Sunday, November 27, 2011

The Hammer Beam Low Table . . . Creating Corbels and Making Sliding Dovetails

With the greater part of the base complete, it is time to tackle the decorative corbels that adorn each foot of the table.  I began with the idea that a stacked and constructed piece, incorporating a large tenon, would add visual flair and lock the curved beams in place.

I construct each corbel with five small pieces of stock -- a tenon, an upper internal piece, a lower internal piece, and two external cheeks.

The tenon and the internal pieces ready for glue up.
As the individual components are all small, I use "rub joints" and line up the pieces by eye. I start with a sub assembly of the internal pieces and add the external cheeks once these are dry.

Using my initial drawing as a guide, I place an arm on my bandsaw's circle cutting jig and set the it for an 8 3/8" radius cut.  Aligning the drawn piece against the curve, I'm able to hot-glue two stops on the jig to transfer this curve on the glued blocks.  A bit of carpet tape holds the block in place and they pop out of the jig nearly complete.  A quick chamfer on the front edge and a bit of sanding results in a finished corbel.

The end grain was sanded quite easily with the pneumatic sanding attachment on the drill press, and as I'd already checked the fit of the tenon in the base unit, everything came together with just a bit of work with the shoulder plane.

I'm joining the two base assemblies together with a stretcher that attaches with sliding dovetails.  I always enjoy making sliding dovetails.  There is a great deal of instant gratification as the pieces come together in a perfectly fitting joint.  It also helps that I have a method, and a pair of bits, that make it very easy.

The stretcher is 3" wide and will attach dead center at the top of each base assembly.  Because I want this to be a stopped mortise, I will only cut the groove about 2 1/2".  I start by finding center on the mortise stock and begin to make a 3/8" groove on the router table.  I bring the depth to 1/2" in a number of passes.  Experience has taught that if I create this groove, and follow it with a single pass with my Lee Valley 5/8" x 14 degree dovetail bit (#16J1755), set to a depth of 1/2", I get a perfect sliding dovetail mortise without stressing the bit with too much stock removal.  The only challenge with this technique on the router table is that you must make the bit change in a way that allows you to keep the fence in the same place.

The advantage to the router table is that once you have made the pass with the dovetail bit in the mortise, you have automatically set the right height for the bit to cut the tenon.  All adjustments to the width of the tenon will be made with the fence. I like to start by marking a deep mark in the tenon stock with a gauge to help eliminate tear-out.

I adjust the fence, using test stock of the same thickness and with passes on each side, until I have a snug fit.  Patience is key here as you dial this in -- one overly-ambitious cut and you will need to start over.  Using a jig, or your own steady hand, cut the tenons to size.

As these are stopped joints, you will need to remove some stock from the tenon to complete the joint.  I start by sliding the tenon into the mortise until it stops.  I set my gauge to this depth, mark the tenon on the bottom of the stretcher, and remove with excess with a hand saw.

Although this joint will never be seen, I do get some weird kind of satisfaction as it comes together without a hitch.

In order not to bore you (too much!) I did gloss over some of the specifics of how I went about these parts of the project -- but I'd be happy to go into more detail if you have any questions.  I'm also open to being shouted down if you see some glaring error, or know of a better technique.

Next up -- the breadboard top.  Cheers!

Monday, November 21, 2011

Workshop Tips You may Already Know -- Drilling Perfectly Centered Holes

I'm often the last to discover a brilliantly simple method for overcoming some devilish workshop problem. So, in case this little technique isn't in your bag of tricks, I thought I'd pass it along.

Say you have to drill a hole through stock that is one diameter on the entry side and a second diameter on the exit side.  To make matters worse, there is no square reference surface to place against a fence, and you cannot tolerate any tear-out on either face.  This situation presents itself when making an electric guitar with a string-through bridge -- the strings come through the metal bridge, enter a 1/8" hole, pass through the body, and exit a 5/16" hole that contains a ferrule.

There are many ways that you could attempt this, but this method is foolproof.  Start with template with holes that match either the entry or exit hole dimension.  In the case of a guitar, the metal bridge works a treat.  Position the template on the correct side and affix using carpet tape.  Using a drill press, and allowing the bit to find its way into the template hole, drill most of the way through the piece.

