Category Archives: Furniture

My second commission – part 12


Let’s start with a confession. I did stuff around with some of the photos in this post. It is the first time that any of JNSQ’s photos have been altered, as far as I can remember anyway.

The previous posts in this series can be found here.

In this edition we will take a look at some of the joinery and the first phase of preparation of the top for finishing.

The first bit of joinery I attempted was to fit a small block to the top of each leg construction. It will be the point that fix the centre of the leg to the centre of the top. All the other connection points will allow for wood movement, but these two will not. This means that the top will be able to move freely with changes in humidity, but the centre will remain fix to the centre of the legs. I think this is called a T-bridle joint. One feature of my Langdon mitre box and saw that came in handy here was it’s ability to set the depth of cut. Obviously you can simply do this by hand, which would also be much quicker. Where the mitre box might have an advantage is when you need to do heaps of these joints with the same dimensions. In this case it was an opportunity to work out how to set the mitre box for a job like this. That way it will be easier next time.

A router plane works well for the cheeks of the bridal section. It was a bit of a challenge to hold such a small piece while cutting the cheeks to depth. The solution was two dogs and a Veritas gadget. That could be a good name for a progressive rock band or a retrogressive gin mill (Two dogs and a Veritas gadget), come to think of it.

Next up were the slots in the top of the legs.

The two aprons are also jointed to the legs by means of T-bridal joints. Here I am marking the exact location of the shoulders using the leg.

That was followed by the same sequence involving the mitre box, router plane and careful chisel work to perfect the shoulders.


So then on a crisp and bright winter’s Friday morning I started to flatten the bottom side of the top. Seeing that it is the first table top of this size in Kershout that I am doing by hand I thought that the bottom side would provide an ideal opportunity to work out which method works best. The major challenges posed by this top are the schizophrenogenic nature of grain and the extreme hardness of the wood. As every self-respecting JNSQ Woodworking reader should know by now, we deal almost exclusively with feral boards from the ancient Knysna forest. Each of the trees that the boards for this top were sawn from would have been over 500 years old.

If you deal with wood like that it is my opinion that one has a real responsibility to do the best possible job of allowing the story of the tree to be told. In my estimation that means a delicate balance between careful surface preparation and leaving certain imperfections that relates to the history of the piece of wood. George Nakashima’s immortal oeuvre of work (which inspired this design) lends itself perfectly to getting the most out of feral hardwood such as what I chose for this top. How much and which imperfections are left to tell the story is of course in the eye of the beholder.

Anyway, I started experimenting with various different tools to see what might work best in flattening such a challenging top. The techniques I tried included a belt sander, a low angle jack plane with a toothed blade, a standard no.3 smoothing plane (45° frog) with a back bevel of 25° creating an effective pitch of 70° and a shop made fore plane with a blade pitched at 50° (aka York pitch).

The belt sander has always been one of my least favourite tools. It makes noise, it is all over the place and seems to be the best possible tool to turn a flat surface into the famous Valley of a Thousand Hills. What I found was that it is less harmful in such hard wood, but still not an option if you are aiming for a superior surface. The low angle Jack plane (12° bedding angle + 38° micro bevel for an effective pitch of 50° and a tight throat) worked diagonally to the grain clearly had the measure of the wonky grain, but would have taken too long for what would suffice for the bottom side of the top. I did not aim for a perfectly flat bottom side.

Next up was the back-bevelled smoothing plane. It worked even better (in terms of finish) than the low angle plane, but was difficult to push due to the high effective pitch and therefore even slower at removing material. It is also important to mention that this strategie seize to be effective in difficult grain when you try to take a fat shaving.

So I rolled the dice and tried the shop made fore plane (50° effective pitch) diagonal across the grain. It wreaked havoc in a semi controlled sort of way. This particular blade has a fairly substantial camber and it took no prisoners in the process of removing the necessary material in a timely fashion. In the pictures below you can see the characteristic scalloped appearance of a surface smarting from such treatment.

This is one of my attempts at manipulating the photos to highlight the pattern left by the plane.

The wooden plane in the picture below enforced the above damage. Of note in the picture is the bottles of water I consumed during this arduous labour of bellicosity.

Here we have an example of a part of the history of the tree that is often neglected to be told. Yes I know some of you will think I have lost the plot. Probably something along the lines of: “The #%$@&* hippy has been smoking too much pot.” The reality for me is however that the wood I in my collection have all sorts of imperfections and it would be impossible to create anything of reasonable size without these imperfections exposing themselves. I have therefore made peace with having to incorporate imperfections and try to design in such a way that the eccentricity of the stock enhance the aesthetics of the piece I am building.

Here are a few more tweaked pics with an array of tools that were used to tidy up the bottom side of the top.

Then finally it became time to employ some of the lessons learnt on the bottom side to the face side of the top. It took me three full days of planing at 45° to the grain with a toothed blade in a low angle jack plane to get the top as flat as I wanted it. The two pictures below were taken after the first day.

The dogs on my assembly table came in quite handy during this brutal process.

The toothed blade created these beautiful patterns in the areas that were approaching flatness.

This was the end of day two.


Once the entire (well almost) face side were in the same plane I removed the bulk of the rhombi left by the toothed blade using a no. 112 scraping plane. It was the first time I used this particular tool for a huge job like this. I prepared the blade the way that is recommended by my woodworking icon David Charlesworth. In my case a 45° main bevel, 50° polished micro bevel with a 75° burr set up in the plane with the blade leaning forward at 20°. The plane is an absolute joy to use when set up like this. You have to make sure you take very thin shavings of course. Some sanding with my shop made sanding planes took care of the rest of the rhombi.

While grappling with the rhombi I took short breaks to tidy up the cracks in the top. They all had lots of loose splinters of wood and other ancient bits of debris inhabiting their depths. This task was mainly accomplished by using a very old pocket knife that used to belong to my grandparents.

