Sunday, June 29, 2008

Double bevel inlay

Inlay adds lots of time to the oud's construction process. For some of the instruments I've made, I would guess that the inlay aspect alone takes as much time as the rest of the entire instrument. I also won't use an inlay technique that either yields an unsatisfactory result, or that I'm not proficient at. So I'm always looking for ways to speed up the process while maintaining or increasing the quality level.

A few years ago I read an article from an old marquetry text describing the technique known as "double bevel marquetry". This is a well-known technique among marqueterians where the two pieces of inlay material are sandwiched together and the pieces are cut simultaneously at a slight angle. The result are two wedge-shaped pieces of material. The piece from the bottom layer (the wider wedge) raises up into the cavity formed by removing the top piece (the smaller wedge). The kerf (saw cut) made by the blade, combined with the tilt angle create the perfect amount of "slop" between the two pieces and they fit together flawlessly. Technically this is not inlay, the results however are the same. This is a very quick method since it completely eliminates the need to scribe and mortise the background material.

The "inlay" material, in this case, bone, is tack glued to the background material (here, rosewood) only at the ends, so once the cuts are made, the pieces are easily removed.

In this sample the background piece is pushed up to the level of the inlay, and because it is tapered, it wedges in position, flush with the inlay material (the bone). This is where tweaking the angle comes into play. Once the perfect angle is found for the material thickness and blade kerf, the background piece will seat perfectly flush with inlay material, leaving very little levelling necessary once the design in finished.

I attach the pattern with spray adhesive directly to the inlay material. Here I've already cut out the inside portion of this design and pushed the background up to the level of the inlay. The fit is perfect.

This is a view of the backside of the design with both background and inlay pieces removed. The taper can clearly be seen in the small curly part of the cavity.

The background piece is inserted into the cavity, and bottoms out flush with the face of the inlay piece.

Another example.

I use a high-end scroll saw on its slowest setting to do the cutting. I use jeweler's saw blades. I also use a magnifying lamp which greatly improves my ability to cut exactly to the line.

The inlay is cut using the same bevel angle, but the direction of cut is the opposite as the background (interior) pieces, so the instead of the background rising up to fit within the inlay (for the interior pieces) the inlay now drops down to fit within the background.

I always cut out the interior pieces first (in this example, it's just the one fleur shape at bottom right), then fit them into position without glue (the wedging action keeps them in place firmly) while cutting the inlay portion. This keeps the inlay portion strong while cutting.

The paper pattern is removed and the waste pieces are discarded or saved if they are large enough for future use

The background piece (or pieces) is glued into the inlay piece.

Then the inlay piece is glued into the background.

A light press and it will seat flush.

If the angle was right, the inlay will seat almost perfectly flush. Since all the cutting of both layers was done simultaneously, the grain matches perfectly. And if the cutting was done with smooth continuous motion, the fit will be perfect, with no gaps between the inlay and the background. I can even deviate from the pattern if I like during the cutting with no adverse effects to the fit. There is however one gap that must be filled, and that is the hole drilled to insert the saw. With the fine jeweler's blades I use, this is not an issue, since the holes are so tiny. I try to find a small curved corner of the design to drill this hole to further disguise it. The fine curved lines in the leaf portion of this design are simply saw kerfs only, will be filled with glue, and take on a dark appearance once the final piece is installed in the instrument.

The piece is then cut to final shape and inlaid into the instrument in the typical manner. Scribing around and wasting away the mortise with a router or router plane.

The finished piece.

Thursday, June 26, 2008

Bench #3 - A leg vise breakthrough

I like mechanisms that are smooth and effortless to use. And I definitely want my leg vise to work that way. After building brackets to hold my skateboard rollers, it looks like I've achieved that. The results are just what I was hoping for. The vise chop moves in and out in a perfectly straight line all along the length of the parallel guide. The chop is completely supported by the parallel guide and the rollers, not the vise screw. So the screw will turn smoothly since there's nothing pulling it down.

Here is one of the two brackets I made to hold the skateboard rollers. I used the smooth portion of a 5/16" bolt for the roller shaft (I cut the head and threads off). I locked this in position with a couple 10-24 set screws threaded into holes tapped in the flanges of the bracket. The roller is kept from rubbing the inside of the flanges by a couple washers placed between the bearings and the flanges.

The brackets attach to the legs with a couple 3/8" cap screws, tapped directly into the legs.

This is the rear bracket, on top of the parallel guide.

...with the parallel guide in place. I positioned the cap screws so there is about 1/4" vertical adjustment room.

The front bracket.

The same bracket with the parallel guide and chop.

View from underneath.


Here's a video showing how smoothly and easily the vise chop moves.

Next I'll try to devise a similar mechanism for the sliding leg vise. This poses a problem since the sliding leg vise has much less space below the parallel guide.

Wednesday, June 25, 2008

Bench #3 - Leg vise rollers

My prototype rollers weren't the greatest. So as I was leafing through my Grainger and Reid Supply catalogs to get some ideas for better rollers, my older and much wiser brother suggested I use some skateboard wheels. So I headed down to the local skate shop. I told the owner about my vise plan over his billiard cue lathe (he had some cue stick blanks on the counter in some of the nicest bird's-eye maple I've ever seen) then he ducked into the basement and came back with a box of mismatched skateboard parts from probably two decades ago. He tossed me three large urethane wheels and a bag of 4 ball bearings, plus a couple shaft bolts. I wasn't sure if I could use the wheels or not. And he wasn't too confident that the urethane could be turned or sawed. He said the wheels were mine if I wanted them for nothing, plus $5 for the bearings. So I took the parts home and pondered my options. I figured for $5, it was worth a try.

