Batchin’ ’em Out

If you’ve been here more than once, you know I’m a hand tool guy. To be clear, I do have power tools that complement primarily hand tool work. My lunchbox thickness planer does the donkey work once there is a tried and true reference face and a squared reference edge off the jointer plane. A double bevel compound miter saw quickly cuts stock to rough length. And you can pry my benchtop hollow chisel mortiser from my cold, dead hands. I also have a small drill press (that at this point is used exclusively for accurate drawboring), a collection of battery-powered DIY tools (a drill driver, a circular saw and a random orbital sander), and a trim router kit for when I’ve truly given up on things.

But doing hand work efficiently is more than just leveraging power when, as and if it makes sense. When there are multiple parts to cut (there always are), it helps to think like a one-person assembly line. Each step in the assembly line is a repeated task. Sure, variety is a the spice of life. But just like a blade and fence setup at the table saw, you want to set it once, do all the cuts, then move on. It’s the same thing for body mechanics at the workbench.

So let’s talk about tapered octagon legs by hand.

The first two tapers on each piece (on opposite sides) go pretty quickly in the face vise. If you work to opposite sides, the other profile is still square and therefore easily held in the vise (in my case, a leg vise). Do that eight times.

If you have a twin screw vise that can hold tapers securely, great. Stay at the face vise. But I don’t, so I move to the tail vise. Pinched between the dogs, the legs sit flush to the bench on the tapers I had just planed to make a square taper on all faces. Do that eight times.

And then you have this.

Now lay out the octagon(s). If you have a lathe and will taper across the entire length, you’re nearly done at the workbench. But I don’t have a lathe and I like to start the taper where the round tenon ends, so in addition to the octagon at the foot, I also lay out an octagon on the top where the tenon will go. A cradle jig that goes in the tail vise holds the work and I taper from square down to octagon at the foot and also from square down to an octagon at the top. Do that thirty-two times (16 long tapers and 16 short tapers).

And you end up with something like this.

That short taper makes it easier to center the round tenon cutter I have for my drill driver, btw. I use a 1 1/2″ tenon cutter, but that’s just a rough cut. With chisel, spokeshave and rasp I take that round tenon down to 1 3/8″ to ensure it’s centered on the blank (it rarely is straight off the tenon cutter). It also helps to bore a 1 3/8″ hole in some hardwood (or at least wood that is harder than your blank) with the bit you’ll use for the mortise to test the fit now and again. Do this four times.

Almost done now! Yes, that’s a Mets colored Nalgene. #LFGM

Finally, I go back to the corners (where they were tapered from square to octagon) and plane in the full tapered octagon from the tenon to the foot. I find taking the facets down evenly first (so the facet is parallel from the tenon to the foot), and then incrementally increasing the facet width at the top near the tenon by counting strokes, works best. Again, if you taper the full length, this step is unnecessary.

The finished leg.

It goes without saying, but I did one leg first to work out the process and then batched out the other three with the process described above. Are they perfect? Of course not. But we are not machines (and should not strive to be machines). And I enjoy the hand made aesthetic far more than machine-wrought perfection.

Okay, I lied. I made a second one to test the process. I’m actually at step 2 for the other two legs.

There is a great rhythm one can get into when batching out parts at the bench. Hehe, batching.


Headlight Fluid

Working with unprepared stock is usually no fun. I prefer to work from rough lumber purchased from the yard (because it’s both more economical than pre-surfaced stock and there is more variety available). But for things where 3/4″ white pine will suffice, it’s hard to beat the home center common boards. Spend an hour digging through the pile and you’re bound to find clear, straight grain stock. And if you don’t, come back in a week and try again.

But even the clearest, straightest grain stock tends to cup and twist once it’s been in the shop for a week. Re-flattening can take it down to 5/8″ thickness or even less (depending how severe the cupping is), and it adds time and aggravation to the process.

But what if cupping and twist didn’t matter? If you’re just gluing and nailing something together, couldn’t you instead flatten the boards with some clamping cauls and go to town? The answer is, of course, “yes”. And that’s the approach I took when knocking together a quick Japanese-style toolbox for a recent trip out of town (don’t worry; I quarantined both before and after).

Like so?