At this stage I shaped the curved ends of the top. As you can see I marked out two lines using my fingers as a fence. These lines guided the removal of waste to create a very gentle yet quite wide bevel. Once the bevel were established, the end grain area were rounded off ever so slightly using the same technique. My no. 9½ Stanley block plane did most of the donkey work and was then followed by a low angle smaller block plane, which was in turn followed by gentle sanding.

When I got a bit tired of the top I continued to chip away at the last bits of joinery.

Once the two aprons were fitted to the legs with very precise bridal joints, I started working on the massive beam that connects the legs at floor level. The Witpeer beam was laminated and squared up more than a year ago. It gave the wood a very generous time frame within which it could settle all possible disputes the fibres might care to raise (so to speak). It turns out that a very dense laminated beam like this stayed pretty much dead straight in all it’s  dimensions, but managed to go out of square by what appeared to be a full mm. That was fixed by hand planing a face side and face edge perfectly square with each other and using those reference surfaces to square up the others with my electric planer.

I transferred the inside measurements of the joinery from the aprons to the beam.

Using the above reference point I took the beam to the Windsor leg to mark out the exact location of the other side of the fairly complex stopped bridal joint (my own name not necessarily correct terminology) which will marry these two structures.

This is how far I got with this joint at present.

It was now time to break in my precious polissior that one of my favourite woodworking personalities and über craftsmen Don Williams (of The Barn on Whiterun fame) sent me earlier this year. That entailed rubbing the heads of the grass/straw on a rough piece of scrap wood and tidying up the appearance on a spindle sander.

I can thoroughly recommend reading Don’s article on this epic tool from the past.



I used the Polissior to burnish the top after perfecting the finish with gradually increasing grid sander paper on a orbital sander. I went all the way to 600 grid and did two rounds of wiping the surface with a damp cloth to raise the grain before sanding it back down with the 600 grid. You can see the effect of the burnishing in the pictures below.

Aoife helped me to apply a tung oil/turpentine mixture. We kept the surface quite wet for 30 minutes by reapplying the mixture where the wood absorbed it and then wiped it down with a clean and dry cloth.

As you can see it was one of those unbelievably satisfying moments in woodworking where the wood rewards you for months of painstaking elbow grease. Kershout is simply one of the most beautiful species of wood on the planet. I want to reiterate that there were no pigment added what so ever. This is what it looks like after tung oil mixed with turps were applied!!

The top will now rest for two weeks before we will apply beeswax with the polissior. Stay tuned my brethren!!

My second commission – part 11


As promised we will look into the process of jointing, gluing, and inserting dovetail keys into the top of the table in part eleven of our journey.

The rest of this particular chronicle can be found here.

The Kershout boards in the picture below were prepared up to this point towards the end of last year and has since been kicking it with my 1969 MGB in a separate garage.

The first task is to arrange the boards as best as you can with regards to colour matching and balancing out defects. This is where you whip out your artistic licence. This is after all a tribute to the legendary George Nakashima.

I took the opportunity to see what the trapezoid leg would add to the overall look. The top looks very light in colour (in this picture), but I can assure you that it will be transformed to a very dark reddish brown once the finish is applied. The Kershout dovetail keys contrasts exquisitely with the lighter Witpeer boards that makes up the trapezoid leg. I also like the darker lines created by the defects on the leg. It was strategically place to balance out from an aesthetic point of view. We will see later in this post how the reverse of the mentioned timber combination has a similar effect with regards to the top.

As you can see here my bench really came into it’s own working on the edges of these boards during the jointing process. I first prepared the edges so that they were close to the desired configuration, which is a very slight bow in the length.

Then the boards are clamped together with the two edges that will mate (so to speak) flush with each other and folded much like book-matched pieces before opening the “book”. This nifty trick leads to a cancelling out of the minute error that might arise in squareness of these edges with regards to each other. This technique is sometimes referred to as match planing.

Didi gave me a few pointers.

The Kershout is so ridiculously hard that I had to resort to using an alternating attack with my Lie-Nielsen low angle Jack plane armed with a toothed blade and a Shaw’s Patent Sargent no. 14C armed with an aggressively cambered blade.

Once the artillery softened up the enemy, I moved on to this shop made jointer plane to finish off the job.

I find my Festool Domino to be a very useful tool to keep the edges flush during glue-up.

It has become my custom to do only one of these edge joints at any one time given the short window to get the job done in our dry climate. Each joint is then left in the clamps for at least 16 hours. In other words, I tend to leave the glue-up for my final task each day. It is usually done at around 17h00 and left over night until around 09h00 the next morning.



Ready for the final glue-up.

I had to buy a set of 1.3m long 1″ pipes for my pipe clamps in order to do this final glue-up. Of course, as you would expect, my 1.2 meter wide assembly table was too narrow to accommodated the clamps for this glue-up. The situation therefore necessitated some problem solving on my behalf.

As you can see here a piece of wood (for each of the bottom clamps) was cantilevered off the edge of the table held in place by a clamp through a dog hole. Oh! … and yes, in case you wondered, it is my daughter’s “Biscuit finds a friend”. My English is not advanced enough to indulge in such haute literate.

As I have mentioned before, a mere mortal tends to sweat like a Gypsy with a mortgage during our sweltering rainy season. Didi is the master of African Climate Control (aka toplessness).

… and Bob’s your Uncle.

I modified the strip of wood that links my trammel points to draw a curve to soften the appearance of both ends of the top.

Marking the location of the dominos like this helps to remember where they are when further shaping is done.

The waste was removed with an electric jigsaw. It is a crappy old Black & Decker that I bought many moons ago while still living in New Zealand. I do not use it very often to start with and do not recall ever calling upon it to munch through Kershout. As most things you do for the first time there were a few lesson to be learnt. These things (for lack of a better insult) cut on the pull stroke, which translates into a messy splitting out of fibres at the top edge. Therefore (in hind sight) it is desirable to have the bottom of the top facing the jigsaw when doing this job. Secondly, I realised that I used a blade that was too aggressive, which did not help either.