The key to this technique is a purpose built jig with a post that matches this first hole.  In this case I use a small length of 1/8" steel rod and allow it to protrude from a piece of MDF by about 1/2".  Now, making sure that you have enough clearance between the bit and the jig (to allow sliding in the workpiece before plunging for the hole), center the bit on the post.  Insert the bit for the second diameter and double-check that the bit is still centered.

Now it is just a case of placing each hole over the post and drilling to the proper depth.  The resulting holes will be centered over the smaller holes, and keep the perfect alignment of the template.  One caveat -- place a thin piece of waste board on top of your stock when removing it from the post.  The snug fit can (will) make you pull hard enough to recoil the piece right into the bit above.

This technique comes from the archives of TDPRI -- a great source of information for building Telecaster-style guitars.  BTW, I'm finding loads of tips from the guitar-building world that have broad application in making custom furniture.  So I'll pass them along from time-to-time.


Tuesday, November 15, 2011

The Hammer Beam Low Table . . . Building the Base

There are many reasons to avoid designing and building simultaneously.  Materials and time may be wasted, and there's no guarantee that the whole thing will hang together -- visually.  But it does give you the maximum degree of flexibility for the greatest amount of time.  After fiddling with the design on paper, I decided to flip the entire piece and and place greater visual weight along the floor.  I wasn't exactly sure, as I started to saw the white oak, how the joinery would come together -- but once underway, it became pretty straight forward.

Given that I'm looking to give the appearance of post and beam construction, I knew that each piece would be fairly thick, and given that I didn't want to invest in 12/4" timber, I knew that I'd be gluing up for thickness.  This gave me the chance to "build" more complex joinery in pieces, simulating large mortise and tenon joints without all the chiseling.

I start by gluing up three pieces to make the vertical post.  Because I want to create a shadow line against the bottom support rail, I sandwich a 3/4" thick piece between two 7/8" pieces.  I take care to keep the grain running in the same direction to facilitate the hand planing that blends them into one visual post.

This support rail is an odd-looking little construction (3/4" thick), that will slide into the post's bottom. creating a bridle joint of sorts.

I then attach a series of 3"x7"x 3/4" "cheeks" that make the rail look to be one solid piece.  As long as every cut is square, you create a perfect 90 degree angle.  As all glue surfaces are long grain-to-long grain, I'm not worried about the strength -- but they will be reinforced later.  Once dry, I'm ready to move on to the curved "supports."

I drew the curves using the method outlined in the contemporary chair project -- attaching flexible bending sticks to a block the width of the leg, and moving it until I had the curve I liked.  It's then into pattern-making mode to cut and refine a template that reproduces that pattern.  I remember to make the template longer on both ends to facilitate the entry of the router bit without encountering end grain.

As I want the final thickness of each support to be approx. 2", I select some chunky 8/4" white oak stock. I Surface, bandsaw, and template each piece, arranging the stock in such a way to avoid routing against the grain.  I do not cut the straight sides with the router, rather I plane to the line on the long straight side, and cut the bottom of the support with a newly-sharpened saw.

A quick pass with a my 4 1/2 plane over this very straight-grained oak results in a glass-like finish.  I've found that adding a high-angle frog to the mix improves almost all of my smoothing jobs,

The support is joined to the post and rail by a joint (that must have a name), that slides along a loose tenon on the post, and is fixed to the rail by means of a dowel.  There will be a corbel that attaches in a very robust fashion that will also secure this piece.

The support on the right will slide along the tenon, cut to provide a long grain-to-long grain strength . . .  

. . . and the dowel pops into place
The next step is to create the corbels that will act as feet on the table, provide visual interest, and secure the supports.


Tuesday, November 8, 2011

Shopmade Cam Clamps -- Revisited and Refined

Last year, I made a pair of cam clamps to secure workpieces in my Hot Rodded Stanley Mitre Box.  They were long on function but short on form, and it set me to thinking about my own personal criteria for the make-or-buy decisions on jigs, tools and shop furniture.

I realized that shop-made tools only end up in my hands if they a) are cheaper than the store-bought version, b) perform better or at least as well as the store-bought version, and c) look better than the store-bought version.  In spite of my workman-like attitude towards tools, I'm surprised to find that their aesthetics are as important to me as the look of my shop or the view out my window.

