There are lots of recommendations available for warming up to saw joinery but here I will concentrate on two aspects:

• The progression of the warm-up
• Core muscle activation

The progression

Any good warm-up should include aspects of the main event. To prepare for sawing dovetails, for example, saw to a series of lines that mimic dovetails. As you begin, recall and concentrate on basic technique and mechanics without being primarily concerned about hitting the lines perfectly. You’re like a baseball player before a game, at first taking easy batting practice pitches while just trying to execute sound form and make good contact. Address any neglect of the fundamentals.

Then bear down and try to make a couple of dead-on cuts. Observe the results, sharpen your mind, and clean up your technique accordingly. Find your familiar physical and mental groove.

Make sure there are no deficiencies in your tools and setup, including the lighting. The warm-up also gives you a chance to sense the density and grain of the particular wood at hand and make appropriate adjustments in technique.

For work that you do frequently, the warm-up should be very brief. Even if you’re a bit rusty, it should only take a few minutes, provided your skills are fundamentally sound.

An exercise to engage the core

Only when the core – glutes, hips, upper back – is strong, engaged, and balanced, can the peripheral parts – shoulder, arm, and hands – move with accuracy and precision.

Try this exercise. Make a small, shallow pile of sawdust on your benchtop or scrap of wood. Attempt to create “kerfs” in the pile by pushing the dust with the teeth of your saw without the teeth making contact with the benchtop.

It can only be done with your core muscles engaged, along with a balanced stance.

When sawing joinery with a backsaw, the saw should not be helping to support you. If it is, it is being partly diverted from its primary function, which is to make a kerf, and it won’t be as consistently accurate.

The hand without the saw can rest on the bench or work piece to aid in balance. It should bear the weight of no more than itself and the arm.

By the way, core activation does not mean being stiff. Think of the shock absorbers on a car. They are very strong but allow movement, always maintaining an equilibrium that allows all the other parts of the car to function smoothly and precisely. This discussion is about sawing with a backsaw but even with a handsaw where the entire body moves more, the core is still in primary control of all the motions.

Note to readers: Uncommon tips 1-6 can be found here. More on the way.

Corner blocks (corner braces) are a practical, effective way to strengthen furniture, particularly post and rail assemblies.

As previously discussed here, a properly designed and executed mortise and tenon joint will itself rarely fail but the wood around the mortise still can break. Two feet of leg extending below a table apron can impose huge leverage on the wood in the area of the joint.

Corner blocks are routinely used by chair makers but it surprises me when they are absent in tables where they could have been included. I use them whenever possible – belt and suspenders. Of course, sometimes there is no room for them, such as when a drawer is in the way, and sometimes they would disturb the appearance of the piece. The different mechanical stresses involved in casework make corner blocks generally less useful but there too they are still probably underutilized.

Working with tables

If the table aprons are at 90°, it is easy to make corner blocks with the table saw and miter gauge. The long side of the triangle could be as little as 3″ for a small table and 4 – 5″ or more for larger work. For thickness, 1 – 1 1/4″, using a single row of screws, is usually enough for small to medium work, though for large pieces, 2″ or more with a double row of screws is more in line.

I make corner blocks with a notch to accommodate the inner surfaces of the leg. I think it is best to leave a tiny gap between the block and the leg to avoid possible problems with wood movement that might affect the tenon shoulder line. However, in the past I have not always made a gap there and that has not caused problems.

If the aprons are not at 90° or are curved, I find it easiest to place a blank of wood diagonally spanning the aprons and simply trace the inside edges of the aprons and leg onto it, provided the top of the legs and aprons are flush. Then I bandsaw to the lines and, if necessary, refine the result with a plane. Alternatively, one straight cut, if required, could first be made on the table saw and used as a starting alignment. If the legs extend beyond top of apron, measure out and cut a notch, then place the blank and trace.

Another approach that is sound, though not my preference, is to make the block span the aprons but completely clear the leg. This sacrifices some glue area but should still be sturdy enough and seems easier to make.

Prepare the block by drilling and countersinking clearance holes for the screws. The center hole for the screw that will penetrate the leg is at 90° to the long edge of the block. I make the holes for the screws that will penetrate the aprons at 75°, not 45°, to the long edge of the block so there will be a good bulk of wood around the clearance hole. This also allows a little more screw length to penetrate the apron.

The blocks in the photo above show the features discussed.

I glue the blocks in place with Nexabond 2500M CA glue after sizing the end grain for about one minute, or similarly with quick set epoxy if the sawn surface is at all rough. After the glue sets, I drill pilot holes into the legs and drive those screws. Then I reset the masking tape depth stop and do the same for the apron screws, being extremely careful not to drill too far into the aprons.

