Chisel Handles and Rust Bluing

It’s been a little while!  But, the blog’s not dead.  The workbench was finished back in early April, in a mad rush.  In a rush, because at the end of April we had to move halfway across the country, from OK to GA.  Since then I’ve been busy setting up a new shop and working on some things, and now I need to get back into the blogging habit.

I’ll do a wrap-up post on the workbench soon, as well as a tour of the new shop, but for now I’ll just do a little post on tool maintenance/restoration.

As a sort of New Years resolution, I’ve been working on restoring and otherwise getting into shape some newly acquired as well as old, defunct tools.  A lot of my old vintage chisels were kind of ugly, with random pitting scarring the original finishes.  Some of them had developed an attractive patina, but others looked kind of leprous.  For the latter category, I decided to try rust bluing, in order to get a consistent color and sheen.

In addition to my western chisels I also worked on a few Japanese tools- a couple of new Koyamaichi shinogi-pattern chisels (for chopping dovetail pin sockets) and an old kanna from ebay.  The Koyamaichi chisels were not ground to as fine a side land as I’d like, so I ground the sides down to nearly a knife edge.  Of course this removed the forge finish from the sides, so I decided to rust blue these as well.

Rust bluing is a conceptually simple process- clean and finish the steel to as fine a finish as you desire, carefully degrease, then cause the steel to rust as evenly as possible.  Then immerse the item in boiling water until the red rust converts into black oxide (magnetite).  Gently remove any loose oxide with a fine wire brush or steel wool, then rust again and repeat the rest of the process.  After several (up to a dozen or more) iterations, a deep jet black finish can be attained.  This was the traditional finish for firearms prior to modern hot bluing, and the black oxide finish produced is very tough and corrosion resistant (certainly better than cold blue, and is said to be tougher than hot blue, if not as capable of a glossy finish- it is usually more of a satin or matte).

The rusting process is enhanced by wiping the parts down with a corrosive solution.  There are commercial products sold (Pilkington’s seems to be popular with gunsmiths), but none were locally available. As this idea more or less occurred to me on a whim, I didn’t want to want to wait for shipping and decided to try a recommended home-brew solution consisting of hydrogen peroxide, vinegar, and table salt.

I’m fairly happy with the results for a first attempt, though none are quite perfect:


The chisel fourth from the left is a bit lighter than the rest because I tried buffing with pumice, for some reason.  Not recommended.

There is quite a bit of information on this, however almost all of it is for firearms.  Specific to woodworking tools, Stanley Covington on Sawmill Creek has several old posts describing the topic.  Another source, which is I most closely followed, is here.

I lost count of the exact number of iterations I used, but I would guess around 10.  Using the quick bluing process from the link above, all the rounds could be done in one full day.  It is mostly waiting for things to boil and dry.

In addition to the bluing, I made new handles for the western chisels, to replace old ones which were either too short or too ugly.   They are simple tapered octagons with whittled socket tenons, of white oak and red oak.

Thanks for looking – Robert

Chisel Handles and Rust Bluing

Roubo Build #8 – Installing the Top

We left off this series’ last entry with the undercarriage completed and ready to accept the top slabs.  Now I’ll bring the tops to the same stage, which involves a bit of tuning and the creation of four large mortises.

A key design feature of the Roubo-style bench is that the front edges of the top slab and the legs are flush and coplanar.  This makes for a large and versatile clamping surface.  In order to facilitate good alignment of the top and base, I need to true the front (outer) edges of the slabs and make them square to the top and bottom surfaces.

Here’s one slab ready to work:


That’s the front edge facing up.  I check it for square:


Not so good.  But, I didn’t expect them to be square since the alignment during glue-up was a bit random.  This isn’t too hard to correct with hand planes – I use my #7 and #8 Stanleys, planing more on the high side (as shown by the square) at first.  After a few minutes it’s brought down to square:


I repeat the process for the other slab.  Then, the last tuning step is to check the squareness at the end of the slot for the dog block…


…which looks dead on.  This is pure dumb luck, but I’ll take it.  I fully expected to have to get out a paring block and true this surface.  It needs to be square because it will form one “jaw” of the end-vise.

With those tasks complete, I move the tops down onto the floor, flip the base upside down and set it on the slabs:


I’m going to mark the positions of the leg tenons onto the underside of the top slabs.  First I need to carefully align the base and top.  In the photo above I’m obviously only concentrating on the front slab.