We modern woodworkers get hung up on perfection, sometimes. To us, everything needs to be piston fit and machinist flat, even when it really doesn’t matter. But, in reality, only some surfaces need to be perfect. Our forebears knew this (just look at the hidden surfaces in any period masterwork). If the wood is going to restrained, it doesn’t really need to be perfect in the first place.

And so, once glued and nailed, and then reinforced with floorboards and gussets, also nailed on (and maybe even glued), any wayboard boards on this toolbox will be restrained from future movement. Might it blow apart in the future from some unforeseen stress? Maybe. But probably not.

Thing thing is solid as a rock.

And, in any event, if it does, it’s just some home center common grade white pine 1×12’s joined with glue and nails. I can always knock it apart and add some dados to keep things more in alignment.


Keeping it Together

Fair warning: this post is actually a “how to” on a method for reinforcing a joint where you’re joining two boards at a right angle by screwing through the face grain of one board into the end-grain of the mating board.  If that interests you, please proceed.

I recently built a “The Naked Woodworker” workbench, partly for the intellectual exercise of it and partly because my brother needed a workbench for his recently-expanded garage.  I have mostly good things to say about the design and the ease of construction.  I was able to put the entire bench together in less than 12 hours time of shop time stretched over two days (one long, one short).  Had I let the wood acclimate a bit more before construction, I bet could have done the whole thing in a single day.


This was about 6 total shop hours in.

The bench is mostly glued and screwed together, but there are two joints where boards are joined at right angles with just screws through the face grain of one board into the end grain of the mating board.  One such place is the top rails of the leg assemblies (seen above).  The other is the number of bearers stretched between the aprons to which the top is eventually screwed down.

Screws into end grain, especially late growth softwood, is not the strongest joint.  In an abundance of caution, I sized all the end grain and glued it as best I could.  But it was still a bit shaky in places.  So when using up the last bits of construction lumber to make a shop fixture, there were a couple of places where screws into end-grain just wouldn’t cut it.  Instead, I utilized a 3/4″ oak dowel like a bench bolt to give something for the screw to bite into.


I’ve been using my 18 gauge brad nailer more, these days.

Please note, I cannot take credit for this technique.  I learned it from a Popular Woodworking video on making a quick and dirty first workbench.  It shows up in the first half of the linked video.

First, bore a hole to match the dowel (3/4″ in this case) and glue it in place with the rings perpendicular to the direction the screw will penetrate.  While not critical, this will reduce the likelihood of the dowel splitting and weakening the joint.  Anything over 1″ is probably enough.  I went the full 1.5″ that my drill guide could handle.

This hole is 1 5/8″ on center, meaning there is a full 1 1/4″ of material for the dowel to lock against.


The boards cupped a bit after planing.  More stable stock would not have needed this screw.

Next, drill a pilot hole for the screw, all the way through the dowel.  For cleanliness, I first countersank the hole, then finished it off with a long drill bit.  Red oak is tough, even for self-drilling deck screws.  Better not to risk it.  An extra long bit lets you sight to ensure the pilot hole passes through the dowel.


Luckily, this extra long bit (the only one I own) was perfect for the screws in use.

Finally, drive the screw and flush up the dowel.  I use a flush trim saw and either a chisel or a plane, depending on how much material remains after sawing.


Never to be seen again once the top is attached.

If done right, this joint is tremendously strong (at least compared to screwing into end grain alone).  Bench bolts are not terribly expensive, but oak dowel and screws are undoubtedly cheaper.  And, to be fair, this method requires less prep and fuss.

And less prep and fuss is what shop fixtures are all about.


Rethinking my Life Choices

A funny thing happened on the way to the workshop the other day.  I had four, 8/4 White Oak boards to laminate into a tabletop for the new compact Nicholson Workbench.  At over 20″ wide, the lamination would be far too wide for my lunchbox thickness planer.  And I needed as much thickness as possible for the final lamination so the workbench top would be as stout as possible.  So keeping everything aligned through the various glue-ups was paramount.

So I turned to something that cannot by any stretch be classified as a hand tool.  A self-centering dowel jig.


I absolutely adore this thing.

Using dowels for alignment actually serves two purposes.  First, it does the aforementioned aligning so any minor bowing along the length of a single board does not otherwise ruin the straightness of the glue-up.  Second, it reinforces the glue joint so if the glue fails, the entire thing doesn’t just fall to pieces.  It’s not as good as dominoes, obviously, but it’s also way cheaper.

