A Restorer's Notebook

A desk was recently in my shop for restoration. This is an English 18th C. slant-front desk. It had some very thorny problems, know in restoration parlance as "inherent vice". In this instance, the design had a serious cross-grain construction issue where the feet are glued onto a rigid molding, glued onto a frame, that's nailed and glued onto the bottom of the case, front, sides and back. The back parts of the frame are two small blocks glued to the bottom to which the back of the back legs are glued. This construction does not allow for the sides of the case to move seasonally without breaking the glue join to the frame. Furthermore, the back parts of the back legs are connected to the blocks glued to the bottom, which is moving seasonally, while the sides of the back legs are glued to the frame, which does not move.

3/19/2023

A recent restoration had me scratching my head. How should I attach a panel to a frame when both items will be shrinking and expanding in opposite directions? Initially the two parts moved together so they were attached with corner blocks. But now, the frame is firmly attached at the top of the case while the panel is well attached to the bottom of the case. What I came up with was a glue block that is in two parts joined with a sliding dovetail. I'm sure I'm not the first to invent such a block but I thought it might be worth passing on. It really took about 10 minutes to make and solved the problem. In the picture you can see that it attaches to the panel on the right and the frame below. Unfortunately these two items move seasonally and in opposite directions! I foresee using these blocks more in the future as they are pretty easy to make and could be mass produced should you need a lot of them. While this application was for a floating panel in a clock, I wouldn't hesitate to use them where tops are attached to a frame. Queen Anne dressing table tops were often simply attached with corner blocks (or worse, nails!) and they always seem to get a crack in them because they aren't allowed to move on the frame.

4/21/2021

When gluing with urea formaldehyde, epoxy or liquid hide glue I sometimes worry about creating starved joints, i.e. when all the glue has been squeezed out of a join and very little is left holding things together. Modern glues such as Titebond and other aliphatic glues have enough solids that make the glue resistant to flow. This helps to prevent all the glue getting squeezed out. Hot hide glue, once it starts gelling, will be highly resistant to flow, even sometimes leaving too thick a glue-line if you don't get clamps on it fast enough. However, liquid hide glue and urea formaldehyde glue have very long open times and may flow out of the join. Similarly, raw epoxy, without fillers, may do the same. Often I wait a few minutes before putting the wood together to allow some thickening to occur.

Another tried-and-true method was to tooth the ground with a toothing plane. This gives the glue a place to "hide out" and stay in the joint. This was trade practice in hammer-veneering where the crafts-person is squeegeeing out the glue as it begins to gel. If you push all the glue out of the join, the veneer will not stick. The toothing insured that some glue would remain in the glue-line. I have found toothing marks on joins other than veneered grounds.

If you find you are using a glue that you worry might run out of the join, a toothing plane might be just the guarantee against a starved join. Pictured are two types of toothing planes. The old wooden plane has a finely toothed blade that scratches the pattern into the wood. The newer plane, a Veritas block plane, has been fitted with a toothing blade that cuts the serrations into the wood which also is very effective for working gnarly woods without tear-out.

1/23/2021

There is a long running debate on how to cut dovetails. Should the pins be cut first and the outlines transferred to the tails or the other way around? As a restorer I have to know how to do it both ways as sometimes I am replacing the sides of a drawer which have the tails and sometimes a back or front which have the pins. But the method that has made the most sense to me and the one I use in furniture making is to cut the tails first and transfer the marks to the pins. One advantage to this method is that you can cut tails on two boards at once as in cutting the left and right sides of the front and back joins of drawer sides or the top and bottom (or left and right) of a box or case.

In debates with my colleagues we often speculate how the old-timers did it. Is there a way to deduce the method by which dovetails were cut by examining the dovetails on drawers and cases? If you saw that the left and right side of drawer sides had identical tails, then you would deduce that they were cut at the same time. By drawing the pattern of dovetails on a piece of transparency paper and then flipping it onto the other board, you can see whether they line up exactly.  I offer that this would be pretty strong proof that the tails on both boards were cut simultaneously.  There is no way that I know of to simultaneously cut the pins. Cutting to a template might give you a similar result but was probably not common trade-craft. However having non-matching dovetails doesn't necessarily mean the pins came first, only that the tails were not cut simultaneously. In the example above, you can easily see that the left and right drawer sides were not gang-cut.

