced during the initial milling process, I labeled the ends of the boards for the tabletop so I could easily put them back in order for the glue-up.

The first step in dimensioning lumber is to flatten one surface on the jointer. To easily assess my progress, I scribbled chalk marks on the surface to be flattened.

Low spots will retain the marks.

When all the marks are gone, the surface is flat.

The next step is to joint one surface 900 to the one you just flattened.

Be sure the fence of the jointer is set exactly 900 to the bed and keep the board pressed tightly against the fence. A fraction of a degree off and you risk losing silverware in the gaps of your tabletop.

It never hurts to check the accuracy of one tool with another. Here I'm confirming two sides are precisely 900 to each other

With the flattened face of the board down and the 900 edge registered against the fence of the table saw, rip the opposite edge to make the two parallel.

Joint the freshly-sawn edge to remove any imperfections made by the table saw blade.

My plans called for the top to have a finished thickness of 7/8". With that in mind, and to keep waste to a minimum, I looked for 6/4 boards. Unfortunately, the local hardwood dealer didn't have much 6/4 walnut. As a result, most of the wood I purchased was 8/4 or even 12/4. Rather than wasting time and material planing boards down from 12/4, I decided to resaw the boards to a rough thickness of about 1-1/4".

This was one of the 8/4 boards being resawn. After running all of the boards through this process I was left with some great veneer and thin stock which I'm sure will find a home in future projects.

With all of the boards surfaced on three sides and resawn to a rough thickness, it was time to go to the planer.

A quick coat of Carnauba wax on the bed of the planer helps the lumber pass smoothly through the tool and protects the tool from corrosion. I routinely apply it to the jointer beds, table and band saw surfaces, hand planes, etc. You'd be surprised what a difference it makes.

Although the planer could accommodate up to 15" of width at each pass, I sent only one board through at a time. So the center of the planer knives didn't dull faster than the edges, I changed the point of entry every few boards. I would send a few through the center of the planer, then the left side, the right, etc.

It's also important to note that I would send all the boards through at a given thickness setting, then change the setting in increments of 1/32" or 1/16". This method ensures that when you're finished planing, all of your boards will be exactly the same thickness.

You'll recall the plans called for a top with a finished thickness of 7/8". With that in mind, I stopped planing at about 15/16". That gave me about 1/16th of an inch to work with after the top was glued up.

As I mentioned earlier, it's always wise to check the accuracy of one tool with another. My intention was to stop planing at about 15/16". I used the scale on the planer as a rough estimate of the boards' thickness but I used the dial caliper to ensure I didn't go too far.

A quick side note: all of this surfacing generated a ton of chips. I filled and emptied this 35 gallon drum three times while working on the tabletop alone. (For the record, I let this one get way too full before emptying.)

I can't say enough about the utility and convenience of a centralized dust collection (CDC) system. It took an exhausting three-day weekend to install but it was well worth the time and expense.

I'd also suggest getting an air cleaner. The dust collector filters 99.9% of the particles (as small as .2 microns) from the tools to which it's connected and the air cleaner picks up most of the airborne stuff the CDC misses.

Back to business. Once all of the boards for the tabletop were milled to their final width and thickness, it was time to join them together. Because this type of joint is long-grain to long-grain, no structural reinforcements are required. However, to ease the alignment of the mating parts, I cut slots for #20 biscuits.

Because the appearance (grain and color) can change dramatically after a board is surfaced, I took the time to arrange the boards once more before proceeding. When I was comfortable with the layout, pencil marks were made at each location requiring a biscuit.

The slots were cut on the workbench and to keep the area clean, the biscuit joiner was hooked up to a wet/dry vacuum.

The last couple of slots are cut and it's time for some assembly. Finally!

When the boards weren't being milled they were stored inside the house. Perhaps this entire process is fraught with overkill, but I wanted to take every precaution to keep the boards stable. It was to this end (and because I don't have enough shop space) I glued up the tabletop inside.

I gathered the necessary items (wax paper, clamps, plastic sheeting, glue bottles, clamps, acid brush, biscuits, clamps, dead-blow mallet, cauls, clamps and a straight edge) and did a test glue-up. Did I mention the clamps? 

