|
|
Individually Styled and
Crafted
Fine Hardwood
Furniture
by
Andrew
Pitts ~
FurnitureMaker
Work in Progress
Chest of Drawers 2007 No. 3
Article completed 24 August 2007
|
Editor's
Note: In consideration of those using a dial-up connection, I've
changed the format for faster downloads.
The images are now thumbnails, which may be clicked to open larger
images.
Disclaimer:
This discussion chronicles the making of a piece of furniture in my
workshop. My intention is to bring the reader virtually into my shop to
generally see how I do the work and share in my thought processes.
Although I try to point out how the tools work and the applicable
safety considerations, this discussion is not intended to be a text on
how to work with tools, nor how to execute operations with the tools
shown. I am not providing instruction in woodworking methods.
Woodworkers attempting to imitate my methods do so at their own risk.
Click on images to
enlarge
A few months ago I decided it was time to build a
piece of furniture on speculation, the kind of piece that stretches my
limits and begs the right person to see it and appreciate it (and take it home!). I
have been tending toward curves more and more lately, and suddenly I
had a germ of an idea to make a chest of drawers that would have curves
on all sides. In fact, since I
like to show off my drawers, I thought that perhaps I could make the
chest with no sides at all -- the drawer sides would be the chest sides
and would show off the hand cut dovetails in their
full glory!
So .... I sat
down at the computer and, using my computer aided design (CAD) program,
I started to draw. This, to your right, is what resulted.
Now the coloring of the rendering is not the actual color I expect the
wood to be -- I colored the parts to clearly define them. When I am
making a rendering for a client, I always try to match colors as close
as I can to what I expect the finished piece to be, but in this case I
am the customer and I already know what the colors of the woods will
be, or do I? Hmmm....
I started out
by making the legs. First I drew cardboard templates of the legs, then
traced the shapes to the wood. These legs were made by band sawing
thick red oak planks to the shape of the legs, then laminating a layer
of 1/2 inch cherry to the right and left side of each leg. When the
cherry ages a bit, the contrast between the oak and the cherry will be
very pleasing. Since each of the four legs are different shapes, the
process of shaping and laminating them was somewhat time consuming.
Here you can see two of the legs in a simple jig I used to hold them to
the correct spacing for further construction. After the legs were
roughly shaped, I constructed the drawer dust dividers, made of
sycamore and red cedar. This was my first use of sycamore, which
resembles lacewood in figure. The tree I milled for the sycamore had
been growing for years along U.S Route 360 in Heathsville on the
property of the Heathsville United Methodist Church. That is a story in
itself. The tree was almost dead and we decided it must come down
before a wind storm took it down, either falling on Route 360 or the
church (both bad outcomes), and in either case bringing down the power
lines with it. I decided to mill the tree into lumber, thinking it
would yield some decent pieces. Well, it seems this tree had been hit
by vehicles negotiating the "s" curve through Heathsville (I was told a
cement truck was one of the victims of the tree) and was full of
"shake". That is where the annual rings inside the tree separate,
making the lumber come apart like onions when the milled wood dries.
When the milling and drying was complete, I was the proud recipient of
a short stack of 1-2 ft long pieces of wood, enough to make my dust
dividers and some other small works! I did get a load of firewood, though!
Incidently, the tree must have been a good sign post, since I recovered
enough 20D nails from within to build a small house! So, the sycamore
and cedar made very nice frame and panel dust dividers.
Here it gets
tricky (OK, it's been tricky from the start, but here it gets REALLY
TRICKY). With the right hand pair of legs firmly held in position with
clamps and braces, the dust panels must be fit in place. The challenge
is to take curved legs and fit the panels so the tops and bottoms are
parallel. Complicating matters, each panel is a different size to fit
within the curves of the legs, and
the edges of each panel are beveled to the same angle as the
corresponding legs! I made spacer blocks (some of which are seen in the
photo at right) that are the exact heights of the drawer openings
between the panels, and the blocks helped hold the panels in the
correct orientation while I attached them to the legs. The panels are
attached to the legs with two dowels at each corner. I did not want to
spoil the look by drilling dowel
holes straight through the legs from the
outside (although that would be the easy way to do things), so I
drilled the holes on the panel edges using a horizontal boring setup on
my Shop Smith, then fit dowel points (little metal pieces of dowel with
a point on the end) in the holes on the panel, and placed the panel
alongside the leg in the correct position. Giving a little tap on the
opposite edge of the panel, the dowel points made little depressions on
the legs in the correct locations to drill the dowel holes. The panels
were then dry fit to the right hand legs with dowels in place, and
that's what you see in the photo.
There are two things I forgot to mention. First, why would anyone build
a chest of drawers with no sides? I already mentioned that I wanted to
show off my drawer construction -- I always regret that folks seeing my
pieces in a gallery are reticent to pull out the drawers to inspect
them. I build my drawers as pieces of art in themselves, but they are
seldom appreciated as I intended. (stop your crying, Andy!) But there
is more to it. Think about a factory produced chest of drawers you may
see in a furniture store. Pull out a drawer and look inside, and you will
most likely see a framework of secondary wood (less expensive wood
intended to not be seen) with rough joinery and absolutely no finish.
On the outside of the piece is the more expensive wood,
the primary wood, often a veneered surface. Now, look at a piece of my
studio furniture. You will not find a frame of secondary wood, but
instead the woods used throughout are meant to be seen and appreciated.
The carcase of a chest is the
primary wood, and it is finished
on the inside just like on the outside (go ahead, pull out a drawer of
mine and check inside the drawer pocket, or lie down on your back under
my table and check the bottom -- all primary woods and all finished,
all the time). So ... this particular piece takes the concept to the
extreme. Not only is there no secondary wood frame, but there is no
primary wood carcase, either!. All you see are the primary wood legs,
panels, and drawers, and now you cannot miss seeing the dovetailed
joints on the drawers!