If there is room for only a thin corner brace that cannot support a screw, a dovetailed attachment to the top of the aprons could be used. Otherwise, dovetailed attachments are difficult to make and unnecessary.

These are so eminently practical that it seems they should have been around for a century but it has been just several years since Lee Valley started manufacturing Chris Becksvoort’s clever idea. Since then, I’ve been using them whenever there is a need for a substantially long slot to accommodate the movement of a screw caused by dimensional changes in wood related to humidity.

One of the most common uses is for screws that go through cross grain support pieces and secure a table top. Another is at the back of drawer runners that are cross grain to the sides of a case.

The washers come in two sizes, designated #10 and #14, and both are thoughtfully made to convenient dimensions. The #10s shown here are slightly less than 1/2″ wide and 1″ long, with a slot slightly greater than 3/16″ wide. They are 3/64″ thick.

It is possible to rout slots for these but I find it easier to simply drill two 1/2″ holes with their centers 1/2″ apart using a Forstner bit in the drill press. Pare away the remaining web with a 1/2″ chisel. Next, without changing the fence setting, drill 3/16″ through holes on the same two centers. Then drill overlapping holes in between and gradually drill away the waste to form the slot.

The finished slots, and the washer and screw in place are shown below.

For this type of assembly, I prefer square drive, hardened, deep-thread, washer head screws, #8 in this case, available from McFeely’s. (Technically, this is a combo drive head but who in his right mind, given the choice, would use a Phillips driver instead of a square driver.) Of course, the depth of the large slot must be worked out according to the thickness of the stretcher or runner, the thickness of the piece that the screw will bind to, and the length of the screw. The view from the other side is shown below.

True, the same slot construction can be done without the washer and in fact, those that I have so made have functioned well for many years. And there are other good approaches to this issue. However, when sizable dimensional swings must be accounted for, it has always been too careful a setup done with some doubt about the possibility of the screw head binding. Perhaps if it was socked down too tightly in dry wood, I’ve wondered, it might get stuck in the swell of wood around it and not slide.

These washers make things simple and remove any doubts. The screw head will not catch on the metal washer. The construction is clean and sure. Thanks to Chris and Lee Valley for this handy hardware item that should be in routine use.

The glue line of a properly constructed mortise and tenon joint will almost never break from external load. Other things might break but not that.

To get a good sense of this, let’s think about what’s going on in just a small joint with a tenon 3″ wide by 1″ long. There are 6 square inches of glue surface, which is equal to that in an 8″-long edge joint between ¾”-thick boards. Now imagine trying to break that edge joint, not in tension as in hammering down on the unsupported joint line, but in shear! The wood will break, the glue line will not.

The mortise and tenon joint is strong because even in a fairly small joint there is plenty of glue line and it is stressed in shear. So if you’ve fit that decently, don’t worry; it is very unlikely to break. The tenon shoulders transfer much of the stress to that glue line. (By the way, the glue line of a half-lap joint can be stressed in tension if, for example, a frame undergoes severe twisting forces.)

The tenon itself is stressed mostly in tension and compression along the grain, which are also quite strong. So don’t worry there either, because a reasonably sized tenon is also very unlikely to break.

Furthermore, for the purpose of strength, there is no point in fitting the tenon tight to both ends of the mortise. That does not make the joint strong.

If something is going to break, it is most likely to be the wood of the stile or leg, which can succumb to stress in tension across the grain. This is especially so if the joint is designed with injudicious distribution of wood among the components.

Thus, make sure the stile or leg will be strong around the joint. In general, the walls of the mortise ought to be at least as strong as the tenon. And though it seems less popular lately, a haunch is a good idea when joining an apron at the top of a leg.

Also, hygroscopic cycle changes in the wood will stress every mortise and tenon but don’t let this be any more than it must. Don’t let a tenon and excess glue bottom out in the mortise and don’t jam a tenon to each end of the mortise (see above). Placing a peg too far from the shoulder will tend to make hygroscopic movement eventually produce a gap at the shoulder, though placing it too close to the shoulder will make the mortise wall more liable to break.

Inspecting old broken or cracked furniture and other wood structures, wherever you can find them, and thinking about why the failures occurred, is one of the most useful habits a woodworker can have. I’ve been doing this for several decades – I suppose because I would rather not have the same things happen to anything I make.

There are the good reasons for using a diamond stone as a nagura. This is not a novel idea – the intent here is to present a clear rationale for it. However, there’s also a significant practical problem involved.

All of this applies to synthetic finishing waterstones. I think most or all of this probably also applies to Japanese natural finishing stones but I defer to those with more knowledge about those.

The reasons for a diamond nagura:

1. It’s fast. The slurry is raised faster and the surface of the stone is refreshed faster than with any other type of nagura that I have tried. Whatever you perceive to be the benefits of these effects, as discussed in the previous post, they arrive faster with a diamond nagura.