So the mating surfaces of the boards would not be marred by the clamps, I cut some cork pads from an old sanding block and used them during the glue-up.

Rather than try to glue up, align and clamp nine boards before the glue began to set, I thought it best to assemble the top in sections.

Here's the first section. To help keep the top flat, I whipped up a few cauls from some 2x4s I had lying around. I took care to ensure the surfaces contacting the wood and the clamps were dead-flat and parallel. It took all of 15 minutes at the jointer/table saw and made the glue-up go much more smoothly. I wrapped the cauls with wax paper so they wouldn't stick to the tabletop.

The assembly process is as follows: 1) while the boards are on their sides, squeeze glue into the biscuit slots, 2) using a roller, spread glue evenly on the mating surfaces, 3) insert the biscuits into the slots, 4) brush away any squeeze-out from the slots, 5) lay the boards flat and bring together with finger pressure, 6) align marks of the biscuit slots, which often requires a few taps with the dead-blow hammer, 7) apply light clamping pressure to edges - don't forget the cork pads, 8) clamp cauls in place, 9) finish tightening edge clamps and 10) check once more for alignment and flatness and adjust as necessary.

I only placed the cork pads on the surfaces that would be joined. I didn't need to take this precaution on the outermost edges because those surfaces will be removed when the top is cut to final width on the table saw.

A benefit to waiting until after the entire top is glued up to cut it to its final width is that by doing so, you ensure the two sides of the table will be straight and parallel.

A 24" combination square serves as an adequate gauge of flatness. However, I ordered my Starrett 36" straight edge shortly after taking this picture.

This is the second half of the tabletop.

Here, both sections are glued up and clamped in place. Now it's really starting to look like a tabletop.

Interestingly, the glue manufacturer (Titebond) recommends removing the boards from the clamps 30 minutes after applying clamping pressure. I thought the longer you left the boards clamped-up, the better. Apparently that's not the case. In fact, Titebond specifically instructs the consumer not to clamp parts for 24 hours.

I let the top cook in the clamps for about an hour, then removed the clamping pressure and let the top sit overnight. For fear of waking up to a misshapen top, I left the cauls lightly-clamped in place.

With the help of my lovely wife, I carted the unwieldy top back out to the shop. I used a card scraper to remove the dried glue squeeze-out and level the mating surfaces.

Some woodworkers recommend using a damp sponge to clean up the glue squeeze-out shortly after tightening the clamps. This approach, however, can force glue into the pores of the wood and adversely affect your finish. Although it was more work, I chose to remove the glue after it dried.

With the tabletop free of all glue residue and the surfaces leveled, it was time to crosscut the top to its finished length. A 50" tool guide with integrated clamps made this a snap. I used a 24" combination square to clamp the guide square to the length of the table.

I mentioned earlier that I was going to wait until after the glue-up to cut the table to its final width. However, after removing the top from the clamps I measured the width of the table at four or five different locations. All of the measurements were identical and thus, the sides were parallel. With that in mind, I decided to leave the width as-is rather than wrestle a couple hundred pounds of walnut the size of a full sheet of plywood simply to remove an inch or two. Although the top wasn't exactly as wide as I had designed, it was good enough for me. I made a slight adjustment to the plans to accommodate the change in width and no one (except us) will ever be the wiser.

The distance from the trim saw's fence to left side of its blade was 3-11/16". I made a mark where I wanted the end of the table to be, measured back 3-11/16" from there, clamped the tool guide and made the cut.

Since a circular saw blade cuts up through wood, tear-out will be worse on the surface on which the saw rides. I'm cutting with the top up here, but that's not a concern because I'll be milling blind haunch tenons for the breadboard ends. Any tear-out in this step will be removed when I mill the tenons.

The tool guide is relocated to allow the straight-cutting bit in the router to remove a little less than an 1/8th inch from the rough-sawn edge.

Prior to this step, I milled a 1/2" groove in a scrap piece of walnut on the router table. The groove was the same width as the mortises I'd be milling in the breadboards.

I set the depth of cut on the plunge router by making some test passes using one of the cutoff pieces from the earlier cross-cut step. I'd make a couple of passes on one side, flip the test piece and run the other side. Using the dial in the upper-left hand corner of this picture, I made very fine adjustments in this router's cutting depth until I removed just the right amount of material. I wanted the material that remained to be slightly proud of the width of the groove. I'd true it up later with a shoulder plane.