The second point I wanted to mention is why
the dust panels are frame and panel construction. Frame and panel means
the dust panels are made with a frame consisting of a front piece, back
piece, side pieces, and a middle divider piece all joined with open
mortise and tenon joints, and the inside panels fit loosely in a groove
cut on the inside edge of the frame. Why frame and panel, as opposed to
simply using solid wood? The answer is that wood moves in width, across
the grain, seasonally with humidity changes. It actually changes
dimension, sometimes quite radically. However, it does not change
noticeably in length (the reasons for this will be the subject of a
future installment). For example, if the bottom panel of this chest
were solid wood, it might expand across the grain as much as 1/4 inch
from winter to summer, as the relative humidity changes! Imagine the
drawer in that space fitting nicely this summer when the piece was
built, then jamming in place come winter when the dust panel shrinks.
Not good for my image. So I use frame and panel construction, where the
frame pieces are not so wide, and the panels of cedar are made so they
can float in their grooves and change dimension freely without
affecting the overall dimensions of the dust panel. The other
construction option would be to use plywood or veneered medium density
fiberboard (MDF, or furniture board), which are dimensionally stable.
That is how most factory furniture is built. Well, I just like solid
wood better than plywood or veneer, and I like the look of frame and
panel, so that's the sum total of it. The lesson here is that whenever
looking to buy a piece of furniture, make sure there are no wide
expanses of solid wood that, when it moves, will either cause a big
problem like drawer binding or will cause the piece to fail, such as
when large pieces of solid wood are glued together with grains running
perpendicularly to each other (Remember seeing cracked or warped table
tops on antiques? That happened when the apron was glued at a right
angle to a solid top and the top changed in width seasonally but the
apron did not change in length).
Here is the
final
fit up carcase, with all the legs attached to the dust panels, but
still held in place with clamps and braces. The left hand legs were
attached to the dust panels in the same manner described above.
Now that the carcase was basically assembled, it was
knocked apart so I could trim the legs and shape the tops of each leg,
plus smooth the legs, chamfer the edges, sanding the legs and panels,
and basically do all those things that are easier to do with the piece
unassembled. After glue-up all I wanted to have to do was apply the oil
based finish.
Speaking of glue-up, now was the time to glue all
the
dowels to the legs and dust panels. This, unfortunately, was easier
said than done. Glues all have limited working times, so a complicated
glue-up needs to be choreographed like "West Side Story" to be
successful. Since I was using pre-compressed dowels, that is, dowels
that are squeezed into a slightly smaller diameter at the factory, I
knew that as soon as I applied glue to them, they would start to swell
and make assembly all the harder as I tried to slide the dowels into
the now tight holes. My approach was to glue the dowels into the
panels, first,
and then glue the panels to the legs, one pair of legs at a time. I used Franklin Titebond
III glue for this job, kind of a "super Elmer's glue" because it is
easy to work with, is usable
straight from the bottle with no mixing of parts (like epoxies), is
strong, has a longer working time than many other glues of its
class, and is waterproof. OK, so the waterproof part doesn't make any difference in
this project, but it's good to know, isn't it? Titebond stays
plastic forever, meaning it can "creep" over time
and let parts shift a little bit. For this application, creep was not a
problem since there are no stresses that would cause appreciable creep.
But, sometimes creep is a real concern. When I glued up the cherry
strips to the outside of the legs
(remember that, way back in the story?), I used Unibond 800, a
two part urea based glue that sets up strong and stiff, even brittle,
and does
not creep. That way, if the bending stresses in the cherry tried to
straighten the legs out, the Unibond 800 would hold firm and prevent
that from happening. All my bent laminations and veneering use urea
glues for just that reason, to prevent changes in the laminations due
to creep. Getting back to the glue-up at hand, gluing
the dowels into the panels was not stressful, as each dowel simply fit
into a hole with glue and that was that. Gluing the panels to one pair
of legs was a little more stressful, since each panel had to be drawn
home tight up to the legs and clamped before the glue set. However,
this was done one panel at a time and was still pretty easy to do.
Gluing the third leg on was big time stress, though, because now I
needed to align all the panels at once to all the holes in the leg and
draw the leg up to the panels before the glue set. Just the friction
between 14 expanding dowels and their holes required a bar clamp at
each panel level to provide
enough force to bring the leg home. This required fast work with
everything needed (clamps and blocks) prestaged. And the fourth leg was
tougher than the third -- I don't know why; maybe I was just getting
tired by then. Here is what the final piece looked like. Note that the
dust panels are actually structural components holding the legs
together into one "carcase". You can now see how the drawers will fit
into their pockets and actually form the front, back, and sides of this
chest.
Now the carcase is almost complete. It still
needs the top and bottom aprons, filling out the look of the piece. The
process of constructing and installing these was the same for the top
and bottom, so I will use the photo at left as the example, where I am
gluing up the bottom aprons. If something looks wrong, it is just that
the chest is upside down while I work the bottom pieces. Before I get
into
that, though, look closely at the bottom of the top dust divider (which
appears at the bottom of the photo!). Can you see the faint outline of
my oval shaped logo at the very back of the divider? Just in the center
back ... see it? Well, I burn in a logo with a branding iron, then
number and date every piece of furniture I make. This piece is 2007 No.
3, meaning it is the third dated piece of 2007. I always do this in
an out of the way place, usually on the bottom or back of a piece, and
if you look hard, you will find it. I figure that my furniture is made
to last for generations, so in a hundred years or two someone will
discover my "signature" and wonder who this guy was. Not everything I build
gets numbered, though. Only the studio furniture pieces. But I digress.
The apron
pieces not only fill out the look, but they are my structural insurance
to make sure the carcase stays stiff and is rugged. After all, a chest
of drawers can take a beating. The way I made the apron assembly was to
cut the correct angles on the front and back red oak aprons so they
would fit exactly between the legs, then cut the side aprons to fit
exactly between the front and back aprons. Only the front and back
aprons actually attach to the legs with screws -- the side aprons are
doweled into the front and back aprons. Confused, yet? Well, think of
the four apron pieces forming a "picture frame" that gently sets within
the legs and you have the idea. After making up that assembly, I
carefully drilled a hole right through the leg and into the apron
assembly, then used a brass screw to pull the apron assembly up tight
to the leg. The screw is countersunk into the leg, and later I will cut
a wood plug to fill the hole. If I use a plug of the same cherry used
in the leg, you might never see it as it will blend in well. On the
other hand, if I use a contrasting wood, like walnut for the plug, it
might be an accent point. I haven't decided yet how I will plug the
holes, but I saved cutoff scraps from the legs so I could make cherry
plugs if I need to. By the way, I glued the aprons to the dust panel at
the same time to make the assembly extra strong. Did you notice that I
use brass screws? I could also have used stainless steel, but never
plain old mild steel. The tannic acid in the oak will corrode steel,
possibly causing a failure of the fastener in the future. Not good for
my reputation! So, I keep a good supply of brass and stainless steel
screws on hand, since I use oak a lot. I just use those fasteners for
every wood -- keeps it simple.