Aligning the front slab consists of setting the front surfaces of each leg flush with the front edge of the slab, and the right edge of the front right leg (which is on the front left in the photo above) flush with the end of the dog strip.  Checking with the blade from my combination square:



I think you can see in that last photo why I made the third tenon shoulder on the front right leg.

Once I’m happy with the alignment I take a utility knife and mark around the tenon:


This was a bit awkward, especially on the front and back of the tenon.  When using a knife along the grain, the knife wants to follow the grain direction, which is not exactly where I want it to go.  The awkward positioning  makes it difficult to maintain control, so some of the markings are a bit wonky.  Mainly I just made sure that the ends of each mark were accurate, as I will use those to position the paring block that will determine the final dimensions of each mortise.

The results:


Now to make some mortises!  I will use the same technique as I did on the undercarriage mortises from the second entry in this series.  I install a long 3/8″ end mill in the plunge router and set up a guide jig.


I think my homemade mortising jog works nicely, but it doesn’t accommodate pieces this large, so it ends up getting modified by screwing extensions (maple battens) onto the router base plate, and clamping auxiliary stops in front of and behind the router to serve as fore-and-aft stops.  It looks like quite the Rube-Goldberg device, but once I had all of the parts together it could be moved from one mortise to the next pretty quickly.


And this is what the setup produces (along with a big mess that is not pictured):


I am staying a bit off the knife marks with the router.  I will pare the walls down to the knife marks by hand.

Heres’ the paring block set up for working an end-grain wall.  I use a couple of vintage chisels for this- a 1/4″ Ohio Tool and a 1/2″ Witherby.


Making progress:


All four mortises done:


I’m just about ready to install the tops now, but there’s one last thing.  I want to chamfer the tips of the leg tenons to make it a bit easier to get them started into the top mortises.  A block plane does the job:




Now to set the slabs on the tenons and see if anything fits!  This is a pretty big moment- the first point in this project where it will actually function as a bench.

It took some persuasion with an engineer’s hammer, but they did indeed fit:


I think the front slab required a little too much persuasion, and it won’t seat quite perfectly on the top stretcher.  I will trim the front leg tenons a bit the next time I remove the top, which should solve the problem.  I still need to cut a groove for the sliding deadman, anyways.  But for now, I’m thrilled to finally have this pile of parts transformed into a bench.

Thanks for reading!





Roubo Build #8 – Installing the Top

Roubo Build #7 -Finishing Up the Base

With the tops flattened, I will return to the base, which is currently just fitted together, but not fastened permanently.  There are some small tasks remaining.

“Small tasks” is misleading, actually.  It’s not that they’re huge tasks; there just seem to be a lot of them, and each has hidden complications.  Many of the operations here involve drilling large diameter, deep holes. I have a small drill press, which while better than nothing, is not well suited to this scale of work.  A good brace and bit would be the preferred tool for much of this, but my bits are no good.  Alas.  I make do with the drill press and cordless drill.

The overall plan for the base is to permanently attach the front and back legs on each side via the short stretchers.  I’ll glue the tenons and reinforce the joints with drawbore pins.  Then, the left and right assemblies will be attached with a de-mountable mortise and tenon joints with the long stretchers, using barrel nuts to secure the joint mechanically.  This will allow for relatively easy transport of the bench – each top will come off, then the base can be broken down into the two assemblies and the two long stretchers.  Each section can be carried single-handed, up and down stairs and narrow corridors if necessary.

I will also need to cut tenons on the tops of the legs, and drill oblong through-holes in the top stretchers for the lag bolts that will secure the tops to the base.

Here we go:


First I disassemble the base, and bring the legs over to the bench to begin preparing the drawbores.


Each joint will get a pair of drawbore pins.  Here I’m beginning to mark for drilling by determining the placement of the holes.  I make the initial marks on the edge of the mortise I’ll be pinning through.  Since I’ll be drilling across the mortise , I need to transfer the marks around the corner.


Once I’ve done this for both short-stretcher mortises on the first leg, I can just transfer the positions to the other three legs:


Then I mark the other dimension of the hole location.  I need to drill holes where the lines intersect:


Set up on the drill press with one of my adjusting stands as a support:


I’m drilling 1/2″ holes for 1/2″ red oak dowels.  The holes will not go all the way through the leg- there is not enough travel on the drill press quill.  I could finish the hole with a cordless drill, but I think there is enough material on either side of these mortises for a blind hole to be sufficient.  The holes extend about 1″ past the far mortise wall.