2. The slurry consists of grit solely from the finishing stone; no new grit is added. This removes the uncertainty of introducing another grit, often unknown, from a stone nagura, along with the uncertainty of the amount of it that gets into the slurry based on the relative hardness of the bond in the main stone versus the nagura.

3. It is very capable of crushing the grit in the slurry. I first learned about this several years ago from So Yamoshito, a Japanese tool vendor in Australia and expert on Japanese natural stones. I wrote about it then. The rationale for specifying 1200 grit diamond is that it is fine enough to readily crush the loose fine grit in the slurry yet coarse enough to raise the slurry quickly. The latter effect is apparent.

I can’t directly prove the crushing theory. Furthermore, for it to be of value, the crushed particles would have to retain good cutting ability as finer particles. After working with this for years at the sharpening bench, it does seem borne out by the blade edges it produces.

By the way, what about just using the slurry created by flattening the finishing stone with a coarse diamond stone, say 220 grit? Yes, that’s pretty good but the crushing effect is better with the 1200 diamond. Also, a lot of water is used in flattening and the process tends to swipe the slurry off the finishing stone.

Now for the problem. When you rub a 1200 grit diamond stone, even an Atoma with its surface made of tiny dots of grit clusters, on the wet finishing stone, it sticks like crazy. This is very annoying and then it tends to carry away much of the slurry when you remove it.

I tried using smaller continuous surface diamond stones but they were no better. Then I tried a DMT 6″ x 2″ inch 1200 diamond stone with the “polka dot interrupted surface.” This reduced the sticking but still not well enough. It needed to be smaller.

The little DMT polka-dotted pocket stones were too thin to grip in my fingers. So I hacksawed a 2″ x 2″ section off the 6″ x 2″ stone, which you can see above. It works pretty well. The small size and the perforations eliminate most of the sticking.

Note that I do not consider flattening to be a function of the nagura. In fact, a reasonably evenly-distributed rub of the small diamond nagura should not significantly change the flatness that has already been established well by a coarse diamond flattening plate. I flatten stones at the end of a session when they are fully wet and so they are ready to go for the next use.

The best solution, I believe, would be a 1200 grit diamond nagura, about 2″ x 2″, with narrow channels extending to the edges that would reduce sticking and allow the slurry to flow away from the nagura and remain on the finishing stone. I am working on prototypes using 1″-thick ABS plastic for the base and various applied diamond surfaces. I’m hoping this results in a nagura that is the bee’s knees, but in any case I will report on this soon.

Several reasons are usually given for using a nagura on fine grit waterstones. These include: to raise a slurry, to remove unwanted deposits in natural Japanese stones, to refresh the cutting surface of the stone, and to flatten areas of the stone.

Let’s think about what’s going on when a nagura is used, recalling what we can directly sense at the sharpening bench.

The slurry

When the little nagura stone is rubbed on the finishing stone, a paste, or slurry, is generated. It is sometimes claimed that the slurry actually does the sharpening, but it seems questionable whether loose abrasive particles in the slurry are really cutting steel. There are microscopic photographs of blade edges and stone surfaces, but to my knowledge, no direct visual recording of the actual cutting action at a microscopic level. We can observe the effects but not the actions that produced them.

The thin edge of steel plows most of the slurry but perhaps some loose particles are held by the stone’s surface texture, enabling them to cut. Maybe it burnishes the steel. Maybe it creates a variable grit surface on a synthetic stone somewhat like in a natural stone.

In any case, we can sense that the slurry improves the feel and ride of the blade on the stone and reduces sticking, all helpful effects. So, whatever it is actually doing, the slurry at least feels good.

The next issues are what composes the slurry and what happens to it.

Are there particles of the finishing stone, the nagura stone, or both in there? Particles of the softer (more loosely bound) of the two stones will presumably predominate. This should be considered when the two differ in grit size. For example, a nagura that is softer and coarser than the synthetic waterstone with which it is paired will be probably be counterproductive.

With fine natural Japanese waterstones, nagura selection is an art unto itself. Consult a knowledgeable purveyor of these stones. The nagura also is used to remove defects in natural stones that can damage the blade edge. This function is, of course, not relevant for synthetic stones.

So, what happens to the particles in the slurry? Are they left intact or crushed to some degree? If the nagura could crush loose grit to a finer size, that would seem to be an advantage assuming these crushed particles retained their cutting ability.

The surface of the stone

We can see and feel that a nagura refreshes the surface of the stone by removing metal and glazing. Much like dressing a grinding wheel, cutting particles are better exposed at the surface, ready to cut steel.