The chest was
now half complete. The second half of the job was to build the drawers.
They have cherry fronts, pecan backs, and maple sides, and are fully
dovetailed. I picked out the boards several weeks earlier and rough
milled them to near the final dimensions so that they could "rest" a
while and get all their pent up stresses worked out. Kind of like after
a day at the office (or in my case, the ship). Any time you take a
board off the stack and start to cut it up, the pieces will move around
a bit because you just released stresses that were in the board since
it was dried, or in some cases since the board was in the tree in the
forest! I rarely use a piece of wood as soon as I cut it from the board
unless the application is one where a bit of movement is not an issue.
Now, as I said, the drawers all have dovetail joints, and since the
drawers are not at all nice rectangular boxes, but are more like
trapezoids, a router dovetail jig was useless and all the dovetails
were all cut by hand. That is not to say that I did not use my bandsaw
where possible, but as opposed to using a jig with a router where it is
just zip-zip and the joints fit perfectly, hand cut dovetails are laid
out by hand and each cut is made one at a time with saws or chisels.
I'll go through the process here, since cutting dovetails by hand is a
lengthy process and kind of an artform.
The photo to the right shows basic dovetails, consisting of tails (the
board to the left) and pins (the board to the right. Now, imagine
fitting the pins into the tails with the boards at 90 a degree angle
and you get the idea of how dovetails fit together. Cut correctly,
dovetail joints do not need glue to hold the box together, but they are
always glued, anyway, to keep the joints from coming apart over time
and to add stiffness. By the way, I've shown "through" dovetails, where
the tails extend the full thickness of the pin board. For the drawer
fronts, I used "half blind" dovetails, where the tails are not cut the
full thickness of the front, therefore you do not see the ends of them
when you look straight on at the front of the drawer -- more on that
later. Sometimes I like to cut through dovetails on all four corners to
show off the joinery, but in this case you will be able to see the
joinery all the time, anyway (remember, this chest has no sides).
Cutting dovetails
requires the correct tools, and probably the most important is a good
saw. I like to use Japanese pull saws like the Dozuki saw shown. These
are great saws! The teeth are formed so that the blade cuts on the pull
stroke. That means that the blade is under tension as the handle is
pulled, so it tends to draw very straight and cut straight lines.
Additionally, since you are not pushing a sheet of steel, such as the
action in a western type saw, the blade can be much thinner, and
therefor the saw kerf (the wood removed in the cut) can be thinner than
with a western saw. Look how fine the teeth are! This saw enables me to
"split" a scribed line so the dovetails fit with very little additional
paring using chisels. I bet that what I am saying is like a foreign
language for a lot of you, so why don't I go through the drawer making
process with photos so you can see how this all is done.
The first step in cutting
dovetails is to lay them out so you can see where the cut lines are. As
I mentioned, the idea is to scribe very thin lines and then saw them so that the edge of the
blade splits the line. If successful, the joints will fit together
first time with no chisel paring of the pins or tails. Here I am using
a marking knife, which is just a small double edged knife, and small
square to scribe some lines. Some sharp eyed folks may notice that in
this photo I am scribing pins on the front boards. Now, woodworkers
will argue all day whether the pins or the tails should be cut first.
I've done dovetails both ways, but for this project the layout made
more sense to me by cutting the tails first, then transferring their
shape to the pins board. Understand that there is a lot of brain work
going on before the first cut is made. You have to visualize what the
drawer will look like, where the bottom will fit (exactly), how the
dovetails joints will look, and so forth. In this case, I had to do
this while thinking trapezoids instead of rectangles. Ouch! Motrin
headache!
The next step
was to cut the sides of the tails with a saw. Since the tails could
easily be cut on the bandsaw, that was my choice. That was much faster
than cutting with a hand saw in this case, but is not always my choice
of methods. The drawback was that on the bandsaw it was very difficult
to split the scribe line, as bandsaw blades cut quite a wide kerf and
wander a bit. If you want the technical explaination, bandsaw blades
normally have quite a bit of set on the teeth, meaning that the tips of
the teeth are bent to one side then the other to form a kerf wide
enough that the
blade can be steered along a curved line
- lots of wiggle room. That is the strength of the bandsaw, but for
cutting dovetails a little less set on the teeth would help make a straighter
cut. So, I had to make the cut a bit wide of the scribed line and then
pare the tails smooth with a chisel. Since the tails were cut first, it
did not really matter whether they were cut to the exact size scribed,
since the next step was to trace the tails on the pins board prior to
cutting the pins (exactly on the line, this time). In the photo to the
left you can see my great level of concentration making sure that only
the board contacts the blade. The band saw is kind of "sleeper" tool,
as it doesn't make the shriek of the table saw and the blade looks much
more benign than on the table saw. But I believe the bandsaw blade can
remove fingers just as efficiently as the table saw, so I concentrate.
In the photo to the right, you can see how the tails look marked and
then cut with the bandsaw.
After cutting
the sides of the tails, the waste between cuts had to be removed. These
waste pieces were the shape of the pins, and thus were pretty small, so
to remove them I had to chisel them out. It was important to get the
base of the pin "sockets" exactly positioned so the pins would fit
snuggly, so I used a guide board to direct my chisel perpendicular to
the board and exactly along a scribed line. The guide board is a thick
piece of hard maple with sandpaper glued to the bottom to prevent
slippage once clamped in place. I cannot stress enough the importance
of cutting all the pin sockets in a straight line across the board and
to the exact depth of the pins. Error here makes for sloppy fitting
dovetails. Some folks use a big round lignum vitae mallet to drive the
chisel, but I prefer a small hide mallet I inherited from my father. It
just feels right as it provides just the right amount of umph to
control the chisel through the cut.