Ideally, one would drill for drawbores prior to creating the mortise, in order to avoid this sort of tear-out inside the mortise:


However there isn’t much you can do to avoid the tearout if you forget, and that’s what happened here.  I just clean the splinters out with a chisel and move on- there will be plenty of glue surface area left, plus a couple of stout drawbore pegs to make double-sure the joint is solid.

Now all the holes are drilled in the legs…


…and the short stretchers can be installed to transfer the hole positions onto the tenons:


Above, I’m using a 1/2″ transfer punch, shown here:


In general, transfer punches are fantastic and much better than using a drill bit to transfer a mark.  However, the point on this 1/2″ punch is quite fat, and in soft wood like doug fir it makes a huge depression.  This would be fine if I just wanted to drill on the center of the mark, but when you need to offset by a slight amount, the new offset mark ends up inside the “cone” of the indention created by the punch.  It’s awkward to make the offset mark, and when you go to drill on the offset mark it’s difficult to predict where the bit is actually going to go.

Of course, I went and marked all of the tenons before I realized this.  I would have rather used a 1/2″ bradpoint drill bit, some other kind of transfer punch with a finer point, or this same punch with a much lighter touch.

Anyways, here is how the offsetting works:


I set a pair of sharp dividers to my offset distance, 1/32″.  I can probably get away with much more, but since I’m using store-bought dowels, I’m being conservative.  I’ve had success with this offset before.


I set one tip of the dividers in the transfer mark, then lean the dividers over to make the offset mark.  The offset is always towards the tenon shoulder, in order to force the pin to draw the shoulder up tight.

You can see how the large divot from the transfer punch makes this awkward.  The result, after deepening the offset mark with a scratch awl:


The smaller mark centered in the divot is the original mark.  Now I just go back to the drill press and drill on the offset marks as best as I can.


Now I’ll make the actual drawbore pegs.  The raw material:


As I said before, these are 1/2″ red oak dowels.  It is typically recommended to avoid commercial dowels, and use riven (split) wood shaped into a dowel (by whitling or driving through a dowel plate).  That recommendation is based on considerations of grain strength, and the fact that 9 out of 10 commercial dowels will have significant grain runout.  Grain runout could lead to the pin breaking while being driven in and bending through the offset holes.

In fact, I looked through every single 1/2″ dowel at Lowe’s (more than 50 I would guess) and managed to find 4 that weren’t terrible in this respect.  Of the ones I chose, only about half of the length is usable on each.  So I cut my 16 pegs from the good sections, and I reckon these will do just fine, especially with a modest offset like I’m using.

For reference, this is what grain runout looks like on an oak dowel:


The “V”s in the grain (often called cathedrals by woodworkers) signify where the grain direction is running off the edge of the dowel, instead of straight up and down its length.

The good sections are cut into lengths to become pegs.  They are cut to be about 3/4” longer than the hole depth.  I need to taper the leading end of the peg to assist it in snaking through the offset holes, and I need to chamfer the back end to resist splitting while being driven in.  I’ve found that chucking the dowel into a cordless drill and using a belt sander is a fast method for this:



In the last photo you can see a block of paraffin – I like to wax the pegs to help them drive in easier.  Spinning the bit in the drill is a good way to apply the wax.  I can finish shaping and waxing a peg in perhaps 30 seconds this way.

Before I started shaping, I had to consider how far back to taper the end of the peg.  This is not as critical with a through-hole, but since my holes are blind, too much taper means not enough bearing area for the peg to do its job.  Since I have about 1″ of hole depth past the far wall of the mortise, I chose to taper 1/2″ of the peg, leaving 1/2″ of bearing surface on the far side of the mortise.

All 16 done:


But before I can drive these into the joints, I have some other tasks to complete.




These are the barrel nuts (Bench bolts) I ordered from Benchcrafted:


Pretty nice stuff, and not too badly priced, either.

These are a little bit tricky to install, as they require you to bore two intersecting holes, one of which is very long and into the end of a board.  Drilling an accurate hole of any length into the end of something has always been a challenge for me.

I start by finding the center of the mortise for the long stretchers:


Then I transfer that to the back (outer face), where I’ll be drilling.  Where the transferred lines intersect, I make make an indention with a scribe, which will give the brad-point drill bit a definitive place to start.