As I've had my nose in several acoustic guitar-building books, it has become clear to me that I will need cam clamps in quantity if I am to proceed.  I wanted to see if I could come up with a consistent method-of-work, and quantify the key elements that I've found to be critical in building cam clamps that move smoothly and lock securely.  I also wanted to use only materials that were readily available at my local home center or found on the internet.

It seems that jaws that are 7 1/2" long, with each jaw being 1 1/2" wide and 3/4" thick result in a useful size for luthiery work.  For the purposes of this test, I'm making some jaws from hard maple and some from white oak.   I'm placing the oak on hot rolled steel bar (1/8" x 3/4") and the maple on aluminum bar (1/8" x 3/4".)

In general terms, the clamp consists of two jaws that ride on the bar.  A top jaw is fixed to the bar and the bottom slides freely up and down.  Pressure is applied by engaging a cam lever that, because of the placement of its fulcrum, pushes a lower arm out and pinches the workpiece against the top jaw.  I use roll pins to secure the top jaw to the bar, to create a friction fit for the sliding bottom jaw, and as a fulcrum for the cam arm.

While I'm giving exact dimensions for every step, the most critical geometry comes into play around the sliding mechanism of the lower jaw and the placement and pinning of the cam lever.  I'll deal with these in the step-by-step instructions and I hope that my experience (and mistakes!) will take some of the mystery out of creating a really useful tool.

Making the Parts

1. I start by milling all stock to the correct dimensions and I create identical sets of top and bottom jaws.  Using the tenoning jig on the tablesaw, I cut a 1 1/2" kerf on the back end of each jaw set.  As the bar stock is 1/8" thick, it fits nicely in this channel.  I also cut 12" lengths of the bar stock with a pair of bolt cutters (far superior to sawing in my opinion.)

2. Next, a quick trip to the router table (with a 1" straight cutting bit) to place a relief on the inside of all the jaws.  Right away, the oak gives me more trouble, chattering, and requiring three incremental passes in order to avoid tearout. The maple responds well to one deep pass.  I built a cradle to hold each jaw tight to the fence and keep my fingers well away from the bit.

3.  With the relief cut, I then take the lower jaw to the bandsaw to cut the section that will flex under the pressure of the cam.  I leave 3/16" between the relief cut and the cut line and cut back to within 2 7/8" of the back.  I drill a hole at the terminus to prevent the piece from splitting.

4.  Switching to a 5/16" dado blade in the tablesaw, I prepare to cut the curved channel into which the cam lever will ride.  I  flip the lower jaw upside down and set the height of the blade to just brush against the flexible arm, I want to cut the channel 2 1/2" along the top of this lower jaw.  I do this by aligning the overhead fence with the back of the jaw as it just touches the blade.  I then mark on my notched push stick a line 2 1/2" back from the edge.  I now know that when the mark on the back of the push stick meets the overhead fence I've advanced the jaw the proper distance onto the dado blade.  Turn off the saw, wait for the blade to stop and remove the jaw.  Repeat this for all of the lower jaws.  I also extend the distance that the cam contacts the arm by removing some material with a mortise chisel.  This increases the length of throw on the cam arm.

The mortise chisel removes material the cam lever has access to nearly 1" of the flexible arm

5.  I used white oak to create the cam levers.  I start with a blank that is 4" x 13/16" x 1/4".  I use a 13/16" Forstner bit to scribe a circle at one end.  Using a saw, or a file, or a disk sander, remove material until a half circle remains.  I then taper the cam arm to a narrower radius at the back end, but this is primarily for aesthetic reasons.


6.  I start by securing the upper jaw to the bar stock in a little jig that holds the bar at a 90 degree angle.  I snug the bar tightly into the kerf

Now, drill the holes at the drill press and drive in the roll pins.  It is important, particularly with the steel bar stock, to drill at a medium high speed with little pressure, to keep from bending the bit.  This last piece of advice comes from experience.