With the depth-of-cut set, the tool guide is moved once more and the router is used to begin milling the tenons. To reduce tear-out and lessen the load on the router, each face was milled in two passes

I wanted the fit to be snug but not so much so that I needed to force the two pieces together. I used a shoulder plane to fine-tune the tenons to fit the groove in the test piece I ran earlier. So I would be able to tune the entire width of the tenon in a single pass, I was sure to cut the length of the tenons to slightly less than the width of my shoulder plane (1-1/4").

To ease the transition between the tabletop and breadboard ends, I used a block plane to chamfer each edge. A sanding block proved a better method for this since the plane, as sharp as it was, caused some tearout.

The dimensions up to this point have been pretty straightforward, even crude. However, when laying out for the mortises and tenons (whose measurements are often referred to as "critical dimensions"), I checked my math with a fractional calculator. Perhaps it's sad, but I'm a little slow at manually calculating half of 3' 11-5/8".

To crosscut the breadboard ends to their final length, the stop at the chop saw outfeed table is set about 1/4" longer than the finished width of the top. The extra 8th inch on each side will hopefully keep the top from extending past the width of the breadboard ends when the top inevitably expands from changes in its moisture content.

The laser guide on the chop saw indicates precisely where the saw kerf will be (and where my hands shouldn't).

I laid out the spacing of the tenons (and thus, the corresponding mortises) on a story stick, then transferred those layout lines to the tabletop and breadboards. Since I wanted the grooves/mortises of the breadboard ends to be concealed, the cross-hatched section is where the tenon shoulder will be removed. This type of joinery is more time consuming than a simpler tongue-and-groove approach, but the result is a cleaner, more refined look.

Here I've placed the same story stick on one of the breadboard ends to locate the mortises.

When it's laid out accurately, the story stick approach saves a lot of time and greatly reduces the chance for errors. Even if you make an error when you layout the stick, your parts should still come together well since you'll be copying that same error on all the parts.

With the layout on the breadboards complete, it was time to move to the router table.

I needed to mill a groove in each breadboard to accept the long, 1/2"-wide, 1/2"-thick tongue of the tabletop.

With that in mind, I set the height of the 1/2"-diameter straight-cutting bit to cut a groove in the breadboards 1/2" deep.

I've highlighted the layout lines in the next several pictures so they would be clearly defined. In this picture, the layout lines represent the location of the mortises.

To conceal the groove, I slowly dropped the breadboard stock onto the spinning router bit, taking care to plunge the bit into the stock where the mortise would be located.

I used a featherboard to keep the stock pressed tightly against the router table's fence.

I slowly ran the piece through until the bit was located within the layout lines of the mortise on the other end of the breadboard. When it reached that point, I shut down the router, let the bit stop spinning and lifted the piece off the table.

DON'T LET GO of the piece until the router bit stops turning. If you do, the vibrations from the router might move the board slightly and the spinning bit can chew up and even throw the board.


Next, I used a hollow-chisel mortiser to cut the mortises into which the tenons would be placed. The center mortise was cut to fit its corresponding tenon snugly. To accommodate seasonal movement, the outer mortises were cut about 1/2" wider than their corresponding tenons.

By stopping the router bit within the layout lines and using a hollow-chisel mortiser to complete the groove/mortises, the ends were cut square. That saved me from having to square them up using a chisel (or having to round the tenons using a rasp).

I could have accomplished these tasks any number of ways, e.g. using jigs and a plunge router or using a stacked dado cutter and chisel work. I chose the router table/mortiser method because the setup is fast and the results are acceptable.


Here's the groove and center mortise in one of the breadboard ends.


Back to the tabletop.

There are a couple of things going on in this picture. 1) I've completed the tenon layout (as I mentioned earlier, the cross-hatched sections will be removed). 2) I'm using a little glue and painters' tape to repair some accidental tear-out that occurred when I was planing the tenon.