Now I like to think that with a jointer plane and some car I can have a joint that will never fail.  And it probably won’t.  But the peace of mind of the reinforcing dowels is nice to have.  It matters more for larger timbers, though.


And spiral dowels are cheap.

The most important thing, though, is to make sure your dowel holes align.  This is more about keeping track of how you’re flipping the boards than anything.  Otherwise, you’ll use your extra dowels to fill in erroneously-bored holes.  And that’s no fun.

Trust me.


Short Cuts Make Long Delays

For an upcoming project, I need perfectly flat, perfectly straight stock.  But I only need 1/2″ thickness out of 3/4″ boards.  So I’m taking the laziest possible approach: skip planing.  But because the boards start off too thin for true skip-planing, I am pulling a page from the planing sled handbook and using blue tape to fill the hollows.


If I had a super power, it would be “cutting corners”.

Each outside edge is planed perfectly straight and without twist, and the blue tape will ensure the board rides to the planer table evenly and without pressing flat.  Thereby, one side of the board will be made perfectly flat.  With the tape then be removed, the board can be flipped end over end and sent back through the planer for perfectly parallel faces.

Speaking of parallel faces, I apparently left my winding sticks at my parents house, hanging off my old workbench.  I needed some for the above skip-planing.  So I made some out of scrap 3/4″.


About as basic as they can get.

At 1.5″ x 15″, I’m sure they will get re-purposed for actual furniture.


Engage the Forward Stabilizers

I’ve been a bit under the weather, so not much has gotten done in the workshop.  That’s not all bad, though.  It gave a Douglas Fir post, which I bought to be the top rails on the Japanese-style Saw Horses, more time to acclimate.  A good thing, too: it was sopping wet when purchased, as though left out in the rain for a day or two before the pallet was brought to the rack.

Over the last couple of weeks of drying, the two 36″ lengths that will become the top rails stayed quite straight.  There was barely any twist either.  What did happen, though, was some severe checking on the ends and on one face.  It makes me think the wood was greener than it should have been (or was left out in the rain for a very long time).


This is what happens when wood dries too fast.

None of the checks were structural, nor very deep.  Not even the long check down the face grain seen in the above picture.  That does not mean, however, that they shouldn’t be stabilized.  There will after all be mortises within a couple inches of each end.

My strong preference for stabilizing end grain checking (and knots, for that matter) has always been thin viscosity cyanoacrylate glue.  Like the Ents at Isengard, I just keep pouring CA glue into the checks until the gaps are filled and the bubbling stops.  Then I give it about five minutes to set before I seal it up with spray activator.


It looks gross, but it works.

I will let those sit for a day or two to fully harden inside, then I’ll square up the ends with a chop saw.  The end result should be pristine end grain and clean glue lines in the stabilized checks.




A Rare How-To

If you follow me on Twitter, you may know that I spent most of #blizzard2016 hand-cutting mitered half lap joints for a cherry side table.  The finished piece will have eight such joints, and although I’m only about halfway through the first of two frames, I’d like to reflect a bit on the process.


The goal is to have four each of these.

First off, some defined terms.  For the duration of this post, I will refer to the lower piece in the picture above (the E-G piece) as the “Angled Piece”.  The upper piece in the picture (the E-F piece) will henceforth be known as the “Recessed Piece”.  I’m sure they have proper names that have no relevance to what I’m about to say.

On each Angled Piece, there are three cuts to make.  The first two apply to all lap joints: a shoulder crosscut and a cheek rip cut.  The third is unique to the mitered half lap: a 45° miter cut across the face. My preferred order is (i) cheek, (ii) shoulder, (iii) miter, but whatever you do, always make the cheek rip BEFORE you cut the miter.   Otherwise, you will lose your guiding kerf on the rest of the rip as you get to full depth.


Don’t be a hero.  Start with the easy cut.

Once all three cuts are made, and after the cheek is flat and parallel to the face (whether off the saw or by router plane), it’s time to square the shoulder to the reference edge by paring down to the knife line.  Then is the most critical step: true the miter to 45° using the same reference edge to which the shoulder was squared.

When I cut the first joint, I thought to just straighten the miter, transfer the mark to the Recessed Piece and leave well enough alone.  After all, it doesn’t have to be perfect; it just has to be consistent.  Since my shoulder was square, this would have worked fine, assuming I sawed perfectly and my full knife line was intact.  But alas, the saw jumped out of the kerf and took part of my knife line with it.  So fitting the joint became trial and error angle finding with a shoulder plane.