I then drill as far as I can with the drill press.  But it doesn’t have enough quill travel to get to the backside of the mortise, so I have to continue with a cordless drill until the bit comes through the bottom of the mortise.


That’s the hole through the leg done.  Now I insert the long stretchers into the legs and use the existing hole through the leg as a guide for drilling into the long stretcher.

I think I started getting a little frustrated at this point, and I didn’t take pictures for most of this process.  I’ll try to describe it with my sparkling prose instead:

The 1/2″ drill bit isn’t long enough to go more than a couple of inches into the stretcher while the leg is in place, so I can only get the hole started.  To finish, I needed to remove the stretcher form the leg.  The results of drilling into the stretchers were not very impressive – placing the long 1/2″ bolt into the hole I just created showed it to be out of plumb in both dimensions by a considerable amount.

*I did all of the initial drilling with a standard-length 1/2″ brad point bit.  I think it would be worth buying a longer bit just for this job, to enable the stretcher to be drilled as far as necessary without removing the leg (which is serving as a guide).  That was at least part of the problem.

The big problem with the out of plumb hole is that it’s hard to know where to drill the intersecting hole for the barrel nut.

I ended up marking a pencil line down the length of two adjacent faces of the long stretcher.  The pencil marks are centered on the hole at the end of the stretcher’s tenon, and parallel to the edges of the stretcher (plumb, in other words).  I would use the lines to drill for the barrel nuts, so I needed to “adjust” the existing holes to match the plumb pencil lines.

The adjustments were made by enlarging the existing 1/2″ hole first with a 9/16 twist bit, then a 5/8 twist bit.  Whereas I would normally try to follow the existing hole’s path when enlarging, here I consciously tried to steer the bits to match the pencil lines.  It sort of works.

I also went back to the legs and enlarged the 1/2″ holes to 9/16″, to give some wiggle room.  Then I chucked a forstner bit into the drill press and drilled for the barrel nuts.

And I’ll be damned, it went together drama free:


Both together:


The barrel nuts really work well, drawing the joint together solidly.




In the split top Roubo design, the tops are attached to the base primarily by tenons on the ends of the legs, but these joints aren’t secured with glue, wedges, or pegs, which enables the tops to be removed for transport.  Instead, the tops are secured in use by lag bolts which fasten the tops down to the upper stretchers.

It makes sense to do this prior to assembling the legs to the stretchers, so that I can use the drill press.  The plan is to drill two 3/8″ holes side by side, then take out the waste between the holes to create a single oblong hole.  The hole is oblong in order to allow the top to expand and contract – the head of the bolt can move back and forth a little bit along the stretcher.

I calculated that I would need no more than 1/4″ of allowance for movement, so 1/4″ was the offset I used between the two holes on each stretcher.  This turned out to be a mistake – not because 1/4″ was an inappropriate allowance, but because it is not feasible to drill long 3/8″ holes 1/4″ apart without the bits wandering towards the other hole.  At least with the equipment I have.  I think I would have been alright with a 1/2″ offset.

Again the hole needed was deeper than the drill press could go, so I had to drill from both sides.  With the wandering issue, I ended up creating a waisted hole (narrower in the center of the stretcher than at the either end).  This would not permit the lag bolt to move enough.

I did what I could to fix this with a gouge, but the depth of the hole meant I couldn’t get all the way to the middle, where most of the unwanted material was. A 3/8 twist bit got it as close as I think I can do, although it wallows out the sides of the hole, which makes the shape look out of whack but doesn’t affect anything of substance. Lastly I cleaned up the counterbore on the underside of the surface, leaving this:






The primary attachment for the tops are through four tenons, one on each leg.  The shoulders of the leg tenons should form a co-planar surface with the tops of the upper stretchers.  To achieve this, I’ve waited to cut the tenons until I had the stretchers installed and could nail down their final positions.  So, with the short stretchers and legs mated I can take a marking knife and make a small mark on each leg, indicating the upper surface of the stretchers:


Then after hammering the assembly apart I take a square and extend the little mark all the way around the leg, defining the future tenon shoulder:


These tenons will be a bit simpler than those on the stretchers- having a shoulder on only two sides instead of four.  The exception is the front right leg, which will have a shoulder on three sides- the third shoulder is to allow some material (1″) between the edge of the mortise in the top and the recess for the tail-vise.  This should become clearer as the build continues.