I used a jigsaw to remove the bulk of the waste between the tenons...
...a dovetail saw to clean up the corners...
...and a sharp chisel to pare the tenons to fit their corresponding mortises.
Here's a shot of a completed haunched tenon.
Now for the breadboard ends:

The first thing I did was flip the top over (so it's upside down) and temporarily clamped the breadboards in place. These 98" Bessey K Body clamps aren't something I bring out much but they sure are nice to have around at times like this.

With the ends clamped in place, I made a couple registration marks so I'd be sure to have the ends in exactly the same position when it's time for glue-up. If they're off by even a fraction of an inch, the holes for the pins won't line up.
With a cordless drill and a 1/4" Forstner bit I'm ready to make holes for the pins that'll help hold the breadboards in place. I've marked the bit with masking tape because I don't what the hole to go all the way through the breadboard.

Forstner bits are great for this type of application because they make very precise, flat-bottomed holes. I wouldn't recommend using a traditional drill bit for this type of hole.

I've marked the location of the tenon with masking tape and I'm taking great care to stop as soon as the tape on the bit hits the wood.

In retrospect I could've drilled these holes all the way through (using a piece of scrap on the other side to prevent tearout, of course), thus making the pins a part of the table's appeal, not just a structural design element.

Here's the hole in the breadboard ...
... and its corresponding hole in the haunched tenon.
To accommodate seasonal expansion/contraction, the holes in the outer (haunched) tenons are elongated using a coping saw.
The holes are curved slightly toward the table on purpose. This way, as the top expands the breadboard stays nice and snug with the top.
Now for some more assembly:

I have to say I love this part but it's also a little nerve-racking. Make a mistake here and you're in big trouble.

When installing breadboard ends, only put glue on the center tenon. If you glue up the entire breadboard, the wood will likely crack during expansion/contraction.

To hold the pins (1/4" walnut dowels, in this case) in place I put a little glue on the part of the dowel that will contact the outermost part of the hole. This way, I was sure the pins would move freely in the expanded holes of the haunched tenons.
I'd rough-cut the dowel using a Japanese flush-cutting saw, then move on to the next hole.
This is a picture of one of the breadboard ends with all (three) pins glued in place.
When the glue dried I flushed up the pins using, you guessed it, the Japanese flush-cutting saw ...
... then I cleaned up the area with some 220 paper and a piece of scrap wood.
This is a close-up of the area after sanding.
Now to the legs.

I was lucky enough to locate a single, eight-foot long, 16/4 piece of walnut (man was that thing heavy). Since the finished thickness of the legs was about 3", I was able to mill all four legs from this board rather than gluing up multiple thinner boards.

In this picture I'm cutting the board into sections slightly longer than the finished length of the table legs.

In addition to its thickness, the grain pattern was another reason I selected this specific piece of walnut for the table legs. Once the piece is ripped in half, the face of each leg essentially takes on the grain pattern of quartersawn wood. From an aesthetic standpoint, quartersawn is generally considered to be a better choice for table legs than flatsawn wood, for instance.
Here I'm ripping the rough stock in half at the band saw.
After milling four rough-cut legs I went through a similar set of steps to flatten and dimension that stock as I did while preparing the boards for the tabletop. The steps I took were 1) joint one surface flat, 2) joint an adjacent surface 900 to the one you just flattened (and confirm that angle), 3) referencing a jointed face with a combination square, scribe a line at the approximate finished width of the leg (as seen here) ...
... 4) Rip the leg to a rough width (as seen here), then 5) run the leg through the planer (with a jointed face on the planer bed, of course) until each side is surfaced to its final dimension.

Steps 3 and 4 are optional but recommended when the rough dimension is markedly different from the finished dimension. Taking this step saves you time and extends the life of your planer knives.

Here I'm cutting one of the legs to its final length (Figure 1). Using a stop on the outfeed table (Figure 2) ensures all the legs will be exactly the same length.
After cutting the legs to their final length (and choosing the inward-facing sides), it was time to layout the mortises that would receive the tenons of the aprons. I've highlighted the layout lines in this picture so they would be clearly defined. The mortises were positioned to create a slight (3/16") reveal between the legs and aprons.

In the steps to come I'll be setting up the mortiser's depth, linear and flip stops to make these repetitive cuts. With that in mind, I really didn't need to layout all eight mortises. I only needed to layout two: one for each relative orientation. However, I decided to layout all eight so that I could visually confirm they were placed in the right locations and on the correct faces.