After the pin
sockets are cut to depth, the sides of the sockets (the sides of the
tails cut on the bandsaw) must be pared smooth. The photo to the left
shows paring of the tails with a chisel, another operation requiring
great concentration. I use two hands to hold both the handle and the blade exactly where I
want it to be, in fact I like to hug the work closely to
steady myself and get better control of the chisel when paring. I am
paring the maple, here, and that is a very hard wood and tough to cut.
I frequently hone the chisel on a very fine Japanese water stone so
that it is razor sharp. To test the blade, I shave a little hair off my
left forearm -- if it cuts hair, it will cut wood! A very
sharp blade cuts with less force, and therefore is more controllable
and is safer to use. Once the tails had been pared, I could transfer
their shape to the end of the pins board using my marking knife. After
the ends of the pins board are marked, I use the square shown a few
paragraphs earlier to scribe lines perpendicular to the end of the
board to indicate the sides of the pins.
Now that the pins were marked, I
used the Dozuki saw to cut their sides. Here the cuts had to exactly
split the scribed lines, or at worst vary a little wide so the pins
could be pared with a chisel. I have found that no matter how
accurately I saw, I always have to at least kiss a few pins with the
chisel to make an exact fit, but the idea is to limit the amount of
chisel work by sawing accurately in the first place. After I cut the
pins with the saw, I had to remove the waste for the
tail sockets. Again, I used the bandsaw to speed this process, since I
can saw out the waste almost to the depth line, which in all cases must
be chiseled exactly. After bandsawing out most of the waste, I clamped
the board to my bench with my sandpaper coated maple block and used my
chisel to cut the depth of the sockets. With all that done, it was time
to test the fit. Dovetails should fit snuggly, with only a little
tapping from my favorite mallet to bring them home. If I kept all my
cuts true and all my corners sharp, the pieces will fit together well
the first time, but invariably there is still a bit of paring to be
done. Now, that's that for the through dovetails on the back joints of
each drawer. Now to the fronts with their half blind dovetails.
The process
for cutting the half blind dovetails was almost the same as cutting the
through dovetails, with a few changes. Since I had already cut the
tails on the side boards, which were made exactly the same front and
back, all I had to do was transfer the shape of the tails to the front
board to mark the pins, just like previously described. It was the
cutting of the half blind pins that was a little different. Since the
pins did not extend the full thickness of the front, only a partial saw
cut was possible. The photo at left shows how I did this, with the saw
at a 45 degree angle to the face of the board. Now,
you might ask how I
completed the cut for each pin side. I did this in kind of a brute
force way. After sawing, I used a thin piece of stiff steel (a thin
cabinet scraper, to be exact)
and using my mallet I drove the corner of
the steel into the sawed kerf to make the cut square to the required
depth. After all the cuts were
made, it was
time to remove waste and chisel the depth of the tail sockets. I could
not use the bandsaw to remove the waste as I did with the through
dovetails, so I used a Forstner bit in the drill press to remove the
bulk of the waste. I did this, because it is faster to only have to
chisel to the final scribe line instead of chiseling out a lot of
waste. Here you can see how I did the drilling. The Forstner bit is a
special drill bit that cuts a flat bottom hole very accurately. After
the waste was removed, it was a fairly simple process to cut the
sockets to the correct depth and then pare the sides of the pins, as
needed for a great fit.
Here is a final dry-fit
drawer. The next step was to construct the drawer bottoms. There are
two ways to do this. Many times the back of a drawer is cut short so
that the drawer bottom can slide into a groove in the sides and front
and be removable. The other way to install the bottom is to completely
capture the bottom in a groove that goes all the way around the inside
of the drawer, including in the back. In this method, the sides, back,
and front are glued up with the bottom in place. This is the method I
chose for this chest, since the backs of the drawers would be visible
(remember, there are no sides nor back to the chest). I decided to cut a 1/4"
groove around the inside of each drawer, 1/2" up from the bottom of the
sides, then fit red cedar panels into the grooves. The cedar
fragrance would make using the chest nice.
The first
task was to select the cedar lumber and mill it to rough size. I wanted
the drawer bottoms to be about 3/8" thick, and I would then rabbet the
edges down to 1/4" to fit in the grooves. A rabbet (not a rabbit) is
simply a "ledge" cut into the edge of a board to step down the
thickness. Getting back to the cedar ... I was fortunate to come into
several cedar logs last year that milled beautifully. I like to use the
wood for the bottoms of boxes and such, since the color is very nice
and it is fragrant. Of course, the days I am working the cedar the
fragrance almost drives me out of the shop, but it's nice nonetheless.
To the right you can see me planing a batch of cedar blanks to get them
uniform in thickness and take out the cup left in the wood after
drying. Cup is when a board curves from edge to edge, and almost all
wood that is dried has some cup due to the shrinkage of the annual
rings.
I took the planed cedar
blanks, which were a little more than an inch thick, and used the
bandsaw to "resaw" them to about 1/2". After resawing them, each
of the 1/2" boards developed more cup as I released the drying tension
in the wood. I then set the resawed boards aside for a day or two to
relax. As thin as these pieces were, they would settle down fairly
quickly. I then selected boards to use for the narrow drawers, but for
the wide drawers I had to edge glue two board together to make a wide
drawer bottom. The advantage to resawing the wood was that I obtained
bookmatched sets of boards that would glue up into nicely figured
bottoms. But, I still had to do the gluing, so first I jointed the
edges of the required boards. Jointing boards makes the edges flat,
square, and true. It can be done on a power tool called the jointer, or
with a hand plane. I used a combination of both, getting the edges
close to perfect on the jointer then making them really perfect with a
hand plane. To get a lasting edge glued joint, there must be perfect
contact between the boards with no gaps. These joints are as strong or
stronger than the wood, itself. Once the boards were jointed, I simply
glued up four pairs to make four wide bottoms. After the glue dried, I
flattened the bottoms using my stationary drum thickness sander, which
is a lot like the thickness planer I used on the cedar blanks, but uses
sandpaper on a wide roller instead of a blade to cut the wood. The
sander is very good for taking out slight imperfections in thin pieces
like these drawer bottoms, plus it smooths them well.