Otherwise the process for making these tenons is the same as you’ve seen here before.



All four done!  Notice the tenon on the top left- this will be the front right leg with the 3-shouldered tenon.




After finish planing the short stretchers and inside surfaces of the legs, I can return to the drawbore pegs and assemble the short stretchers into the legs.  I’m using liquid hide glue for extra assurance, although glue isn’t technically necessary with drawboring.



With the stretchers inserted and prior to knocking the pegs in, you can get a picture of how the offset in the drawboring process works:


The offsets looked closer to 1/16″ at this stage…the variance from my intended dimension is probably a function of the awkwardness involved marking the offset off the huge divot created by the transfer punch.  But I am actually thinking that 1/16″ should be the preferred dimension for this sort of large scale joinery, especially in a soft wood.  1/16″ is pretty modest in fact, compared to what I’ve seen online.  Anyways, it’ll work.

The pegs are all driven in with my rosewood mallet.  No drama with this, thankfully.  I am sure the straight grained dowels and wax helped this along.


With both assemblies pegged, I can go ahead and trim the pegs flush…


…and finish plane the inside and outside faces of the legs:


I’ve not planed the front and back surfaces yet, because I want to flush them with the long stretchers.  Before assembling each side frame with the long stretchers, I plane the top, back, and bottom of both long stretchers, leaving the front face to be flushed with the side frames.  They are very close to flush as-is, but the legs are proud of the stretchers by 1/64″ or so.

Now I assemble the whole base unit, and tip it on its side for planing flush:


One side done:


With both sides done, I can flip the base upright and reveal the (substantially) completed base.


Most of the remaining work will be done to the tops, beginning with the next entry in this series.


Hope you’ve enjoyed.  Thanks for reading.











Roubo Build #7 -Finishing Up the Base

Roubo Build # 6 – Flattening the Tops


This is where I left off the previous post.  A bit rough looking, you might say.  The task at hand now is to take this mess and make the top and bottom smooth, flat, free of twist, and parallel.  I’ll do one surface by hand, one shaving at a time.  The opposite surface will be handled by the planer, which works beautifully with a flat surface to reference from.

Here are the tools I’ll be working with:


On the left, a circa 1890 Stanley #8 jointer plane.  I use this for fine tuning a flat surface, taking relatively thin shavings.  In the middle, an old Sargent transitional jack plane, used for hogging off thick shavings in the initial stages of flattening, and for getting rid of dried glue.  On the right, a circa 1930 Stanley #7 jointer, set up as a trying plane and used for taking medium shavings.

Before I start flattening, I want to get rid of most of the excess length.  I strategically cut off most of the extra length from the worst, most misaligned ends.  This should enable me to preserve more thickness.  An old Disston crosscut saw does the job here:


Now I can start by getting rid of the worst unevenness with the jack plane.  I alternate working along and across the slab, mainly focusing on reducing the strips whose edges ended up higher than the rest, until the whole slab is roughly even across its width at any point, with no steps between strips:


At this point, I can start with the #7 plane.  I will work very methodically from here on out.  To gauge my progress I find it helpful to make a cross-hatch grid with a lumber crayon.  The plane removes the crayon where there are high spots, and anywhere that crayon remains is a low spot.  Flattening is mostly a process of bringing the high spots down to the same level as the lowest low spot.


I plane along the full length of the slab, overlapping strokes from one side to the other.  Then I will do traversing passes, going across the width of the slab, working from one end to the other.  After a few rounds of each, you start to get a sense of progress as less and less crayon remains. As I go, I periodically stop and check for twist.  I use my longer pair of winding sticks:


This is what you want to see- the top of the back stick being eclipsed evenly by the front stick as you lower your head:


I lucked out on these slabs as I never had much twist to deal with.  I suppose the cauls I used, screwed down to a decently flat benchtop, helped with that.

Had I found twist, I would have to take corrective action by planing more on the two high corners.  Then I would proceed with more straight-through strokes, then I would check for twist again, repeating the corrective action again if necessary.  I had to do a tiny bit of this on one slab.  I’m thankful the twist wasn’t bad, as removing a lot of twist on something this large can devolve into a maddening back and forth game of correcting strokes, losing lots of thickness (and time) in the process.