On the flip side, this indiscretion coerced me into a design tweak that might (or might not) add an interesting twist. You will have to wait and see just like me.

Another reason I chose this shape for the ends of the top, is to enhance the appearance of it being sliced from a massive tree trunk. The idea is that this shape resembles the end of a trunk that was chopped off by axe. If you imagine a board cut from a trunk like the one in the first photo below, it would probably resemble the top of my table as seen in the picture below. That is in my mind anyway, you might feel different.

Then it became time to fashion a few dovetail keys to stabilise the obvious cracks in the top.

I worked out how many is needed of each size.

Here I tried to work out where to place the keys with regards to my sense of (randomly planned) artistic balance. The picture below was not the final version that was decided on, but somewhere towards getting there.

For the design of the keys I chose an angle of 9º, which repeats all through the design of the table. This is an idea you might want to consider. You draw only one key, chopped off at different lengths, and write on the template the number of keys needed of each length. It is then cut out, traced onto the wood as many times as the key tells you and then you chop off the ends and repeat on the next sized key. This way they all have the same shape, but of different lengths in an attempt to add visual interest.

As so.

The keys were liberated from the above Witpeer board by means of a bandsaw.


Another useful trick is illustrated below. Clamping a piece of scrap wood across the top to hold the dovetail key firmly in place while it’s exact configuration gets marked out on the top.

Drilling out the waste by hand in such hard wood is no joke. “Trust you me”, as they say around these parts.

Enter: Lie-Nielsen merchandise in tandem with my trusty shop made Assegaai mallet. I chose the mallet as I needed a bit more heft than what the so-called Je ne sais quoi Persuader can deliver. When working “stone”, the extra heft is a must.

The lazy winter sun give us a better idea of the warm colours of the Kershout as it infiltrates my shop during the late afternoon.

It seems as if this post is riddled with tips, so here is another one. In order to see the scribe line better, one can have a small torch lying on the top to cast a shadow into the line. On my bench this is usually accomplished by positioning the bench light in a similar fashion, but clearly this top is too big to take to the bench.

Once the key enter it’s mortise like this I stop refining the fit. The key is then clobbered home after a frugal application of Epoxy, which acts as lubricant as well as an adhesive. The clobbering is done with a heavy mallet furnished with a thick sealskin face (not pictured).

As you can see here (minus the heavy mallet).

One week later the keys were planed flush using the two planes pictured.

As you can see the Witpeer keys contrasts nicely with the Kershout, much in the same way as the opposite combination works splendidly in the trapeziod leg.

We will get into the preparation of the top for finishing and the key bits of joinery in our next riveting edition of this series.

A Table for a shebeen – part 3


The year is hurrying towards it’s inevitable end and the temperatures in my tropical haven are racing upwards at the same rate. The massive shift in ambient humidity that a good rainy season can induce always wind me up to get joinery that is supposed to last a really long time assembled during the driest part of the year. In my case that means before the end-of-year holidays, as the proper electric storm driven downpours tend to ignite a sudden hike in humidity around mid January. That is of course if we are lucky, because water has been in short supply in this sweltering savanna of ours.

For the purpose of this project I wanted to get the two leg-assemblies done before we leave for a hitherto tranquil spot on the Azanian south coast. Unfortunately the Zuptopian conglomerate has since done it’s utmost to “poison and destroy my brothers”, but let’s not dwell on the bane of my life while we could be discussing woodwork.

As discussed before, I prefer using Assegaai for my custom made drawbore pegs. I usually try and find a perfectly straight-grained piece before ripping strips on the table saw that are then fed to my Stanley no. 77 dowel making machine (not pictured). Unfortunately it is pretty much impossible to split (as apposed to rip) stock along the grain then using the no. 77 as you need perfectly square strips, but if you use fairly straight stuff it still turns out dowels that are far superior to the rubbish bought in local stores. I wrote in more detail on this process in previous posts.


For the sake of trying something new I decided not to go with wedged through tenons as I did with the two workbenches that was built with similar sized stock. I am however partial to the idea of using wedges to ensure maximum strength. Therefore it was decided to experiment with wedges in a closed mortise. You will notice the kerf prepared for the wedge in the monstrous tenons. The sides of the mortise that needs to allow for the wedge to expand the tenon were adjusted. Then it is simply a matter of positioning the wedge in such a way that it sits in the entrance to the mentioned kerf while using pipe clamps to coerce the tendon into it’s mortise. That is of course followed by tapping the drawbore pegs home while the clamps are still in situ. The pictures below also show how I used my Festool Domino to cut the slots for the bits of wood that will fix the aprons to the top.

This was how the two leg-assemblies spent their festive season.



The drawbore pegs were then worked flush by sawing and hand planing.


In order to mark out the exact location of the tenons of the two beams that link the two beams between the leg-assemblies (at an angle), I had to assemble the undercarriage.


By clamping the linking beams at exactly the right angle, I was able to mark out the shoulders of the tenons.


For some reason I found it very difficult to keep to the marked out lines while sawing away the waste by hand. It was the first time that I encountered this problem and still do not know exactly what was going on as I have done many similarly sized tenons in the same wood with the same saw?? Therefore I switched to the bandsaw for the rest of the work.


I usually like to saw close to the line and hone in on it by planing the cheeks to a perfect fit.


Marking out the location of the mortises.



Drilling out waste.

Removed the rest of the waste by vertical chopping.

Then it was time for the final glue-up of the undercarriage. I was very happy with how it came together and it sure is a robust construction.

The undercarriage received a few layers of Woodoc.

Before installing the undercarriage, the bottom of the top received a few layers of boiled linseed oil.

The pictures should do a better job than words to explain how the top was fix to the undercarriage.