After the bottoms were
thicknessed and smoothed, I trimmed the panels to exact size. I cut my
panels a little less than the inside dimensions of the drawers plus the
width of the rabbets. This allowed a little room for seasonal movement
of the bottoms. In fact, I had calculated that for the widest drawer
bottom I would need less than 1/16" of wiggle room on each side of the
bottom for growth from 30% to 60% relative humidity. That's about the
worst I expect my furniture to see, and probably it will not see even
that. Thank you, air conditioning! I used my radial arm saw, pictured
to the right, to trim the panels, since it can handle the width and
make nice perpendicular cuts. This can also be done well on the table
saw with a crosscut sled, but that is for another chapter. 
After trimming the bottoms to size, I cut
the rabbets along the edges. The rabbets were only 1/16" deep by 3/16"
wide, so I used a straight cutting bit in the router table. A router is a
motor with a chuck in which you can install a cutter bit. The router
turns really fast and the bit just "buzzes" through the wood.
Historically the router was a portable tool, but some years back
somebody got the idea to install one upside down through a hole in a
table so the wood could be moved into the cutter in a controlled
fashion. The router table became a mini-shaper (a shaper is a
substantial stationary machine), and is very convenient to use for
smaller work. To the left you can see the router bit popping up from a
hole in the table, and you can see how the wood is run over the edge of
the bit. The yellow thing is a push block to allow me to push the wood
into the blade without getting my fingers near the spinning bit. Great
idea!
After
preparing the panels, I used the same router table with a 1/4" bit to
cut the groove inside the bottom part of the drawers. Fitting the
bottoms into the groove was a fairly straightforward process,
although
the sloped sides of the drawers made using a small rabbet plane to
adjust the rabbets for a perfect fit necessary. A rabbet plane is a
hand plane with the blade extending all the way to the side so that it
can shave away at the ledge of the rabbet. The bottoms will simply
float in the grooves, without glue, so they can move with the seasons
and never crack or deform the drawers. Here is what the stack of
drawers looked like after dry assembly. Kind of neat, I think. Enough
of admiring my work - I now had to drill holes in the fronts for the
handles and apply several coats of shellac polish to the
insides of the drawers and the bottoms before gluing them up. I had
decided that the handles would 
attach to the drawer fronts using blind
dowels, that is, dowels that do not go all the way through the drawer
fronts. Drilling the holes now, while I could put the drawer fronts
flat on the drill press, would make the job easy and accurate. So, I
did some measuring and eyeballing to determine where the handles would
go to best accentuate the curves of the sides. Then I made a small
template of the hole positioning, which would consist of two holes per
handle separated by two inches, and using a center punch marked where
the holes should be cut. Placing the drawer front on the drill press
and setting the drill depth, I drilled the holes. For this I used a
brad point bit, which is like a normal twist drill bit but ground
differently on the end. It has a sharp point in the center to guide the
bit and a sharp cutting spur on each side to slice through the wood,
creating a clean hole. A clean hole is necessary for a good fitting
dowel.
Drilling the
handle holes completed all the machining on the drawer parts, so it was
time to finish the insides before gluing the dovetails. I won't burden
you with a photo of me hand sanding the drawer parts, but trust me,
there was a lot of sanding to do. Although a plane will smooth wood
nicely, and some makers stop at that point, I like to remove any plane
marks and get my surfaces smoothed to P600 grit sandpaper, which is
very, very fine. Normally I will start my hand sanding at P320 grit
(the higher the number, the finer the grit), still very fine and finer
than a lot of folks use for the final sanding. The P320 cut well enough
to remove the marks from the hand plane and gave me a very smooth
surface. I then used P400 grit to get smoother, followed by the first
coat of shellac polish. What is shellac polish, you ask? Well, it is
very thin shellac, applied with a cotton pad. I mix my own using
shellac flakes. I could make it from store bought shellac in a can, but
mixed shellac has a limited shelf life, so I find it more economical to
mix my own when needed. The shellac flakes have no shelf life. For this
project
I used Ultra Blonde shellac, which is
simply shellac that has been dewaxed and bleached to be a very pale
golden color. I chose ultra blonde because it would hardly darken the maple. Shellac in its
raw form is very dark and would color the maple more than I wanted. By
the way, shellac is a very natural finish. It is the resin left on tree
branches by the "Lac" bug. After the bug leaves its
resin, the branches and twigs are cut off the tree and are boiled. The
resin separates from the wood and the mixture is spread on flat
surfaces for drying. The resulting layer of dried resin is shellac,
which is broken up and sold as flakes. The shellac flakes may also be
bleached and the natural waxes may be removed, and that is the kind I
like to use -- bleached and dewaxed. All I have to do is dissolve the
flakes in denatured alcohol to make what we know of as shellac. My
shellac polish is actually 1/2 pound cut shellac, which means 1/2 pound
of flakes per gallon of alcohol. I actually mix 1 oz. per half pint to
make a jar full. Here you see me in the laboratory mixing my brew.
After mixing the shellac polish, I applied it to each surface of the
drawer insides and let it dry. The photo to the left shows me padding
shellac onto the bottom panels. It only takes about 10-20 minutes to
dry, since it is almost all alcohol. When dry, I sanded again with P400
grit paper, then sanded again to P600 grit. Why 
didn't I sand to P600
before applying the first coat of polish? It is because any time you
apply a finish to wood, there is a potential of raising the grain in
the wood. That means that the outer fibers of wood absorb the liquid
and expand, making the surface feel rougher than before the finish was
applied. So, after the first coat dries and the grain is raised, I
simply sand again with P400 to shave off the fibers and get the surface
smooth again. P600 would be too fine to effectively shave off the
fibers, so I don't waste time with it until after the first coat.
Normally I will repeat this process for six or seven coats to get a
really nice finish that people love to touch. For the drawer insides,
however, three coats were sufficient, and I did not want to obscure the
scent of the cedar drawer bottoms, so they received only one coat just
to bring out the color a bit.