Anyways, once I feel it’s free of twist, I’ll keep doing straight through strokes until the low spots are finally gone:



I check for twist once again.  It checks out good, so I bring out my 36″ precision straight edge and check for straightness across the width and along the length:


Since I’ve been doing methodical rounds of through-shavings, for the most part this showed my flatness to be pretty good.  However, I did find some “snipe” on an end or two, where I planed a bit too much as I came onto or off of the slab- this is very easy to do.  The fix was to mark the low area:


And avoid planing that area for a couple of rounds.  In this case, it meant starting the plane strokes a couple of inches in from the end.  Then finish with a round of through shavings, and check again.  Eventually I ended up with something that was pretty good.  I really just need a decent reference surface for the lunchbox planer at this point, and I know that the tops will need to be fine tuned once they are installed together on the completed bench.  Still, it’s pretty damn flat:



…and I’m pretty damn tired.  But satisfied, like I’ve done some honest work, ha!

The next day I set up the lunchbox planer outside to flatten the opposite side.


Before running the slabs through, I took the jack plane and knocked off the worst high spots and the hardened glue drips (which were much worse on the bottom).  The glue chunks can wreck the planer’s knives.

I was a little concerned about the planer handling these big slabs- mainly wondering about the gears it uses to drive the feed rollers.  I will guess the slabs weigh 60 pounds each.  They are also almost as wide as the planer will accept.  However it was no big deal, and all went well.  The DW734 has been a marvelous machine for me so far.

Back in the shop and stacked and stickered:


As you may have noticed, this whole operation has generated tons of shavings.  After all that work, I can think of nothing better than to grab a good beer and watch those shavings go up in a glorious blaze:


Cheers, and thanks for reading.


Roubo Build # 6 – Flattening the Tops

Roubo Build #5 – Making the Dog Strip

If you’re not sure what bench dogs are, imagine a small piece of wood that fits in a matching hole in a benchtop, and the piece can be raised above the top to serve as a stop to butt something against (say, for using a hand plane).  In conjunction with an end vise (more on that later), a board can be clamped while sitting flat on the benchtop.

The dog can be round or square in section, and if I was using a solid, giant slab of wood to make the top then a round dog would be the logical choice.  But since I am laminating strips together to make the top, it becomes much more feasible to make a square dog, which might function a little better for most tasks, and if nothing else, look a bit nicer.

I’ll start with one of the two strips that have yet to be glued to the front slab- the one which will be second from the front.  I thickness it to 1-3/8″ using my lunchbox planer.  Then the side profile of the dog is traced onto the face of the strip:


The pattern for the dogs is courtesy of Benchcrafted, who make the wagon vise I will use.  The dog is canted slightly (2 degrees) so that it leans towards where the wagon vise will be.  To make a dog hole, a recess 1″ deep needs to be cut in the shape of the pattern shown above.

I’ll be making 8 of these dog holes, so it’s worth it to make a quick router jig.  The basic idea is, again, courtesy of Benchcrafted.


I just took a scrap of 3/4″ birch plywood, cut it in half at 90 degrees on the miter saw, then cut the notch on the left side at the bandsaw.  A pine scrap serves as a fence for the jig, abutting the top of the dog strip.  The plywood pieces are then aligned with the pencil outline to get the 2 degree cant, clamped in place temporarily, then screwed to the fence to fix the positioning.  Now this jig can be slid down the length of the dog strip to make each hole.

The router bit I’ll use is a 1/2″ diameter x 1″ long top-bearing pattern bit.


The ball bearing above the cutting edge of the bit will ride along the edges of the plywood, which keeps the bit from cutting past the outline of the dog.  This is done with a plunge base for the router, and I remove wood in 1/8″ increments of depth.


The first few plunges are tricky because the bit is not going down far enough for the bearing to contact the plywood, so it has to be done freehand, leaving about 1/16″ inside the edge of the plywood outline.  Once the bit is going deep enough for the bearing to hit the plywood, the 1/16″ that has been left from the first passes can then be removed.

The end result:


Repeated seven times:


To finish the dog strip, I need to cap the face I’ve just cut these holes into with a 3/8″ thick strip.  I had a couple of extra strips that I rejected from using in the tops, due to too many knots.  However one of them had a clear enough portion on one side that I could use to make a 3/8″ strip.  I had to drag my little bandsaw out onto the workbench and resaw the strip.