Attaching the sliding lower jaw to the bar is one of the more crucial tasks in this build.  If it is too tight, it will not move up and down freely;  if it is too loose, it will require a great deal of throw in the cam arm to get enough pressure to secure the workpiece.  Like most tutorials, I use roll pins, placed in a staggered fashion with just enough play in them to allow the jaw to slide.  However, my method is a bit different as I secure one pin, check the fit, and then decide whether the friction of the bar against the end grain of the jaw is sufficient for the clamp to operate.  If it is too loose, I Install the second pin, protruding only slightly into the kerf to dial in a fit.  Here's how I do it.

7. Slide the lower jaw onto the bar and align it so that it is tight to the back of the kerf.  Using a small square, mark the back of the bar on the face of the jaw.

Drill and install a roll pin so that the edge is against this line (see the photo above.)  It is crucial that it is behind the bar and aligned as above.  If the clamp gods are on your side, you can slide this on the bar and it will move with the smallest amount of effort.  Ideally it will hold its place on the bar when you slide it up and down vertically.  If it is too tight, you can remove the jaw from the bar and file or chisel the inside of the kerf until it moves freely.  If it is too loose, don't worry, we'll address that in a minute.

8.  Installing the cam lever in the slot is best done in one operation (match drilling I think this is called.)   It is important that you drill the hole for the cam so that it creates an eccentric circle -- meaning that as you turn the cam lever up, the effective radius becomes larger and it flexes the cam arm toward the opposing jaw.  Here's an example of how it looks when completed.

This is achieved by  placing the cam in the slot, with the front of the cam arm even with the front of the jaw.  Drill through a spot 3/4" down from the top of the jaw and 5/8" from the front of the jaw.  I have my two shop made marking gauges (kebiki) set to these dimensions for quick marking.  Tape it in place and drill on the drill press.

Drive through a roll pin, slide it back on the bar, and give it a test run.  If everything has worked out, when you slide it together, push a bit on the back, and lift the cam lever, it will push the arm out enough to pinch the workpiece between the jaws with a great deal of force.  If it it doesn't, you can trouble-shoot these pretty effectively.

If it fails to clamp:  Either the fit on the bar is too loose or the cam lever is not pushing the arm away from the clamp.  Check the fit and look at how the bar moves. The easiest fix is to add another roll pin, diagonally and on the other side of the bar on the lower jaw. 

The pen points to the second roll pin installed to create a tighter fit on the bar.  My roll pins have a funky look because they are 1" long and then ground down at the grinder.  It sort of "peens" them.
The pin should be visible from inside the kerf, reducing its effective width.  If it protrudes too far, you can file down the pin from inside the kerf.

If the cam lever isn't moving the arm far enough, tap out the roll pin, remove the cam lever, make another, and reinstall, allowing it to protrude from the front of the clamp when you drill.

If cam lever won't lift far enough to hold its position:  It is probably too big or it is making contact with the jaw before it can fully engage.  Remove the cam lever and sand a flat spot on the top of the circle.  While it is out, remove more material with the mortise chisel to allow it to swing freely.

9.  Using carpet tape and whatever leather you have sitting around, fashion and attach pads to the business ends of the clamps.  I found it easier to attach the tape, then a slightly oversized piece of leather, and then trim with a razor blade to fit, using the sides of the jaw as a template.  It's also a good idea to run a file over the edges of the steel or aluminum bar.

The clamps are now complete!

So what's the verdict on the make-or-buy decision?  First off, I used nearly every one of my stationary power tools to make these clamps, so if you don't have a fully-equipped shop, I'd think twice.  Also, it took more time than I thought, but I wanted to get this right, document it and have a zero percent failure rate -- and that I accomplished.

On the plus side, you can't argue with the economics.  I spent approx $2.00 apiece for clamps that would have cost around $15.00 each (with shipping) and I've heard some real gripes about quality on the clamps at this price point.  I made 20 clamps, so that's around $300 worth of tools.  And if you told me I could build my own Lie-Nielsen plane in this time, and improve upon its quality, I'd do it in a heartbeat.  

Oh, as I expected, I much prefer the aluminum bar and the maple over the oak and the steel.  The maple works more easily and the aluminum is cleaner -- but because it is slick the tolerances on the lower jaw must be tighter in order to grip.   Thanks for hanging in there on this long post, and please feel free to contact me if I can answer any questions or if you have any corrections.  