This is one of the fore-aft linear stops.
This is one of the left-right linear stops.
With the board firmly clamped in place and registered again the round flip stop, it's time to start cutting the mortises.
To minimize heat build-up, cut the mortises so the chips are thrown back toward the hole they just cut.
With this set-up the process of cutting mortises is incredibly simple. Once the board is clamped in place and the stops are accurately set, you slowly plunge the chisel (and the spinning bit within it) into the wood, then you lift the chisel out, turn the wheel that controls the bed of the mortiser to advance the piece slightly, then repeat until you reach the end of the mortise (which should be determined by the other linear stop).

With this set-up one could literally cut these mortises blindfolded (but please don't try that).

The same mortise (like the one above) was cut in each of the four legs. With a few quick changes to the set-up I was ready to start cutting the second-position mortises.
In this picture I'm cleaning up the bottom of the mortise. It'd be nice if they made a mortise bit that left a flatter hole.
Using some scraps created during the resawing process I made a couple of templates for the legs. I wanted to get an idea how different tapers would look.
With the decision made, I scribed a cut line on all the legs using a long straight edge.
Rough cuts were made on the band saw.
Care was taken not to taper past where the aprons would meet the legs.
The tapers were cleaned up on the jointer ...
... and finally with a No. 4 1/2 hand plane.
The legs were lightly chamfered using a block plane, as seen here ...
... and here.
Now for a rather large "oops":

With the mortises cut and the legs tapered I was all ready for the next step: the aprons. I go to do a dry fit of legs and a sample apron I milled, only to find I had accidentally cut the mortises longer than they should have been. $%@#!!!!

After I cooled off (and that took little while) I started thinking about how I could save the parts I'd spent so much time milling. I determined the easiest solution was to cut a sort of spline to fill the extra space of the mortise.

In this picture I'm rough-cutting one on the table saw ...

... and here I'm fine-tuning it with my shoulder plane.
I made a few kerf cuts at the end of the splines in hopes that glue squeeze-out would be minimized.
Here's one of the splines ready for glue-up.
To ensure the splines were glued at precisely 900 to the legs (and thus, would not adversely affect the leg-to-apron orientation) I used an engineer's square.
 Here's one of the legs with the splines glued in place.
I rough-cut the splines using a dovetail saw ...
... then flushed them up using my No. 4 1/2.
I think it worked!!

When finished, no one will ever know the spline's there (except us).

Now, with the CORRECT length mortise, I began laying out the aprons' tenons. Using my marking gauge, I'd take a measurement from the mortise ...
... and transfer it to the apron.
Again, measure from the mortise ...
... and transfer it to the apron.
And so on,
and so on.
With the faces of the tenons laid out, I needed to mark a line for the shoulders (the tenon depth).

I set my combination square to 1", then used it to set my marking gauge. I chose an inch-deep mortise because that was approximately one third the overall thickness of the leg--a rule of thumb in mortise and tenon joinery.

With the gauge set, I marked the sides ...
... and the ends of the aprons.
Using a wide miter gauge in my table saw (fully-extended), I cut the faces of tenons.

I'd register the apron against the fence and make a pass ...

... then I'd rotate the piece, register it again and make additional passes until all four sides were cut to the right depth.
Since I took measurements from the top of each mortise then transferred them to each apron using the top face as my reference, I wanted to be sure to mark the top accordingly.

The tenons were designed to be centered on the aprons but using (and identifying) a single face as a point of reference lessened the chances of another mistake.

Here I'm measuring for one of the shoulder cuts ...
... and here I'm making the cut on the band saw.
A quick adjustment to the band saw's fence and I'm ready to make another series of cuts.

It's important to note that once you've confirmed a measurement and set a tool up accordingly you should try to make all the cuts that correspond to that measurement before changing the set up. In other words, I'd set up the band saw to make this cut, then cut both ends on all four aprons before adjusting the saw. Except for the forethought that goes into it, this is a faster, more accurate approach to milling wood.