This is a Rikon 10-305 10″ bandsaw.  It’s not bad at all for what it is, but apparently resawing 3-1/2″ wide douglas fir is really pushing its limits.  The problem is the 1/3 hp motor- even with a brand new Woodslicer blade it had to be babied to keep the motor from just coming to a stop mid cut.  That said, it did the job, and the cut quality was quite good:


I resawed the strip to be well over finished thickness, 5/8″ or so, just in case things went south during the resawing.  The rest can be wasted with the DW734 lunchbox planer:


I then spend some quality time with the shop vac, and I’m ready to glue up this dog strip:


In case you’re wondering what the third strip is, it’s just the offcut from the resawing that I’m using to help even out the clamping pressure on the thin strip – there’s no glue on it.

After letting the glue cure for 24 hours, I can add the dog strip and the front strip to the 6 strips I’ve already glued together for the front slab.  The two strips I’m adding will have to go through the same gap-finding-and-correcting process I used in the previous post, and then they’re ready to go.  To get an idea of how they go together:


You can see that the dog strip is shorter than the rest.  That leaves a gap on the right side between the front strip and the rest of slab, and that gap will be the thickness of the dog strip.  This gap will be utilized by the wagon vise, which basically just moves a small dog strip/block (containing it’s own bench dog, canted the opposite direction from the rest) up and down that gap, enabling items to be clamped between the dogs.

Here’s everything glued together:


And that’s a wrap for laminating both tops.  Now I’ll need to trim the excess off the lengths, and begin the herculean task of flattening these rather uneven glue-ups.  I’ll be covering that in the next post in this series.

Thanks for reading.


Roubo Build #5 – Making the Dog Strip

Roubo Build #4 – Laminating the Tops

I’ve finally got a minute to update the split-top Roubo-style workbench build.  The last post saw me complete the basic joinery for the base, and get it assembled.  Of course the base is not yet complete- I still need to make provisions for drawboring the short stretchers into the legs, and I need to drill for the barrel nuts that will hold the side frames to the long stretchers.  Not to mention cut the tenons on the tops of the legs, drill for the lag screws that secure the top to the base, and finish plane all the stretchers and legs…yeah, not done!

But I’ve set the base aside for the past week, and started working on the two top slabs. A couple of years ago when I did the initial work, I ripped some douglas fir 2x12s into 3-1/2″ wide strips – 18 total.  Why not just buy 2x4s if I’m aiming for 3-1/2″ wide boards, you ask?  Well, construction lumber in the U.S. has annoying 1/8″ fillets on every long edge, which create gaps when you glue them up face to face, meaning you have to take off a lot of wood from the thickness and/or width of each board to get rid of them.  Additionally, 2x12s are generally cut from bigger trees and have fewer knots and other issues.

Anyways, I ripped these with a circular saw, and they’ve been sitting up on the rafters of my shop for a while:


I took them down, and determined which were the best edges (so that the top surfaces of the slabs will have few or no knots), and the general grain direction of each board (to make planing easier later on), and marked this information on each board.  Then I clamped each group together to see how the glue joints might look.  Of course the factory surfaces of construction lumber aren’t very smooth- meaning if I just glued them up as is, I’d have gaps all over the place.  At best that would be unsightly, at worst it could cause the tops to delaminate in the future.

So, I need to plane each board until it is “locally straight”.  What I mean by that is, I will allow each board to have a general bow over its 8′ length, but with no discrete bumps or hollows.  I will also remove any crowning across the width, since this will definitely cause a gap on the outer surfaces.  For this job I don’t want to use a long jointer plane – it would try to get rid of most of the general bowing, which would take too long and remove too much material.  I will rely on clamps and glue to make up for the general bowing.

A #4 smoothing plane seems well suited.  I sharpened up both of mine – a vintage Stanley and a Lie-Nielsen:


Even with taking the “shortcut” of allowing a general bow, this is still a hell of a lot work!  The shavings are nice, though.



This was enough planing that I got hot spots (not quite blisters) on the base of my right thumb, and oddly enough, on the ball of my right foot.  Anyways, once you get into the groove it’s not a half-bad way to spend the evening.  Once I think I’m done with one slab’s component boards, I clamp them together again and see if there are any gaps remaining.  Invariably there are a few, so I mark where they are:


The pencil “clouds” are highlighting where I see gaps.  Now I can unclamp the group and basically repeat the prior step, except that I’ll focus on the areas around these pencil marks.  Since the pencil is marking gaps (hollows), that generally means there is a bump at one or both ends of the mark.  A 12″ straightedge helps to find these bumps, since it will pivot freely when it’s sitting on one.DSCN1191

Once each board checks out with the straightedge, I go ahead and glue them up.  I want plenty of clamping pressure, so the clamp racks get cleared.  I think I used 18 in total. I also brought out a couple of cauls I previously made.  I screwed the bottom of the caul to the top of the existing bench.  This should keep the boards mostly aligned while I’m clamping, and it will hopefully keep any twist in the glue-up to a minimum.