With the
insides finished, I glued up the dovetails. That was an easy job, since
I had already tested the fit dry. Now the hard part was to begin. I had
to exactly fit the drawers into the carcase pockets. For normal
rectangular drawers this consists of simply hand planing the sides
until the drawers just fit into the pockets, but this was not a normal
situation. Since all the drawers had a side that had to fit along a
curve, I had to curve the sides. That meant hogging out some of the
wood from the side of a drawer to approximate the curve of the
adjoining leg, and then fine tune the excavation for a good fit. Also,
since all drawers must be sized to account for some swelling in more
humid climates (my shop was about 40% relative humidity at the time),
there had to be just enough looseness to keep the drawers from binding
later on, but not too loose to allow them to be sloppy. And, did I
mention that all this had to be
done on curved surfaces? Out to the front
deck of the shop and excavation was commenced using an angle grinder
with a 36 grit sanding disk. Now, from the previous lesson on sandpaper
I am sure you have surmised that 36 grit is a LOT rougher 
than P600
grit. In fact, 36 grit is like having boulders glued to the paper
instead of sand! It really cuts fast, so I had to be careful to not
overdo it. Gingerly I carved away material, frequently test fitting to
the carcase, until I had the curve about right, then I
returned to the air conditioned comfort of the workshop to smooth up
the curve with less aggressive methods. I used a random orbital sander
with P80 grit sandpaper to smooth the curve, and finished up with P120
grit, just to get things smooth enough to fit the drawers. Later I
would sand up to P400 for the finishing. Occasionally I found a cabinet
scraper (a "card" scraper that is simply a rectangle of thin steel with
sharp edges that can scrape away wood) helpful to get the contour
right. For the sides of the drawer that were flat (yes, each drawer had
a flat side), I used hand planes to shave then until the drawer fit
snuggly into the carcase.
Normally, I carefully shave my
drawer sides to just fit into the pocket of the carcase. When the fit
is right, the drawer just glides in and out. However, this piece was
tough because there was no
pocket in the carcase, just legs that the drawer sides had to clear
when pushing home. In fact, if the drawer was slightly misaligned going
in it would slam into a back leg and not close all the way. To solve
this problem, I made center drawer guide rails and glued them to the
dust panels. I cut a matching notch in the back bottom of the drawer
and that guided the back of the drawer exactly between the back legs.
The photo to the right shows the back of some drawers, with one pulled
slightly open so you can see how this works. The drawers move very
smoothly.
With the drawers fit, I applied about five or six
coats of shellac polish to the outsides to make them silky smooth. Then
it was time to finish the carcase, which would be oiled. I used a six
part process, three coats of one mixture and another three of the final
mixture. The first mixture was 1/3 tung oil, 1/3 boiled linseed oil,
and 1/3 semi-gloss polyurethane varnish. This mixture soaked into the
wood and the polyurethane set up a slight moisture barrier. The final
three coats were similar, but instead I mixed tung oil and boiled
linseed oil 50/50, then added enough beeswax to make a creamy
consitency. To melt the beeswax in the mixture I let a can of it set in
a pot of hot water fresh off the stovetop. I did this outside, and not
near a flame or heat source (except for the hot water, of course). I
applied the mixtures after sanding the carcase to P400 grit, and after
rubbing in the first coat with my hands I had to do a little resanding
to smooth raised grain, which was minimal. Between coats I roughed up
the surface a bit with a white scotchguard pad, which is a gentle 0000
steel wool equivalent to give a little tooth for the next coat to stick
onto. By the time all six coats were applied, one each day until
complete, the finish was smooth with a little grip from the beeswax.
After at least a week of curing, the surface can be waxed with a gentle
paste wax, such as Briwax.
I did not take any photos of me applying the oil, because my hands were
oily and I did not want to gum up the camera controls, but I think you
can imagine the scene. On the second thought, spare yourself the
details.
The last part
of the process was to shape the drawer pulls. Normally the design of
the pulls comes to me slowly throughout the building process and when
the 
end 
comes, I can make the pulls quickly. This
piece, like all other aspects of it, was different from the norm. While
the design of the chest came to me in a flash one day (or night), I
struggled with the drawer pulls. In fact, after making two complete
sets I was still not happy with the look. The first set was shaped from
maple, since I wanted to contrast the pulls with the cherry drawer
fronts. Additionally, I wanted the pulls to gradually change shape
between the center drawers and the top/bottom drawers to accentuate the
shape of the carcase. Unfortunately, the pulls ended up looking like
big ears on the drawers. And my trusted critic, my wife Kathy told me
they just did not look right. I am not too proud to show you my first
attempt (sniff), in the photo to the right. My second set of
pulls were smaller and were all the same size and shape, and were made
from red oak to reduce the contrast a little. I though they looked
better, but again my trusted critic suggested I rethink. The photo of
those pulls is to the left. So, what to do? How can such a small detail
make or break the piece? In my next installment I will tell you what I
finally decided (right now I am just confused, but I am confident
success is just around the corner). Stay tuned.
After a weekend of looking, thinking, and gnashing of teeth, I came
upon a solution for the drawer pulls. I decided that I liked the
contrast that the maple pulls (the big ears) gave to the front of the
drawers, but I did not like the size. I also liked the red oak pulls,
but not on that particular piece. Those pulls were conventional, and
the chest definately is not! So my solution was to trim down the maple
pulls and reshape them a bit so they were not so ostentatious, and that
worked. Here is a photo of how I shaped
the pulls. You may think that I use a whittling knife, but that is old
hat. After cutting my thumb to a ragged piece of meat, I got smart and
pulled out my oscillating spindle sander. This tool has a variety of
sanding drums that stick up from a hole in a cast iron table and spin
and oscillate up and down at the same time. By carefully holding the
work, I can let the sander shape the pieces (after they have been
roughly cut to shape on the bandsaw). Of course, before any of that
detail work was done, I drilled holes in the back of each handle blank
for the dowels that would attach them to the drawer fronts. Anyway,
after shaping the pulls on the sander, I hand sanded them down to P600
grit and used shellac polish, as described earlier, to bring out the
soft white of the maple. I finished them with two coats of clear
Briwax. They really look and feel nice, with just a hint of wood grain
so typical of maple. And the drawers, despite the odd shapes, slide in
an out of their homes effortlessly. This chest, for all the work (like
223 hours) came out beautifully! I hope that someday soon it will find
a loving home!
holes straight through the legs from the
outside (although that would be the easy way to do things), so I
drilled the holes on the panel edges using a horizontal boring setup on
my Shop Smith, then fit dowel points (little metal pieces of dowel with
a point on the end) in the holes on the panel, and placed the panel
alongside the leg in the correct position. Giving a little tap on the
opposite edge of the panel, the dowel points made little depressions on
the legs in the correct locations to drill the dowel holes. The panels
were then dry fit to the right hand legs with dowels in place, and
that's what you see in the photo. 