If you instead true the miter to 45°, go ahead and still use the Angled Piece to transfer the mark to the Recessed Piece (it’s easier than using a combination square).  But if you saw less than perfectly, you can always fall back on just truing the mitered recess to 45°.  With a straight edge on the Recessed Piece and a square shoulder on the Angled Piece, everything will come together perfectly either way.


Like this, but without half an hour of shoulder planing.

One more thing: depending on how much material you remove from the miter on the Angled Piece getting it to 45°, you may need a couple passes on the shoulder to bring it back in plane with the end of the miter.  Otherwise, your inside corners won’t meet right at the miter.


The Clamp I Use Most

Apartment woodworking is mostly about making due.  But that can be said about much of woodworking.  And finding the right tool for the job is important regardless of square footage.


The clamp I use the most is not a clamp at all.

I’ve talked about alternative clamping styles before.  I’ve even showcased the machinists granite slab as a clamping apparatus before.  The reality is, sometimes a heavy, flat rock is much easier than an actual clamp.

When (for example) I need to glue back down some face grain that split while cutting a dado, I could use a parallel jaw clamp.  Or I could just put a big rock on top of it.  More times than not, I opt for the latter.  Because if I’ve done my job and my joints are square, weight is as good as mechanical clamping pressure.

Speaking of dadoes, these are for a three-board sushi tray from leftover pine.  If the sizing is okay (about 7″ x 15″), I will likely make a couple more from a tougher wood.


No stopped dadoes, this time.  My masochism knows some bounds.

Have a great weekend, everyone.


A Short Time Ago, in an Apartment Pretty Close By

Last weekend weekend, I cut dadoes. Two, to be precise.  Of the stopped variety.  To install the drawer runners in the dovetailed plant stand.


All dado bottoms flattened with a Veritas router plane.

There was a time when I would not have hesitated to strap a parallel guide clamp on the work and plunge-router the whole thing. A time not too long ago. A time I do not miss one bit.

The dado on the above right was cut with a 1 1/4 inch chisel.  After marking and chopping to depth one side, I then scribed the other side of the dado against the mating piece.  Knowing the vertical chisel chops would compress the fibers and move the knifewall, I intentionally marked the other side of the dado a bit narrow (1/64 or so).  The end result was a wonderfully fitting joint.


No reinforcement yet.  Only friction.

The dado on the right, however, was cut differently.  At the time of day (about 6am Eastern), I couldn’t be chopping with a chisel.  I live in an apartment, after all.  So instead, I scribed deeper and deeper knifewalls with the marking knife to get to depth.  This meant that, without the compression from the vertical chisel chops, the narrower scribe line on the other side of the dado was never compensated for.  I ended up planing the mating piece to fit the dado.  It’s still a very snug joint, just with a different approach.

I have since glued in the runners into place, with a bead of hide glue along the long grain bottom of the dado.  While I suspect this will be enough, I will also reinforce with some nails or buttons.  The drawer is almost done too; I’m merely figuring out the best way to fit the drawer bottom (without a plow plane).

Then it’s time for assembly.



More Routine Maintenance

Last night, I had a nightmare that all my chisels had rusted over and bellied.  So this morning, during my breakfast of mini-bagel and green tea, I made sure they had not.  I actually only got through 1/4-5/8, and will do the rest tonight.


Screw you, subconscious.

I love my 3×8 diamond plates for grinding and sharpening, but they are not ideal for lapping and flattening.  When I need something perfectly flat, I turn instead to the machinist’s granite slab pictured above and adhesive-backed sandpaper of various coarseness.

For flattening chisels or plane irons, I typically use 80 grit, 120 grit and 220 grit sandpaper.  I care more about flatness than mirror polish, so any additional polishing done on the diamond plates in the course of sharpening.  For plane soles, I might go up to 320 grit (which is overkill for everything but a smoothing plane, I know).

I’ve experimented with sanding belts and spray adhesive in the past, but I’m never happy with the adhesion, and certainly loathe the mineral-spirits cleanup of the spray adhesive.  With adhesive-backed sandpaper, I just need a scraper and a spritz of Simple Green cleaner.  The grit on the sanding belts definitely lasts longer, but it’s a small price to pay.