Once that was done it became time to bring in some strength to get this baby to it’s feet.

Now I have to flatten the top by hand and trust me it looks a lot flatter than it is in the pictures below. I hope to complete this task before the end of the year.

My second commission – part 10


My dear reader, I would like to apologise for my extended absence from the wonder world of virtual woodworking via the internet. You would find the reasons quite boring so let’s not waste any time nor effort ruminating on such drivel. This instalment of an apparently mammoth series will concern itself with the addition of the third and final layer of the so-called trapezoid leg. You can find earlier posts in this series here.

Seeing that the third layer would ultimately close up the internal workings of the whole construction, I took the opportunity to unscrew the second layer’s three ‘cross members’ (for lack of a better term). As you should be able to observe in the photos below, the old school mild steel wood screws received a coat of beeswax. This was accomplished by melting a block of wax in a small tin containing these traditional fasteners. The idea with this is that the wax should reduce the effort required to seat the screws and at the same time providing a layer that would resist future corrosion.

The screws were then seated after the surfaces that is supposed to be able to slide ever so slightly with the changes in ambient humidity over the years, were rubbed with beeswax. Whether this is useful (or possibly the opposite) I do not know, but I tried it anyway. Therefore I would urge you to ask someone who knows before following suite. Maybe some of our more experienced and properly trained cadres could assist in the matter.

Seeing that the plan was to fix the third and final layer using panel pins I had to fashion a custom punch to seat the nails below the surface of the wood. A short section of a round file which I picked up somewhere served perfectly well for this purpose. It was shaped carefully (not to take the temper out of the hardened steel) on a bench grinder to fit the head of the panel pin to a T. There are some picks further down to show the business end of my new redneck punch.

As is so common here in Africa, I also had to modify the panel pins somewhat to serve my purpose. In order to allow layer one and two to be able to move relative to each other, these panel pins had to stop short of layer one. In other words they should only fix layer three to the cross members of layer two. That was accomplished by snipping off the required amount, followed by resharpening on the bench grinder.

The two Kershout strips were fitted first, as they needed to be absolutely spot on given the fact that they mirror the spindles of the so-called Windsor leg. Kershout seems to enjoy spending time off  the Janka hardness charts (literally and figuratively) so it hard to say where it rates in comparison to better known species, but let’s just say it tends to take exception when a nail wants to upset it’s feng shui. For that reason I had to drill shank holes for each panel pin, which allowed the shank through and only caught the slightly wider head. This way the panel pins were more inclined to retain it’s linear configuration and the Kershout refrained from flexing it’s muscles.

As discussed in earlier posts, the third layer only needs to add another 8 mm for the trapezoid leg to reach it’s intended thickness of 44 mm. Therefore I decided to challenge my new bandsaw with fairly wide re-sawing in very hard Witpeer. Of course that also allowed me to introduce visual interest by means of a book-matched arrangement of the various pieces.

In order to do that I needed one flat, square and twist-free face side and face edge.

The resultant 8 mm stock were then fitted from the centre of the leg towards the outside. I again used the hitherto unproven technique of rubbing beeswax on the surfaces that is supposed to be able to slide.

I used a no. 78 and a no. 10 Stanley rabbet plane to cut the rabbets that hides the space allowed for movement.

The book-matched pattern is already vaguely apparent.

All the sides were then worked flush.

By hand plane along the grain …

… and by track saw followed by hand plane across the grain.

The small cavities created by seating the panel pins below the surface of the wood were filled with a concoction conjured up by mixing very fine wood dust (of the same wood of course) and epoxy.

Once the elixir had time to set I did a preliminary round of surface preparation.

As you can see the book-matched pattern is starting to emerge nicely. Once it receives oil it should be positively stunning.

Even the opposite side is starting to display a certain je ne sais quoi.

The edges were then treated to some hand beading to hide the laminations.

As you can see it worked a charm.

In our next instalment we will move on to laminating the various boards that was chosen (many moons ago) for the top.

My second commission – part 9


This post disappeared from my site while it went into hiding. Here it is again for those of you who missed it.

Welcome back to JNSQ woodworking. Here is hoping that we will continue to share woodworking banter and ideas in 2017. This past weekend was my first back in the shop and it was a real joy. Much like 2016, my two main projects to focus on will be this so called “second commission” and the table for the shebeen.


I will now report on the work done since our previous update in October of 2016. As usual just a quick reminder of what we are aiming for. A few shots of the model I built while developing the design.


Something that I omitted to illustrate in the previous post is the techniques that were employed to ensure that the spindles end up with zero splay. The first method makes use of a device that we shall call the Tambotie gauge (as I used a small Tambotie off-cut to create this fantastic piece of equipment).

The Tamboti gauge consists of an appropriately sized off-cut clamped to a square.

While reaming the mortises for the spindles you might remember how I made use of a stick with an appropriately sized tenon to check the rake angle.


The Tambotie gauge is used to check that there is zero splay by comparing the gap between the Tambotie off-cut and the spindle on both sides, as referenced off the side of the beam with the square. In this case the spindle is leaning ever so slightly towards the right.


Second of the two strategies again involves a highly complex jig that takes hours to build and set up. I clamped a winding stick to the side of the beam to check wether the spindle (positioned in it’s mortise) runs parallel to it, i.e. zero splay.


Once all eight mortises received this treatment it was time to test how the assembly would fit together. As you can see it came together nicely.


It is probably important to report on the stuff-up I made while drilling the pilot holes for the mentioned mortises. You might remember that I drilled one of these holes in the wrong direction and suffered from a Panic Attack subsequently. The solution I came up with was to turn a dowel of the same wood that fitted the hole perfectly and glued it into place. The hole was then drilled in the correct direction and the picture below show the result. There was only a small strip of the plug visible after drilling the new hole.