III glue for this job, kind of a "super Elmer's glue" because it is
easy to work with, is usable
straight from the bottle with no mixing of parts (like epoxies), is
strong, has a longer working time than many other glues of its
class, and is waterproof. OK, so the waterproof part doesn't make any difference in
this project, but it's good to know, isn't it? Titebond stays
plastic forever, meaning it can "creep" over time
and let parts shift a little bit. For this application, creep was not a
problem since there are no stresses that would cause appreciable creep.
But, sometimes creep is a real concern. When I glued up the cherry
strips to the outside of the legs
(remember that, way back in the story?), I used Unibond 800, a
two part urea based glue that sets up strong and stiff, even brittle,
and does
not creep. That way, if the bending stresses in the cherry tried to
straighten the legs out, the Unibond 800 would hold firm and prevent
that from happening. All my bent laminations and veneering use urea
glues for just that reason, to prevent changes in the laminations due
to creep. Getting back to the glue-up at hand, gluing
the dowels into the panels was not stressful, as each dowel simply fit
into a hole with glue and that was that. Gluing the panels to one pair
of legs was a little more stressful, since each panel had to be drawn
home tight up to the legs and clamped before the glue set. However,
this was done one panel at a time and was still pretty easy to do.
Gluing the third leg on was big time stress, though, because now I
needed to align all the panels at once to all the holes in the leg and
draw the leg up to the panels before the glue set. Just the friction
between 14 expanding dowels and their holes required a bar clamp at
each panel level to provide
enough force to bring the leg home. This required fast work with
everything needed (clamps and blocks) prestaged. And the fourth leg was
tougher than the third -- I don't know why; maybe I was just getting
tired by then. Here is what the final piece looked like. Note that the
dust panels are actually structural components holding the legs
together into one "carcase". You can now see how the drawers will fit
into their pockets and actually form the front, back, and sides of this
chest.
Now the carcase is almost complete. It still
needs the top and bottom aprons, filling out the look of the piece. The
process of constructing and installing these was the same for the top
and bottom, so I will use the photo at left as the example, where I am
gluing up the bottom aprons. If something looks wrong, it is just that
the chest is upside down while I work the bottom pieces. Before I get
into
that, though, look closely at the bottom of the top dust divider (which
appears at the bottom of the photo!). Can you see the faint outline of
my oval shaped logo at the very back of the divider? Just in the center
back ... see it? Well, I burn in a logo with a branding iron, then
number and date every piece of furniture I make. This piece is 2007 No.
3, meaning it is the third dated piece of 2007. I always do this in
an out of the way place, usually on the bottom or back of a piece, and
if you look hard, you will find it. I figure that my furniture is made
to last for generations, so in a hundred years or two someone will
discover my "signature" and wonder who this guy was. Not everything I build
gets numbered, though. Only the studio furniture pieces. But I digress.
blade can be steered along a curved line
- lots of wiggle room. That is the strength of the bandsaw, but for
cutting dovetails a little less set on the teeth would help make a straighter
cut. So, I had to make the cut a bit wide of the scribed line and then
pare the tails smooth with a chisel. Since the tails were cut first, it
did not really matter whether they were cut to the exact size scribed,
since the next step was to trace the tails on the pins board prior to
cutting the pins (exactly on the line, this time). In the photo to the
left you can see my great level of concentration making sure that only
the board contacts the blade. The band saw is kind of "sleeper" tool,
as it doesn't make the shriek of the table saw and the blade looks much
more benign than on the table saw. But I believe the bandsaw blade can
remove fingers just as efficiently as the table saw, so I concentrate.
In the photo to the right, you can see how the tails look marked and
then cut with the bandsaw.
you might ask how I
completed the cut for each pin side. I did this in kind of a brute
force way. After sawing, I used a thin piece of stiff steel (a thin
cabinet scraper, to be exact)
and using my mallet I drove the corner of
the steel into the sawed kerf to make the cut square to the required
depth. After all the cuts were
made, it was
time to remove waste and chisel the depth of the tail sockets. I could
not use the bandsaw to remove the waste as I did with the through
dovetails, so I used a Forstner bit in the drill press to remove the
bulk of the waste. I did this, because it is faster to only have to
chisel to the final scribe line instead of chiseling out a lot of
waste. Here you can see how I did the drilling. The Forstner bit is a
special drill bit that cuts a flat bottom hole very accurately. After
the waste was removed, it was a fairly simple process to cut the
sockets to the correct depth and then pare the sides of the pins, as
needed for a great fit.
Here is a final dry-fit
drawer. The next step was to construct the drawer bottoms. There are
two ways to do this. Many times the back of a drawer is cut short so
that the drawer bottom can slide into a groove in the sides and front
and be removable. The other way to install the bottom is to completely
capture the bottom in a groove that goes all the way around the inside
of the drawer, including in the back. In this method, the sides, back,
and front are glued up with the bottom in place. This is the method I
chose for this chest, since the backs of the drawers would be visible
(remember, there are no sides nor back to the chest). I decided to cut a 1/4"
groove around the inside of each drawer, 1/2" up from the bottom of the
sides, then fit red cedar panels into the grooves. The cedar
fragrance would make using the chest nice.
After trimming the bottoms to size, I cut
the rabbets along the edges. The rabbets were only 1/16" deep by 3/16"
wide, so I used a straight cutting bit in the router table. A router is a
motor with a chuck in which you can install a cutter bit. The router
turns really fast and the bit just "buzzes" through the wood.