I made a quick fixture for holding the wheel so I could get rid of some of the excess material on the bandsaw. The wheels have a recess for a bearing on each side, so I decided to cut away material right up to the edge of the recess, leaving the final width at 1". I was a bit nervous about cutting the urethane. It's a rubbery-like material, and I wasn't sure if the blade might grab the wheel and fling it across the shop like a retro projectile. So I opted to stand behind the saw and pull the fixture into the blade in case something went awry. The urethane actually cut nicely, making mostly dust, and only a few rubbery clumps now and then. I didn't get any grabbing or galling of the material.

To cut the other side of the wheel to final width I decided it was safe enough to push it through from the front.

Now that I had the wheel at the final width it was time to turn down the diameter a little, to about 2-3/8". I mounted the wheel on a chuck and used a worn out tail center (the point was long gone, it's just a stubby taper) to hold it steady.

Surprisingly the urethane turned fairly easily. I was getting rubbery shavings, but the gouge was cutting nicely, and not catching. I used a square nose scraper (it worked!) to finish it off nice and flat. Gone was the high gloss of the original wheel, but this is fine for my purposes.

The ball bearings fit in nicely, and after I build a couple brackets to house the rollers, I think they should work well.

Tuesday, June 24, 2008

Bench #3 - Tail Vise

As I wrote in my first post about this bench, I'm going to stay away from a typical tail vise design. I like the wagon vise from Schwarz's book very much, but for my tastes its not as robust as it could be. So I'm going to beef up the construction with steel, mostly following the design of David Powell documented in Landis' Workbench book.

Powell's design uses a crossbar on each end of the tracks to keep them parallel. This beef's up the construction, but I'm not so sure it's necessary, especially for keeping the guides parallel since the guides are screwed to the underside of the bench top, and reside in rabbets cut into the top. I don't think the crossbars would do anything to keep the guides parallel that the screws and rabbets aren't already doing.

I've designed my tail vise (above) around a German bench screw. In order to maximize the amount of benchtop remaining above the recess cut for the benchscrew, I've placed the guides (tracks in my drawing) as low as possible (they reside in a 1/4" rabbet ), and I've ground off one corner of the nut. The actual construction of the vise will take place after most of the top is glued up.

Sunday, June 22, 2008

Latest instrument

I posted a small update about this oud a few weeks ago. I finally had a chance to upload a page to the website with some more pictures and a video. Below are a few selected shots from that page, as well as a direct link to the video.

Video Link

Friday, June 20, 2008

Sweet infill

I had a chance to play around with the Ron Brese small smoothing plane recently. I say play around, because I'm not actually building furniture at the moment. I honed the iron up to 8000 grit using a Naniwa Superstone, and grabbed a piece of curly cherry. After a few initial flattening passes, I was able to get full width shavings, well under .001". The shavings are fluffy and light. Think facial tissue, only delaminated.

The iron is bedded at 55 degrees, and at this high angle and shaving thickness, grain direction doesn't much come into play. The surface of the cherry was polished, translucent, lustrous, richly chatoyant and ready for any finish. And all this in the few seconds it takes to make a plane stroke. I haven't picked up a random-orbit sander in months.

Creating a full-length, full-width shaving. About 30" long.

Dense and resinous woods are also no problem for the Brese plane. In fact, this piece of rosewood was left shining by the plane. Even the shavings themselves were polished.

The small knot in this piece was planed without tearout of any sort.

Cherry shaving, about as thin as I can imagine.

Wednesday, June 18, 2008

Bench #3 -- Leg vise action

I want my leg vise to open smoothly when clamping large objects. I will already have to stoop to adjust the pin in the parallel guide, so a smooth action, especially when the the vise is open to its extents is important. I don't want to have to fiddle with a parallel guide that gets stuck.

One idea (see above sketch) was presented to me by Rob of Evenfall Woodworks. His idea was to build a rigid frame around the leg vise chop, thus taking weight off the screw itself and allowing a smoother in-out action. I like Rob's idea, but I wanted to minimize the friction between the parallel guide (and Rob's guide bar) and the leg. The chop is a thick and wide piece of ash, and is quite heavy. One of Rob's ideas was to incorporate a wheel at the bottom of the chop so it can roll along the floor when opening the vise. The chop on my leg vise is quite a ways off the floor, and obviously the sliding leg vise will be too. I thought, why not put the wheel under the parallel guide, and place an opposing wheel above the guide, but behind the leg? The bottom of the vise chop is flush with the bottom of the parallel guide, so the wheel will not interfere with the vise closing completely. And the wheel behind the leg is equally unobtrusive.

To test my idea I made a couple prototype wheel assemblies with some scrap metal I had lying around the shop. I've placed the assemblies so the parallel guide is not touching the mortise in the leg, but simply contacting both wheels only. The weight of the chop keeps the parallel guide in contact with the wheels at all times.

This is the vise in the open position. The screw is not installed.

The outer guide wheel.

The inner, and opposing guide wheel. The action of the chop was surprising. My wheel assemblies were not rolling properly (these are just stop collars on a steel shaft--not true bearings), but when they did, the vise rolled open and closed very smoothly and easily. I pulled and pushed on the chop directly in the area where the screw will pass through, for an accurate representation of how the vise will work. I think this rather crude prototype has proven that this idea is going to work, and work quite well I think, especially after I design some proper roller bearings or wheels into the mechanism. I realize that most of the time the parallel guide won't be moving much, but when it comes time to open up the vise for some thick work, I know that fiddling with a sticky parallel guide and heavy chop will drive me nuts. I think this solution will solve that.