I enlisted the help of my girlfriend to spread the glue and tighten the clamps.  This has to be done quickly, since the Titebond III glue I’m using has an open time of 10-15 minutes.  The clock starts once the glue hits the first board.  There was no time for photos of the process, but we used a pair of small paint rollers to apply the glue, which we poured into a paint tray.  It worked nicely, and here is the result:


I’m obviously not concerned with the ends aligning- it’s still about 18″ too long, so there’s plenty of leeway.

The slab in the photo above will be the back slab, which will just be a solid slab about 11-1/2″ wide.  The front slab will be the same overall width, but there is a bit more going on in the front 6″ or so.  Therefore I’ll only glue the rearmost 6 boards at this stage, instead of 8 like the back slab:


While the two cauls did a nice job, I really needed two more, as you can see from the misalignment on the end there.  I could have used my remaining clamps and some scrap wood to bring that closer together, but for some reason it didn’t occur to me until it was too late.  This will probably cost me some thickness on the finished slabs.  I’m hopeful that I can cut off the worst of it when I trim them to length, which I will need to do before flattening to preserve as much thickness as possible.  If I can end up with at least 3″ I’ll be pretty happy.

The front slab still needs to more boards glued onto the front.  One of those will be the dog strip, which will be the subject of the next post.

Thanks for reading.






Roubo Build #4 – Laminating the Tops

Old Projects – Food Pantry Shelves


I thought I’d take a break from the Roubo series to share some old photos of one of my first real woodworking projects.  I made a couple of large shelves for a local food pantry.  I wanted to get some experience with joinery, and I’d say that was accomplished- there’s a total of 80 drawbored mortise and tenon joints between the two shelves!  40 were done entirely by hand, the other 40 were done with the help of a router.

Half of the joints are through tenons:



The sides were created first, with 5 short stretchers tenoned into the rails:


The opposite rail can be brought in now:


After installing the drawbore pins and letting the glue dry, the pins can be trimmed flush, then the whole side can be planed flush.



That gets us halfway there:


I’m amazed at how messy I let my old shop get.  I really like to keep my space neater these days.

Now onto the long stretchers.20150118_135214

The drawbore holes are drilled first into the uprights, and are more or less visible in the photo above.  Then each stretcher is installed, and the tenons are marked with a drill bit poked through the hole in the upright.  Once the center of each hole is marked onto the tenons, the stretchers are removed.


I now offset these marks a bit towards the tenon shoulder (I think I offset these by 1/32″).  The offset forces the pin to sort of bend around the tenon as it’s driven.  This pulls the tenon shoulder tightly against the upright, making a completely solid connection.  Glue isn’t even necessary, although I did use glue for extra insurance.

The pins themselves are cut from 1/4″ red oak dowels.  I selected the straightest dowels I could find at Lowe’s.  Wonky grain could be disastrous if the pin snaps as it’s driven in.


There are 76 pins in each shelf…

They are cut to length and tapered on one end so that it can fit through the offset holes.  A cordless drill and belt sander (and a girlfriend to help!) made this relatively quick:


I used 60-minute epoxy, thinned a bit with acetone.  This assembly takes too long to get together for regular wood glue, which sets in 10-15 minutes.


The shelves are 4′ wide, so ratchet straps were the best way to pull everything together.


Once the shoulders are seated tightly, the drawbore pins can be driven in.20150118_182402

And the assembled frame:



The pins are then trimmed and the front and back edges planed.

Then we’re ready for finishing.  I used a home-brew danish oil, roughly 1/3 polyurethane varnish, 1/3 linseed oil, and 1/3 mineral spirits.  My helper in action:


This is more or less what they look like after huffing varnish fumes for a while:


To complete the project the shelves, which are 1/2″ birch plywood with some birch edge banding, are attached to the long stretchers with some #10 pan-head slotted screws from underneath.  Lastly, some leveling feet are added.


Thanks for reading!



Old Projects – Food Pantry Shelves