After reaming out the mortise there was no evidence left of the blunder on the surface that would be exposed to critical eyes once the tenon gets glued into position. In the pictures below you can however see the edge of the plug inside the hole. Eish, that was a close call. Woodwork has a way of keeping you grounded, isn’t it.


As these tenons run all the way through the beams, I decided to also wedge them. Here I am widening the mortise on the exit side to accommodate the wedges. I recommend reading Peter Galbert’s seminal work “Chairmaker’s Notebook” on how to orientate these spindles and wedges.


Next up I had to  camouflage the laminations with a few carefully placed beads before glueing up the leg.


I used my pre-1900 no. 66 Stanley beadingtool, which I restored quite some time ago. It takes elbow grease beading such incredibly hard wood, but it is very satisfying nonetheless.


I think it accomplished what I intended as the beam now looks like a solid piece of timber.


The tenons were then prepared to receive the wedges.


Made some wedges …


… and prepared for glue-up.


I used a combination of mallet blows and clamps to coax the spindles into position.


Once they were seated to my liking the wedges locked them down for ever (I hope).


This is how the Windsor leg spent it’s December holidays, resting on the assembly table.



This past weekend I continued my assault on the so called Windsor leg. I clamped it to the trapezoid leg and used the latter to mark out the final shape of the former. This way they are exact copies of each other in terms of measurements.


My daughter Aoife helped me to make the necessary cuts using my Miller’s Falls Langdon Mitre Box no. 75. It was quite a tricky operation given the awkward shape and size of the leg , but the Langdon made cutting the 9º angles straight forward.


One day the student will become the master.


The next big drama will be the third layer of wood that needs to be added to the trapezoid leg. I selected a good Witpeer (Apodytes dimidiata) board that ran pretty much through the centre of the tree and made a cut lengthwise along the pith. This gave me two quarter-sawn pieces. From these boards I then selected appropriate 800 mm chunks for re-sawing.  The idea with re-sawing is to created a book-matched pattern to the inside of this leg. This layer only needs to be about 8 mm thick to get the total thickness of the leg up to 44 mm, which fits perfectly into my  ratio of 22:44:66:88 mm (thickness) for the various parts of the table.


You will notice the two strips of Kershout (Pterocelastrus tricuspidatus) on top of one of the piles of re-sawn and planed stock. We will use those to create a type of depth confusion for an observer viewing the table from the Windsor leg’s end. This will hopefully enhance the effect of an construction that defies gravity, but you (and unfortunately I) will have to wait until the next post to see how this works or possibly not?? Here’s hoping (that it works, that is)!


A Table for a shebeen – part 2


The title picture of this post is meant to remind you of what my shebeen looks like. The beautiful ensemble of chicks in the picture demonstrate how happy they can be around a stove. For fear of mass feminist hysteria, I will leave it at that.

It is a case of onwards and upwards with this project, which reminds me of something I have noticed in terms of my woodworking journey. Since about July this year I seem to have entered the next phase in my woodworking. It took the best part of 5 years to set up my shop in such a way that I can build furniture using a set of tools that suites me. That entailed lots of fairly small projects such as building various hand tools and jigs. There were of course also a smattering of mammoth projects such as building the two Roubo-esque workbenches, but it the predominance of small projects made it easy to blog regularly.

More recently however I have started working on projects that is destined to spread it’s wings and leave the shop for good (once finished of course). These projects seems to be more of a war of attrition than blitzkriegs, which makes it harder to find ways to blog regularly. As my friend Jonathan White (from The Bench Blog) usually say (and I am paraphrasing here),  posts should rather be infrequent than of substandard quality.

OK enough of my introspective reverie and on with our topic at hand. Didi and I managed to finish the stage of inserting the monstrous Kershout Dutchmen along the crack in the top. They were worked roughly flush with my shop made scrub plane. The top was then moved to the shebeen to keep it from getting wet when the rainy season arrive. I took great care to set up the sawhorses in such a way that the top would stay wind free.


Luckily I checked with my winding sticks whether the aprons and stretchers were still wind free before continuing with the joinery. I found that they twisted slightly since being planed to perfection a few weeks prior. So back to the bench they went for some more remedial action. I find that my Lie-Nielsen low angle Jack plane with it’s toothed blade works well where you encounter significant knots like in this case.


The mortises were then marked out and the waste removed with a spade bit driven by an electrivore.


In this case the eletrivore was assisted by a carnivore, which is inevitable being resident in Namibia for several years. Vegetarianism (vegans are already extinct here) appears to be illegal (and omnivorous creatures should not get too cosy either) in this fair land and ‘vegetables‘ is certainly a most deplorable swear word.



The rest of the waste were chopped out using chisels and a hammer.


Next up were the two beams that connect the two leg assemblies.

This is what the shop floor looks like after serious hand planing (in 35 degrees celsius weather I might add).


Aoife came for a visit to the shop and took some photos. Her right hand is not nearly back to normal (after the drama we had in February), but managed to get a few in focus.


These long beams slot into each apron/stretcher with a type of cross lap joint. In the pictures below I am marking out the exact location of the slot in the stretcher/apron.


Followed by cutting the slot, …


… which allows the two parts to lock together.


I then drilled the holes for the drawbore pegs. As you might remember I like producing my own custom dowels for this purposed. I prefer using Assegaai (Curtisia dentata) for this as it has all the ideal properties. I use a Stanley no. 77 dowel machine with a ½” cutter head. In the pictures below I am using a ½” brad point drill bit to mark the exact location of the holes in the leg on the tenons of the stretcher/aprons. This mark can then be used to offset the hole through the tenon. Basically you move the hole through the tenon a smidgen closer to the shoulder, which allows the peg to draw the tenon into the mortise when it is driven home.


In these pictures I am busy chamfering the edges of the legs using two block planes. The one is tune to take heavy shavings and the other for very light shavings. The bulk of the timber is removed with the heavy cuts followed by a few fine cuts to leave a perfect surface. I simply run two lines parallel to the edge and plane to the line. Doing this task by hand is one of the most therapeutic in my experience.