Historically the router was a portable tool, but some years back
somebody got the idea to install one upside down through a hole in a
table so the wood could be moved into the cutter in a controlled
fashion. The router table became a mini-shaper (a shaper is a
substantial stationary machine), and is very convenient to use for
smaller work. To the left you can see the router bit popping up from a
hole in the table, and you can see how the wood is run over the edge of
the bit. The yellow thing is a push block to allow me to push the wood
into the blade without getting my fingers near the spinning bit. Great
idea!
attach to the drawer fronts using blind
dowels, that is, dowels that do not go all the way through the drawer
fronts. Drilling the holes now, while I could put the drawer fronts
flat on the drill press, would make the job easy and accurate. So, I
did some measuring and eyeballing to determine where the handles would
go to best accentuate the curves of the sides. Then I made a small
template of the hole positioning, which would consist of two holes per
handle separated by two inches, and using a center punch marked where
the holes should be cut. Placing the drawer front on the drill press
and setting the drill depth, I drilled the holes. For this I used a
brad point bit, which is like a normal twist drill bit but ground
differently on the end. It has a sharp point in the center to guide the
bit and a sharp cutting spur on each side to slice through the wood,
creating a clean hole. A clean hole is necessary for a good fitting
dowel.
I used Ultra Blonde shellac, which is
simply shellac that has been dewaxed and bleached to be a very pale
golden color. I chose ultra blonde because it would hardly darken the maple. Shellac in its
raw form is very dark and would color the maple more than I wanted. By
the way, shellac is a very natural finish. It is the resin left on tree
branches by the "Lac" bug. After the bug leaves its
resin, the branches and twigs are cut off the tree and are boiled. The
resin separates from the wood and the mixture is spread on flat
surfaces for drying. The resulting layer of dried resin is shellac,
which is broken up and sold as flakes. The shellac flakes may also be
bleached and the natural waxes may be removed, and that is the kind I
like to use -- bleached and dewaxed. All I have to do is dissolve the
flakes in denatured alcohol to make what we know of as shellac. My
shellac polish is actually 1/2 pound cut shellac, which means 1/2 pound
of flakes per gallon of alcohol. I actually mix 1 oz. per half pint to
make a jar full. Here you see me in the laboratory mixing my brew.
After mixing the shellac polish, I applied it to each surface of the
drawer insides and let it dry. The photo to the left shows me padding
shellac onto the bottom panels. It only takes about 10-20 minutes to
dry, since it is almost all alcohol. When dry, I sanded again with P400
grit paper, then sanded again to P600 grit. Why 
didn't I sand to P600
before applying the first coat of polish? It is because any time you
apply a finish to wood, there is a potential of raising the grain in
the wood. That means that the outer fibers of wood absorb the liquid
and expand, making the surface feel rougher than before the finish was
applied. So, after the first coat dries and the grain is raised, I
simply sand again with P400 to shave off the fibers and get the surface
smooth again. P600 would be too fine to effectively shave off the
fibers, so I don't waste time with it until after the first coat.
Normally I will repeat this process for six or seven coats to get a
really nice finish that people love to touch. For the drawer insides,
however, three coats were sufficient, and I did not want to obscure the
scent of the cedar drawer bottoms, so they received only one coat just
to bring out the color a bit.
done on curved surfaces? Out to the front
deck of the shop and excavation was commenced using an angle grinder
with a 36 grit sanding disk. Now, from the previous lesson on sandpaper
I am sure you have surmised that 36 grit is a LOT rougher 
than P600
grit. In fact, 36 grit is like having boulders glued to the paper
instead of sand! It really cuts fast, so I had to be careful to not
overdo it. Gingerly I carved away material, frequently test fitting to
the carcase, until I had the curve about right, then I
returned to the air conditioned comfort of the workshop to smooth up
the curve with less aggressive methods. I used a random orbital sander
with P80 grit sandpaper to smooth the curve, and finished up with P120
grit, just to get things smooth enough to fit the drawers. Later I
would sand up to P400 for the finishing. Occasionally I found a cabinet
scraper (a "card" scraper that is simply a rectangle of thin steel with
sharp edges that can scrape away wood) helpful to get the contour
right. For the sides of the drawer that were flat (yes, each drawer had
a flat side), I used hand planes to shave then until the drawer fit
snuggly into the carcase.
end 
comes, I can make the pulls quickly. This
piece, like all other aspects of it, was different from the norm. While
the design of the chest came to me in a flash one day (or night), I
struggled with the drawer pulls. In fact, after making two complete
sets I was still not happy with the look. The first set was shaped from
maple, since I wanted to contrast the pulls with the cherry drawer
fronts. Additionally, I wanted the pulls to gradually change shape
between the center drawers and the top/bottom drawers to accentuate the
shape of the carcase. Unfortunately, the pulls ended up looking like
big ears on the drawers. And my trusted critic, my wife Kathy told me
they just did not look right. I am not too proud to show you my first
attempt (sniff), in the photo to the right. My second set of
pulls were smaller and were all the same size and shape, and were made
from red oak to reduce the contrast a little. I though they looked
better, but again my trusted critic suggested I rethink. The photo of
those pulls is to the left. So, what to do? How can such a small detail
make or break the piece? In my next installment I will tell you what I
finally decided (right now I am just confused, but I am confident
success is just around the corner). Stay tuned.
worked. Here is a photo of how I shaped
the pulls. You may think that I use a whittling knife, but that is old
hat. After cutting my thumb to a ragged piece of meat, I got smart and
pulled out my oscillating spindle sander. This tool has a variety of
sanding drums that stick up from a hole in a cast iron table and spin
and oscillate up and down at the same time. By carefully holding the
work, I can let the sander shape the pieces (after they have been
roughly cut to shape on the bandsaw). Of course, before any of that
detail work was done, I drilled holes in the back of each handle blank
for the dowels that would attach them to the drawer fronts. Anyway,
after shaping the pulls on the sander, I hand sanded them down to P600
grit and used shellac polish, as described earlier, to bring out the
soft white of the maple. I finished them with two coats of clear
Briwax. They really look and feel nice, with just a hint of wood grain
so typical of maple. And the drawers, despite the odd shapes, slide in
an out of their homes effortlessly. This chest, for all the work (like
223 hours) came out beautifully! I hope that someday soon it will find
a loving home!