As for flattening, there are better ways to do this accurately than with a nagura, though with natural stones a nagura might be helpful for some local flattening as it is used intermittently for its other benefits in the course of sharpening.

A solution

There may be more questions than answers here and you may be thinking that this is all a little bit interesting but enough already. I agree, I’d rather get back to woodworking. However, at least restricting the matter to synthetic waterstones, which most woodworkers use, there is a simple solution to all of this, to be discussed in the next post. The background discussion of this post will support why I think the solution makes so much sense.

Woodworking instruction and practice usually make use of easily worked woods such as poplar or pine. This is practical – it makes learning easier and fosters confidence.

However, when moving on to more cantankerous woods, the techniques may not be fully applicable. Not only quantitative changes but also qualitative alterations in technique may be necessary. This may surprise and confound the learning woodworker and, as I often say, that includes all of us.

For example, the adjustment in cutting dovetails in red oak after practice in poplar is not just that you have to swing the mallet harder. The tolerances for sawing and fitting that work for the more compressible poplar won’t produce good results in oak. Chopping to the baseline is also different in oak. It helps more to clear the bulk of the waste with a coping saw, yet once done, there is actually less tendency for the chisel to push back beyond the baseline when chopping if it is done in appropriate increments.

The point is that however you like to do it, it pays to reconsider techniques based on the wood at hand.

Hand tool enthusiasts seem to like chopping mortises with a chisel and making tapered sliding dovetails entirely by hand. Fine in pine, poplar, mahogany, and so forth, but how about bubinga? Similarly, I like to hand plane to the final surface whenever practical but for blister maple, hey, it’s time to reach for scrapers or the random orbit sander.

Likewise, someone working almost exclusively in mahogany will surely have accommodated his techniques to that wood and the design style in which he works. That’s good, but it’s not likely that you can transfer all of those techniques and habits to a substantially different wood or style, and certainly you cannot do so unthinkingly.

Woodworkers work in wood, and wood is a very diverse product of nature. We’ve got all sorts of tools – planes with different angles, saws with different teeth, machines with different cutters, and so on. As for anyone good at any skill, a good woodworker ought to have a range of techniques to thoughtfully employ as needed when building in different woods. Further, it pays to be open to expanding that range when encountering unfamiliar woods.

A good fit of the tenon to the mortise can be described as a comfortable swish fit. You should not need to pound the joint together, nor should the tenon simply drop into the mortise, nor should the tenon wobble in the mortise.

This matters because a mortise and tenon joint derives its strength from the restraining effect of the shoulders transferring stress to the bond between the tenon cheeks and the walls of the mortise. Against this shear stress, a well-made bond has great resistance, stronger than the wood itself.

To create a square assembly, M&T joints must also be true – the tenon cheeks should be in planes parallel to the reference face of the rail. Also, ideally, the tenon should have a good fit over the entire surface area of the cheeks, though perfection is not necessary because the glue does permit some leeway.

Whether the tenon is made by hand or machine, it is very helpful to have reliable ways to adjust the fit using hand tools. Test the corners of tenon into the mortise and feel for tight areas, then check the cheeks for burnishing. Look for bumps and steps from inaccurate sawing.

A rabbet block plane is one of my two favorite tools for trimming tenon cheeks. It starts easily, works right up to the shoulder, and automatically makes a flat surface. Pictured next to it is Lie-Nielsen’s big 1 1/4″ shoulder plane, which also can be used. Though it’s a great plane for other tasks, it and other shoulder planes are a bit tippy for this work.

My other favorite is the Iwasaki 10″ coarse float. One might expect a file or rasp to round over the surface but this tool has such a decisive bite it can be controlled very well. What’s more, it leaves an incredibly clean surface, without tearing, for a tool with such big teeth. The safe edges prevent damage to the shoulder. I use three fingers on top of the tool for feel and control. I like this tool a lot.

A wide paring chisel is another good option, especially for localized clean up. The length of a paring chisel offers considerably more control than a bench chisel. Pressure with left hand fingers on top coordinates with the right hand, which transmits depth of cut via the handle. Still, for a thin shave down of the whole tenon, the rabbet plane works better.

Yet another option is the router plane. I only use this if I think I’ve messed up the trueness of the tenon and need to establish a cheek into a plane parallel with the face of the rail.

Press the sole of the plane onto the face of the rail and start by setting a light cut at the most prominent part of the cheek, then work down from there. Mostly swing the plane, pivoting on the rail face, more than push it, to maintain steady contact and thus depth control. The tool is acting as a gauge to make the cheek face parallel to the rail face.

I generally hand saw tenons because setting up machines is usually not worth it for me for one-of-a-kind pieces. Even with several good options for fine tuning the tenon cheeks, I strive for a good fit directly from the saw, maintaining a one-sided tolerance to avoid having to patch up a tenon.