My second commission – part 8


We finally arrive at the stage of this project that has generated the most anxiety for me. It will be my first attempt at using Windsor technology in any piece of furniture. As you know from the previous posts in this series I have completed the bulk of the so called trapezoid leg, so now we need to grapple with the so called Windsor leg. The rest of the table was inspired by (and leans heavily on) the work of George Nakashima, but this leg is entirely my creation. Having said that I am quite sure there would be similar designs out there, but it was not something I can remember seeing on the net during my research phase.

The idea with this design in general is that it suites the feral nature of the wood I have available. As most of you probably know, Nakashima was a master in the art of allowing the wood to speak for itself. The idea behind this particular leg is to create an elegant counterpoint to the otherwise rugged robustness of the design. It is meant to establish a sense of the top deceiving gravity. In other words, I wanted it to appear as if the top is floating in mid air. There will be a few more details added to the inside of the trapezoid leg to further enhance the mentioned effect, but for that you will have wait.

A quick reminder of what we are aiming for, a few pics of the prototype.


As you might remember the Kershout beams that will form the top and the bottom of the Windsor leg were laminated previously. My first step thus, was to square up a face and an edge by hand planing. I included the monstrous beam that runs between the two legs (at floor level) in this process as it has exactly the same dimensions i.e. 66 mm x 88 mm.


The electric planer then leached off my elbow grease to flatten the opposite sides of the beams.


The spindles presented the first major challenge. I did not want to laminate stock for the spindles as it would be very obvious in the final product. As I discussed earlier in this post these spindles also need to be elegant, which are the two main reasons why I prepared Kershout stock in the order of 1″ (25 mm) x 1″. As you can see the square stock was first planed into an octagon-shaped spindle, before it was introduced to the lathe.

The major challenge of turning these spindles was that I purposefully kept them ever so slightly bowed (lengthwise) as it would aid in turning more natural looking bamboos. Yes Bamboos. I want these spindles to look like bamboos to fit in with the Japanese theme of the table’s design, as well as the feral nature of the timber that was chosen.

That of course is no drama at all, but as anyone knows who’s tried to turn off centre work it is not a joke. You add to that the Kershout characteristic of being ridiculously hard and you have yourself a full-blown drama.


In the end I was forced to use 60 grid sandpaper to do the rounding as every other more conventional technique wanted to rip big chunks of timber out at the fat end of the bow (if you see what I am saying).


As you would expect, while rounding off the last of the four spindles I found that this particular one had an area where the grain developed a brief psychotic episode, which would have compromised the strength of it. Given the delicate nature of the design I could not risk any weakness to one of these four spindles, which meant that I had to start from scratch with fresh stock.



Once all the spindles were round and passed the grain-reading/strength test I marked out the location of the tenons and Bamboo’s nodes. As you can see I used the trapezoid leg as a guide seeing that the Windsor leg will mirror it’s shape. The angle of the spindles is exactly the same as the sides of the trapezoid leg i.e. 9º. The nodes were placed in such a way that the different sets of nodes (across the 4 spindles)  will form (partly invisible) chevrons pointing upwards. The chevrons are at 9º in relation to a line that run perpendicular to the spindle. Now that sounds very confusing, but it is the best my grasp of the English language will allow for.

Once all that was done I had to take the plunge and try to shape natural looking bamboos from spindles that is neither completely round nor completely straight. First, I very carefully placed the nodes using a skew lathe tool. After that I used 60 grid sandpaper to hollow out the internode segments ever so slightly. I purposefully made no attempt to shape them exactly the same. This was a deliberate ploy to have it looking more natural. Unfortunately attempts at making stuff look natural often ends up with it looking tacky, which was one of my main concerns. Luckily it came out looking reasonably good.

In this picture you can see how the slight bow in the length of the spindle made the nodes asymmetrical (for lack of a better term). The node is clearly much deeper on the left hand side than the right. This together with the slight inconsistency in shape and thickness of the internodal segments contributed to the natural look of the bamboos. Oh yes, that and the fact that the spindles are not completely round all over.


Once all four of the spindles were roughly shaped, they went back to the lathe for finishing (sanding with progressively higher grid sandpaper) and the shaping of the tendons.


At last it was time to use a set of hand tools I made back in January. I am of course referring to my shop made Reamer and tenon cutter.


As you can see here I again used the trapezoid leg to position the two beams of the Windsor leg before marking out the location and angle of the mortises.


It all went swimmingly as I first drilled pilot holes on the drill press to keep the spade bit that followed honest. Unfortunately I decided to drill these pilot holes out (with the spade bit) at the end of a very long Saturday. The concentration was going south and it ended up with me drilling one of the holes in the wrong direction!!!!

Enter: Panic Attack! #@!$^%*(&^&*%#@%^!!!!!

Seeing that this is not the first stuff-up I have made in the shop, I know that it is better to just stop working and try to calm down. In this case I added an ice cold G&T to treat the acute exacerbation of my Anxiety Disorder. After my rational mind retuned, it became quite clear how to fix the indiscretion. I had to turn a piece of Kershout to the exact diameter of the hole and glue it into place.

Let this be a lesson to all of us, stop working when you get tired (or gatvol as we would say in this part of the world).


The next day I was able to plane the fix flat and drill a new hole, this time in the correct direction. After that I got to play with the second of my two shop made tools, the reamer.


I drew a square around the hole before I started reaming. This helps to give a visual guide of the progress and alignment.


I tend to do only 3-5 turns with the reamer before scraping out the debris (that tends to clog the reamer) and checking my alignment.


I use this stick that has a tenon of the exact size and a straight shaft as a reference point for my bevel. That way you can make small adjustments as you go.


One done seven to go.