Since I work on many objects that have dovetailed joins, I have taken to examining the dovetail joins and determining whether they match or not and I keep records in my furniture database. To date only 20% or so exhibit gang-cut dovetails. I have only recently begun this experiment so stay tuned. Results are forthcoming.

1/22/2021

 Marking gauges and molding planes have a tendency to wander as our arm motion is not uniform and the grain can pull the blade off the mark. This can be minimized by not trying to make the mark or cut in one pass. Rather, the mark or cut can be made in smaller increments, starting at the end of the cut and working backwards into your previous mark or cut. 

In the case of a marking gauge, by starting the cut close to you and pulling the gauge toward you, then moving a foot farther out and pulling that into the first cut and so on up the board, you will always be starting with the cutter newly aligned to the board. Should you stray, it’ll be for only a small bit and it will end where the previous cut started,  correctly aligned.

Cutting molding or rabbets using any plane with a fence similarly should start at the end of the cut and work backwards. You will always be reregistering the fence on the reference edge and getting a consistent cut and angle.

6/19/2020

 Chisels

As a restorer, the chisel is my most used tool. There is no end to trimming splices and veneer flush with surrounding surfaces. Maybe more than other woodworking trades, I am especially sensitive to less than sharp chisels. What I see in other carpenters' tool boxes points to a very overlooked part of sharpening: the wear on the lapped or bottom side of the blade. If you will imagine what the edge looks like when worn, it will show roughly equal wear on either side of the edge: the side where we grind and sharpen, and the bottom of the chisel. Because I often use the chisel with the bottom being a reference surface, having any kind of rounding there will force you to tip up the handle so the edge will bite.  While this might be acceptable in a plane, this condition in a chisel  makes cutting straight all but impossible. In order to truly sharpen the tool, the wear on both the bevel and the bottom should be gone. This can be done grossly by making the grinding angle quite large so, though it is taking only a small amount of metal, it is quickly grinding the edge back. Less heat will build up that might otherwise ruin the temper of the steel. After the wear is ground back, the cutting angle can be restored by changing the bevel angle, grinding behind the edge, keeping the tip cool, until the desired bevel angle is achieved.

As a note: I think the bevel angle is the least important of the parameters of sharpening. Having a flat bottom surface is first. Having any grind that removes the dullness is next. Until those criteria have been met it is useless to proceed to the last element of sharpening—having the bottom and bevel polished mirror bright. Only in special circumstances are specific bevel angles important, i.e. end-grain, soft wood, erratic or contrary grain.

Flattening the bottom: First, if the bottom of your chisel is convex, that is, the middle bulges out, then flattening is so time consuming that, unless this is a one-of-a-kind tool, it is not worth it. Even a small bump takes forever to remove. I have tried most methods of flattening the bottom and I always go back to carborundum on plate glass. These are powders sold for grinding telescope mirrors and can be found with a little searching on the internet. I inherited my set from a violin maker who, Renaissance man that he was, made a telescope from scratch, grinding his own mirror and using galvanized stove pipe as the tubes! I digress. The plate glass is as flat as you can get and can often be scavenged.  I choose only tools that are flat or have a small concavity on the bottom. To see how much material needs to be removed and where, I begin with a fine grit that will only polish the high spots. This will give you an excellent read on the surface. If it needs a lot of work then I will need to use a coarser grit. What I look for is an even scratch pattern or polish that goes across the whole front edge. Once that is achieved I use another piece of glass and polish the bottom with the next level finer grit. I continue until I get to rouge which will give a mirror finish. It is patient work. Use only a little abrasive at a time. Expect it to take a long time. But then, it is the last time you will have to do this for the life of the tool.

Grinding back the wear: Using the coarsest grit wheel on a tool grinder (a wheel designed for tool steel is best), establish a rather steep bevel and grind until there is no wear on the edge from the bottom. It will go fast because you are only removing a small amount of steel at the tip. Moreover, because it is only a small amount of steel, very little heat will build up. Since the tip has the least ability to dissipate heat, keep it cool, and let it rest between passes. Once the wear is gone, the desired bevel can be established. The grinder will now be starting at the other end of the bevel where there is much more steel—a bigger heat sink. Work carefully until the hollow grind is up to the edge. Keep the steel cool.