In this picture I've started fine tuning the tenons using a shoulder plane, chisel and sand paper.
I'd dry-fit two pieces ...
... and look for areas where surfaces contacted each other with too much resistance.
I'd clean up those areas (here I'm paring a high spot with my chisel) and before long, all the pieces were fitting together nicely.
With the edges eased slightly using a block plane ...
... the aprons were ready to go.
To reinforce the leg-to-apron joints I milled some support blocks from apron stock.

With the blade of the table saw set to 450 ...

... I cut one end of all the boards ...
... set the stop on the miter gauge and cut the end of one board ...
... ensured the fit was right ...
... and made the cuts on the ends of all the other boards.
Doing it this way ensured all four supports were exactly the same length.
On the tabletop there were a couple really great knots that I didn't want to risk losing as the table aged. To keep them in place I squirted some cyanoacrylate glue into them until they were well-saturated. I let it dry then sanded the surface to remove any high spots. Cyanoacrylate glue (e.g. Superglue) is thin enough to permeate into anything remotely porous. This simple process not only strengthens the knot itself but it bonds the knot to the wood that surrounds it.
Now it's time to prepare the surfaces for the finish. To start, I used a 6" random orbital sander and three grits of sanding discs. To keep dust to a minimum I connected the ROS to a wet/dry vacuum.


When hand sanding was required I used various grits of Norton's 3X paper. It costs a little more but it is FAR superior to traditional sandpaper.


The little 5" ROS saw some action too.
To be on the safe side I wore a dust mask through much of the milling, sanding and finishing processes.
Although you'll see tack cloths in some of the pictures above and below, I found that microfiber towels (often used for car detailing) were fantastic at picking up vast amounts of dust. Compared to a tack cloth (a piece of resin-coated cheese cloth), the microfiber towels pick up more dust, load more slowly and they're washable.
With all the surfaces sanded it was time to start applying finish. To raise the grain of the wood I used a "one-pound cut" of super blonde shellac. I've been very happy with all of Behlen's products. A container of their Super Blonde shellac flakes is pictured here.
To make a one-pound cut of shellac I start by putting one cup of denatured alcohol in a small, glass jar, putting that on a scale and "zeroing it out."
I slowly add shellac flakes until ...
... the scale reads one ounce. So, a one-pound cut of shellac is one cup of alcohol and one ounce of shellac flakes. To make a two- or three-pound cut, double or triple the amount of flakes in a one-pound cut, respectively.
The shellac flakes will slowly start to dissolve in the alcohol.
Every half-hour or so I swirl the solution ...
... until the flakes are completely dissolved.
Before using the shellac, be sure to filter it. Here I'm using the piece of an old dress shirt.
As pure as Behlen's super blonde shellac is, there are still some impurities that you wouldn't want on your finished product. In this picture you can see the wax, etc. that got filtered out.
Safety first:

As with any alcohol-, oil- or solvent-soaked rag, be sure to lay them out flat until they dry completely. If they're left balled up, heat from the evaporating liquid can build up to the point of combustion.

Using an alcohol-soluble dye I make my own custom-dyed shellac. When taking this approach be sure to write down the amounts of shellac and dye so the color can be easily reproduced.
To apply the shellac I used a French-polish-type technique called "padding". To make a pad I cut a decent-sized piece of cheese cloth ...
... folded the corners of the cloth in on themselves ...
... placed a shellac-soaked piece of balled-up cheese cloth in its center ...
... then balled that up in my hand.

These steps result in a pad (often referred to as a "rubber") into which you pour shellac.

Using a small squeeze bottle filled with the shellac solution, I start padding on the finish. With just one pass of super blonde shellac you can see what a difference it makes in the appearance of the wood.
When I'm finished with a pad I place it in an air-tight jar. Doing this allows me to save the pad indefinitely. In fact, I've used some pads for more than a year. Eventually, however, the pad will yield to the wood.

When its time has come, be sure to open it up and lay it flat until it dries. Remember: there's a lot of flammable material in there.

Here I've loaded up a pad with dyed seedlac shellac and I'm applying it to the corner supports. Look at that beautiful walnut! Seedlac is one of, if not the least purified types of shellac. It requires more filtering and often takes longer to dissolve but it has a darker, richer color than super blonde (so less dye is required). Since making this table I've pretty much used super blonde exclusively, dying it to my taste. Frankly, seedlac is too troublesome (by comparison) to make.