A table for a shebeen


Actually, from my feverish and extensive research through the length and breadth of this fair land, a table is by no means compulsory in a shebeen. In fact the only compulsory ingredient is a delinquent supply of Zamalek and/or Windhoek Lager. However in the few exceptional cases where it were found to be present, it was a very welcome addition.  So I thought my shebeen should have one, but it took a long long time to find all the ideal ingredients to build a shebeen proof/worthy table.

This project got underway in late July after I managed to secure a beautiful slab of wood from a friend who insisted that he did not want any money for it. In the end I gave him a Record no. 6 fore plane that I restored some time ago as payment. The slab of wood will make an ideal top for a table in Wamboland (our shebeen in the backyard).

Just to remind you what Wamboland looks like, a few pics.


And here it is, a massive slice of Eucalyptus. Unfortunately I do not know much more about the origin and specific species of this wood. I know it came from South Africa and measures up to 2700 mm (length) x 1000 mm (wide at one end and 1200 at the other) x 70 mm (thick). After a bit of research on the net I would say it is probably what is called Rose Gum or Eucalyptus grandis, but by the time it is finished some of our Australian friends will probably give us a better idea of the exact species.

I decided that this will be the underside.
This will be the side facing upwards.


I apologise for the poor quality of this picture due to the contrast in lighting. What I am trying to illustrate is how I marked out the longitudinal centre line on the underside of the slab. Seeing that it is wider at one end than the other, I measured out the centre at each end and tapped in a panel pin. A piece of string stretched between the two pins then helped me to mark out the centre line.


I used trammel points (not pictured) to divide the length of the slab into 5 equal parts along the centre line. The areas to be flattened for the legs were then marked out perpendicular to the centre line at the intersection between the 1st and 2nd fifth, as well as the 4th and 5th fifth. I hope that makes sense to native English speakers.

The first of these areas were flattened using my Festool router with the guide rail. I propped up the rail with bits of wood to stop it from bending. This was a very unpleasant experience. I have developed an allergy to electrical routers. They make too much noise and feel very unsafe to me.


The waste was removed by chisel and a slick (not pictured).


At this stage it was time to start the process of preparing the reclaimed Scots Pine, which will become the undercarriage of the table. I wrote a post on the origin of this wood which you can find here. These beams were first cut to length using my Disston no. 12 24″ crosscut handsaw and then ripped where necessary on the bandsaw. Unfortunately this was also the last job I will ever be able to do with this Chinese made 1982 model bandsaw, as it raised the white flag after an admirable career. Now I am in the market for a good affordable bandsaw, which is something that is probably impossible to find given the #Nenegate #Zuptamustfall #Nkandla #Failedtoupholdtheconstitution #Paybackthemoney #Khwesi #Itookashowerafterwards #eleventyseventyhundred #UntilJesuscomes-induced currency tragedy. RIP my dear friend, you have served us well (in reference to the bandsaw, not you know who).



OK so back to the top. The second of the areas that needed flattening had quite a big bump in it. I decided to use my Festool tracksaw rather than the router.

The bump should be quite obvious to see from this angle.


As you can see below, I again propped up the guide rail with bits of wood until the two winding sticks were lined up perfectly. This ensured that the two flattened areas would be in the same plane (or wind free if you like).


So in this case I made a series of cuts checking the orientation of the guide rail with the aid of the winding sticks prior to each cut.


I got lucky in that this area ended up “open” on one side, which allowed me to get in there with my shop made scrub plane to remove the waste (in conjunction with a slick and a chisel of course).


This is the slick I mentioned earlier. The actual slick handle were removed for this job, so I simply seated an old pairing chisel handle as I did not feel up to turning a custom one for such a quick task. It is a Peck, Stow & Wilcox (PEXTO) for the sake of the tool geeks (like me).


Seeing that there is a major crack in this slab of wood running all the way from one end to the other, I wanted to ensure that it does not fold like a book being closed when the top is handled during the building process. It actually feels quite resistant to such movement, but you never know. In order to lend it some extra support I screwed down these two heavy blocks of wood across the crack. I know, I know, possibly more screws than needed #JonathanWhite!! This is of course only a temporary arrangement, while working on the top. Eventually the undercarriage will fulfil this role.


Here you can see the legs, (some of the) stretchers and aprons after hand and machine planing.


The legs being cut to final length.


Hand sawing of joinery.



The pictures illustrate a technique that I employee fairly often now. I leave an area of waste which helps to stabilise the sole of my Lie-Nielsen router plane. Once the router plane have the area perfectly flat and to it’s exact depth I remove the waste and use the flat area created by the router plane as a reference surface for a pairing chisel.


For some reason I did not take a picture of the part with the waste removed. Sorry.


The next major task is to stabilise that massive split in the top Nakashima-style as approached from the side that will face upwards. For this purpose I had to laminate Kershout stock to created massive Dutchmen (aka doveltail keys).


Now they need a cosy new home. As I have described this process previously I will not bore you with it again. The pictures will probably do the job anyway.


Glued into position using slow setting epoxy.


At this stage I am in a race against time as this top is living outside my shop (due to it’s size) so I need to finish my Dutchmen-placement-activities before the rain comes. If we are lucky we might see a few drops around the start of October. The top needs to be safe and sound under roof in Wamboland by then. Then I can happy toil away at the undercarriage (which does fit in the shop) until it is ready to be joined in matrimony.

My Second commission – part 6


It has been a long time since I posted the previous update on this project, so here goes.

Maybe we should first just remind each other what we are aiming for. Below are two photos of the final version of my protracted design process in the form of a small mockup. In this post you will see how I progress with work on the solid trapezoid shaped leg and stock preparation/lamination of the various parts of the so called Windsor leg.


In the pictures below you can see how I cut dovetail keys from Kershout. I recently learned that the these keys which were made famous by George Nakashima are also known as “Dutchmen”. Why, I do not know, maybe Dutchmen like wearing bowties??


The keys placed on the face side of the trapezoid leg.