Next up, the Hammer-Beam Tables begin to take shape.


Thursday, November 3, 2011

The Hammer Beam Low Table . . . and Stumbling Through the Design Process

I would never have become a custom furniture builder had it not been for my time living in the UK.  My mind was open, I was surrounded by thousands of years of history, and I was exposed to rich palate of art, design and architecture.  And although there were some disappointments (Stonehenge, Marmite, Stoke-on-Trent), there were several things that just blew me away -- The Devonshire Hunting Tapestries in the V&A, Supermarine Spitfires . . .  and the Hammer-Beams in Westminster Hall.  Built in 1393, this ingenious system of timbers changed the architectural world forever, and made a statement about where England ranked in the medieval world.

This image has been bouncing around my head for awhile, and as I start to apply this style to a piece I'm working on, it made me think a bit about how I go about designing a new piece of furniture.  I'd be interested in hearing how you go about this process as well.

For me, I think I start by asking myself a series of questions -- not always in this order -- and not usually in quite such an organized way. 

What is the function of the piece?  Is it a table, a chair or a desk.  Regardless of how creative I want to be, there are conventions that I feel comfortable working within.  If I stray too far outside the norm, I move from working furniture, to studio furniture, to contemporary art -- and I don't have the design chops to hang with that crowd.  (In this case, it is a low table that can be shipped, flat, and may become part of a "line" of furniture that I debut in September.)

For whom am I making this piece?  Is it a custom one-off, or will I want to replicate it.  Is it built on spec, for a specific individual, or for myself.  This will determine whether I deviate from standard dimensions, keep a build diary, or alter my method of work for replication.  (I hope to build again, and it should be robust and appeal to people who prefer arts and crafts style custom furniture.)

I start with a big piece of paper, a T-square, a ruler, and some idea of what I want to build.  This is the base of the table.

What is the major design element?  Whether it stays within a particular genre, or pushes forward a single visual element, it needs (for me) to have a "plot" to it.  "This is my interpretation of a Morris chair", or "I'm going for a Usonian credenza look," is enough.  I like designs that are fairly singular and can be articulated. (English arts and crafts, Gothic/Gothick, Cotswold's style.)

What kind joinery will be used, and is that joinery part of the design (dovetails) or merely functional?  This generally raises other questions about cross grain situations and mechanical fasteners.  (modified mortise and tenon, both for function and show, and the design provides some efficiencies.)

The bending strip allows me to play with the curve using nails to adjust and hold the bend in place.
 Is this a machine project or a hand tool project?  My work seems to be primarily one or the other (with obvious exceptions.)  Before I start, I want to make sure that I have the proper tools on hand.  (Alas, it is another machine project, but there is some hand work.)

Can I visualize the order of operation?  I always get into trouble if I don't thickness all relevant stock at the same time.  Also, if there is a fussy machine set-up or cobbled together jig, I want to do it once and move on. (It's pretty straight-forward.)

Once I like the bend, I create a 2 3/4" radius circle out of fibreboard that allows me to trace a line 2 3/4" away from the curve to create a design element.

 Do I need to create a drawing (probably), make some templates (usually) or build a mock-up? (not that often)  (Absolutely jigs for the curved bits, and a drawing from which to make templates.)

What wood will I use? Which prompts the question, "How will I finish this piece?"  If the wood is particularly pricey, I may actually opt to do the mock-up. (Oak, with QS Oak or Pippy oak on the top.)

What will be the biggest challenge?  There is usually one part of the gig that wakes you up in the middle of the night.  If I'm in game-shape, and in my stride I may start with this to ensure that I can do it properly -- then everything is downhill sledding.  Or, I may want to refresh those skills on another part of the project and "work myself into shape" before I tackle the line of dovetails or that turned leg.  (Bread board top.  It's not too difficult, just time consuming.)

Is this piece "The One"?  Every artisan that I admire must have made his or her first signature piece without knowing that it would define their career.  Can I see the potential for something important in the midst of the inevitable frustrations of a first build.  (Who knows, but I do think there is a theme upon which I can expand.)

A sneak preview of the bottom portion of the table leg.  This is the only part of the design that I feel is complete.

Cheers! - And let me know about your design process!