Something else of note: when finishing a piece of woodwork I typically have a pad for each cut and color of shellac---and I store each in its own jar.

Here, I'm lightly sanding one of the legs an hour or two after applying a coat of shellac.

That's another benefit of shellac: you can apply multiple coats in a single day. In fact, it dries so fast you really don't have to worry much about dust in the workshop. In that regard it's much easier to work with than, say, varnish.

Shellac has its drawbacks though: it doesn't stand up well to heat or alcohol. To help with that deficiency, I employed lacquer and Danish oil top coats. More on that later.

After tacking-off the dust ...
...  I'm ready for another coat.
You can easily see the dramatic difference the dyed shellac makes.
To "beef up" the finish's resistance to heat and alcohol (and wear in general), I applied brushing lacquer (spraying it can be dangerous and requires equipment I don't own).

I used clear gloss lacquer because the semi-gloss sometimes results in cloudiness. Using gloss not only gets you around that issue but it allows you to choose the exact luster you find most appealing. Want glossy? Brush it and you're done. Want semi? Make a few passes with steel wool. Want flat? Make more passes with steel wool.

I gave the lacquer (and the shellac, for that matter) plenty of time to cure--maybe a couple of weeks each. To add the final, protective coat I brought out the Danish oil. To take the edge off the glossy luster, I used steel wool.

I'd hand rub the Danish oil on the piece, put a little oil on a wad of steel wool and lightly rub until the luster I wanted was achieved. I let that dry for a little while then I buffed the wood with a soft towel.

This picture's a good comparison between an untouched, lacquer-finished board and one that's been rubbed with Danish oil and 0000 steel wool.
Here's one of those knots I mentioned earlier. It really adds a little charm and character to the table.
As I often do, I forgot to brand my piece before starting to apply finish.

While each craftsman's product is already his own, I think adding a brand or a signature really personalizes one's work.

With finish applied to all the pieces it's time for some final assembly.
Those little cork pads I made a while ago come in handy, yet again.
To give myself plenty of time to check the alignment of mating parts I used a slow-setting glue.

Still, I was sure to do a couple dry fits to be sure the glue-up process was well-rehearsed. I've done waaaaay too much work to screw up now!

With the mating pieces marked ...
... I apply an even coat of glue to each piece (one joint at a time) with an acid brush.
I glued up the legs and short aprons first ...
... then I glued the long aprons in place (with the help of my 98" Besseys and my little assistant).

I used a couple of closet doors (that I checked for flatness and leveled with shims) to ensure the legs and aprons were registered against a flat, level surface.

With the glue still damp, I measured from corner to corner. The measurements were identical so I knew the table bottom was square. I tightened the clamps and waited for the glue to dry.
After the glue dried I installed the corner supports. Using a notched piece of 2x4 and some more of those cork pads I held the corner support in place ...
... drilled counter-sunk holes and screwed the support in place.

I suppose I could've drilled the countersinks deeper and made walnut plugs for the holes but since these screws will be hidden by the top I thought I'd leave them uncovered.

To allow for seasonal movement I attached the tabletop to the leg assembly using "figure eights". I'll use a Forstner bit to drill mortises into which the figure eights will be placed. In this picture I've made a hole in a piece of cardboard that I'll use to position the Forstner bit.
Here I've drilled a mortise in one of the aprons.
The depth of the mortise is such that the figure eight will be flush with the apron when it's installed. Rather than trying to mark the Forstner bit with tape, I'd drill a little, check the fit then drill some more if necessary.
With the mortise cut to the right depth, I used the hole created by the point of the Forstner bit to make a pilot hole for the screw that would hold the figure eight in place.
To hold the figure eights in place I put a screw through one hole down into the apron (without over-tightening it), placed the tabletop on the leg assembly (and centered it) then screwed up through the other hole and into the top. There are two of these on each end of the table

Now, as the top expands and contracts the figure eights will move freely, from left to right. This is one of the easiest yet most reliable ways to affix a top to its frame.

And here's the finished product.
In case you're wondering: we bought the chairs!

deldavenport.com © copyright 2007