The Dutchmen is meant to stabilise natural cracks in the wood. Of course it also adds a certain je ne sais quoi, as you would expect.


I used slow setting epoxy to fix the Dutchmen about two thirds of the way into the thickness of the Witpeer panel. As you can see they stood proud of the surface of the panel by quite some margin (post insertion).


As a result of how incredibly hard the Kershout is and the amount of material to remove I followed the approach pictured. Several crossgrain cuts with a carcass saw, followed by chisel, followed by Jack plane.


The two edges of the panel were then laminated with the off-cut piece of Witpeer flipped over head-to-toe and back-to-front. This ensures the best colour match possible and per definition results in the grain running perfectly in the same direction. The lamination on the one hand aims to create a component that is twice as thick as the top of this table, which is a choice made in the interest of pleasing proportions (from a design perspective). It also creates space to hide a second layer of strips that run perpendicular to that of the first layer. This is quite useful when working with my beloved feral hardwood as it has a tendency to warp in the absence of something to keep it honest.


As you can see below, I used Dominos to keep the edges flush during glue-up.


After glue-up the edges were cleaned up with a hand plane.


Next came the preparation of the strips of wood that will run perpendicular to the grain of the trapezoid panel. This involved the usual array of tools including winding sticks, straight edge, and hand plane.


As this will ultimately be a layer that will be completely hidden from sight, I used proper off-cut level stock.


Preparing lapjoints  at end of the perpendicular strips.


Followed by the mating lap joints in the raised section of the trapezoid panel.


A useful trick that I employed here is to leave a narrow strip of waste intact to support the sole of the router plane while shaving 90% the floor of the lap mortise (for lack of a better term) perfectly flat and to the desired depth. Then it is easy to use the flat area as a reference surface for your chisel while pairing away the remaining waste.


The gaps you see in the lap joints are there to allow for ample movement of the trapezoid panel with changes in ambient humidity, while the perpendicular strips resist warping. You will also note the slots around each screw which allows for the same thing. I used my Festool Domino to cut the slots. At this stage the leg is ready for it’s final layer, which will cover up the perpendicular one and take the thickness up to 44 mm. More about that in a month or so as we now have to first build the other leg using Windsor technology!!! It will be my first time so I am stressing!!!

This picture was take before the joint was fitted perfectly, but does show the gaps left to allow movement.
Here you can see the lap joint after proper fitting.


At the level of the paint bucket you can see the Kershout I chose for the Windsor leg.



I can assure you that the time and effort it takes to liberate the few pieces of Kershout (aka Candle Wood or Pterocelastrus tricuspidatus)  pictured below from the pile pictured above will surprise you. This stuff is very hard and extremely heavy. It actually sinks in water!!! The Kershout is the deep red coloured wood.

I am sure it is quite obvious from this picture that even after liberation and some planing these pieces still look as wild as a pet rhino.


Below I am laminating the top and bottom beam of the Windsor leg.


In the next addition of this series of posts we will plunge into the world of Windsor technology.


My second commission – part 5


These Witpeer (Apodytes dimidiata) boards had a good 5 weeks to settle after all that hand milling we (yes “we”, we are a team now, I do the work and you get to read about it in the comfort of your own shop/home) did in the first part of April. They were clamped to each other to prevent even a thought of warping. It has therefore become time to glue them together to produce the first of three layers of timber that will ultimately become the trapezoid leg of this table.

In the photos below you can hopefully see my usual artwork on the face side. The boards were arranged with much deliberation on where the cracks and interesting areas of colour change should endup. Once that is done I usually draw a so-called carpenter’s triangle (sometimes referred to as a cabinetmaker’s triangle) across all of the boards to ensure that I can easily get them back in this order after planing individual edges.


I then marked out the location of the dominos I use to line the edges up during glue-up. Dominos are similar to biscuits, just better (as you would expect from Festool). I use the mark illustrated below to indicate where the slots for each domino should go. The ring around the end of the line is there to help distinguish these marks from others. I have found in the past that one can get confused with the lines of the carpenters triangle once the boards gets mixed up.


Here I am preparing the glue surfaces of the edges that go together. The idea is that you fold each board (on either side of the joint) down and away from you, much like you would do with a book were the cover is facing you and you close the book. Then you clamp them together with the two edges flush at the top. This way your error (if you go ever so slightly out of square) would cancel out for a perfectly flat joint once planed. You also want to create a very slight hollow in the length of these edges, which will result in a so-called spring joint. In other words, during assembly the ends of the joint will meet first and very gentle pressure from your clamps will close up the minute gap in the centre of the joint.


As a wannabe exclusive hand tool woodworker (AKA a hybrid woodworker) I have to admit that one power tool I really do enjoy using is my Domino. Why? I am not sure. It might be because it is relatively quiet, fairly accurate and very easy to use. Here I am preparing the small (3 mm) slots for the three dominos I decided on for each of these joints. I prepared only three edges and then glued them up one at a time. PVA glue sets so quickly in our dry environment that I simply cannot risk doing more than one at a time.


The next day the edges of those glued up panels were prepared and glued to each other.



The day after that the two bigger panels were then prepared and glued. As you can see it necessitated a different strategy to clamp in my bench(for edge preparation), but the bench came to the party.


I honestly cannot remember what one calls this thing that I made to stop the leg vise from twisting (when you clamp on only one side of it), but it works like a charm in situations like this.


For the final glue-up I had to first cut 1″ galvanised pipe to appropriate lengths as all my other clamps were too short. Luckily there were a few left in the stash that came from my father’s shop.


Didi assisted me to make sure this joint ends up flat as (this is a Kiwi expression for those of you who thinks that it is just poor grammer).



Once the whole panel was glued-up, I used a Festool TS55 track saw to cut out the trapezoid shape.


For the angle of the sides of the trapezoid shape my Stanley no. 18 bevel was simply set to that of the prototype/mockup pictured below.