Polish the bevel: There are many ways to polish the bevel but this is how I do it. I use a polishing stone 1000 gt or better to polish the bevel. I use the edge and the back of the bevel to establish the angle and polish until the polish goes across the entire edge. Then I polish the same bevel using the finest stone, either my 10,000 grit waterstone or a hard Arkansas. Lastly, I lap the bottom of the chisel, keeping the stone clean and working backwards so I don't run over the wire edge that will come off. You must keep these stones FLAT using carborundum paper on plate glass or a diamond plate. I lap my blades on a hard Arkansas so I'm not having to true up the stone all the time as I would with my waterstones.

Planes

While most of the sharpening rules for chisels apply to plane blades, there are a few more tricks to make sharpening easier. I've always polished the bottoms of plane irons but when one is warped a little or is old and has pitting (from rust) in the steel, a micro-bevel can fix these problems. It also can save hours of flattening and polishing the bottom. Plane blades can have micro-bevels on either side of the cutting edge because they are fixed in the plane body at an angle—the bottom of the plane blade is not used as a reference surface for cutting. A very slight bevel on the flat side can be made using the polishing stone eliminating the need for polishing the entire back of the plane iron. The ruler trick, attributed to David Charlesworth, requires putting a thin metal ruler (or any thin metal) on one side of the stone and placing the edge on the other side of the final polishing stone so the blade is elevated a smidge. The plane blade is passed back and forth until there is a continuous polished edge. This is especially helpful on old plane irons that have been pitted by corrosion where you need to abrade down to get unblemished steel. This method is not useful for a chisel where straight cuts require a flat bottom all the way to the cutting edge.

4/3/2021

It is the perfect word. IN-PAINT: PAINT IN-side the limits of the  repair. Often professionals and amateurs alike try to "blend in" the repairs to the surrounding surfaces, increasing the size of the repair area. However, it is better to match the repair to the surrounding surface using pigment and dye, being careful not to go over into undamaged areas. Since perfect color matching is impossible because pigment and stain do not reflect light the same as wood and do not reflect the same using different light sources, we can only approximate the color and sheen that aged wood and aged finishes have acquired over time using dyes and pigments. It is better to keep our repair area as small as possible so our less than perfect color match is at least as small as it can be. 

This concept can be expanded to include limiting sanding to the repair area only. Sanding around the repair until the  surface patina has be sanded off,  will make the wood go back to its un-oxidized color. For light woods like maple or cherry that have darkened, the sanded area will become lighter. Dark woods like mahogany and rosewood, especially rosewood, which will have acquired a beautiful orange patina, will darken considerably. All could be avoided if the repair is surfaced only within the repair area. Sandpaper may not be the optimal tool. Small scrapers might offer better control.

While replacement of lost material is common, by keeping the repair and subsequent color match confined, the repair will be less noticeable and, even better, less original material is lost.

10/8/2019

The bandsaw maybe the most under-appreciated tool in the modern cabinetmaker's shop. Worse yet, the tablesaw, the staple of those shops, have made their way into the average hobbyist's shop.  Tablesaws excel at cutting sheet goods and wood under 2" in thickness. Great for making kitchen cabinets. They also excel at cutting fingers!* When a tablesaw is your only power saw in the shop, there is a temptation to use it inappropriately, such as for cutting small bits of wood. This is dangerous as the work can bind, get jammed, and get thrown. Personally, I have found a tendency to try control the work (wood) when the saw misbehaves and that is often to inadvertently put your hands closer to the blade. Not good!

The restorer and, I would think, the hobbyist woodworker is less concerned with plywood and can use the bandsaw to rip and crosscut wood safely up to the capacity of the saw. My 14" Walker-Turner can resaw 6 ½" and make 1/16" inlay. It will saw ebony and pine with equal efficiency. And you can safely cut small bits of wood using a zero clearance plate. When the blade breaks, the blade stops and a small part of the blade collapses inside the saw. Better yet, my reaction has been to be startled and move my hands away from the blade! Today's blades have wonderful welds and breakage is almost a thing of the past.

As with any machine, it should be adjusted properly and have a sharp blade installed. Most of the frustration I have found with bandsaws is in using a dull blade and/or not having good tracking. In my restoration studio, this machine does most of the sawing in the shop, with my Japanese ryobas and dozukis doing the rest.

*Thank goodness for SawStop, a tablesaw that retracts the blade when it makes contact with a conductive surface, like human skin! 

9/5/2019