branagan custom

small shop hand tool woodworking

Project: Building a Workbench: The Nicholson (2)


|


The Trestles

First I measured and crosscut the 2 x 4s to laminate legs into 4x4s. 

I cut only the leg sections to the height (length) I would need them after drawing up a quick sketch to get to my planned bench height of about 30”. I had 8 2×4 lengths that were almost the same – I could square up the ends after laminating. 

I focused on just flattening one face on each, prepping them to glue together.

I tried to face the growth rings back to back (curving away from each other), and ran the grain in the same direction if possible to make flattening the side with the glue seam easier.

I planed a slight cup along the boards, and the slightest bow so the joint was just a little bit sprung. For the most part I got a clean, full contact joint, with slight gaps in one or two corners but I planned my joinery to either cut out these gaps or work across them.

The process was to plane the glue surfaces then glue and clamp them together overnight, one by one.

A nice seam on one of the legs

When all four were complete, I assessed the faces and chose joinery locations, and which would face out. I also decided which legs will pair, and face what direction, and marked these in the top end grain so I knew which faces to flatten and square.

In this build I’m prepping only what I need to, to get through the build more efficiently.

I worked on planing one face flat and straight – the face that would be exterior on the assemblies. The joinery will run across the glue seam so one full face of a 2×4 will face outwards. 

I took this part seriously, and I had my winding sticks and straight edge to hand the whole time. I started by assessing the face – checking for high spots with the straight edge, and whether it had a dip or hump along the length. I knocked down the high spots with a coarse jack plane, came back with the straight edge, and checked with the winding sticks at this stage. When it was getting close I switched to a finely set jack and checked straight, flat and twist after every few passes. Since this was a joinery reference face I aimed for completely flat and dead straight.

When I was satisfied I assessed the edge to be squared to the face, first checking how out of square it was and then the high spots with this in mind. I squared the edge that will receive the joinery to the face. Again I started to work with the coarse jack plane, then switched to the fine when it was getting close.

From this process alone I had some lessons learned:

Sharpen often. A fresh edge makes the work immeasurably easier.

Pay very close attention, especially for your joinery faces. I had some spots that were very low, and didn’t want to lose the material from the full face trying to dig them out. A low spot like this can show out of square in just this area, you have to be cognizant that it’s the low spot causing the reading, not an out-of-square edge.

So I worked the edges to flat, straight and square with the faces.

This process was just backbreaking on my setup at the time. Workholding was a constant struggle with my low bench, which was built to fit my old shop (a balcony), hence why I am building a new bench. My bench is low, light and short. I use a planing stop but these legs are too long to sit on them on the bench, so I had to hunch over it, and when I hit a tough spot like a knot, the whole bench skidded across the floor. 

I butted the bench up against a wall and used simply a clamp on the side as a planing stop, to bring the work closer to me. When I needed to work the close end (which was hanging off the bench – it is short), I moved the leg to the far planing stop just to hit the near end. 

“Planing stop” might be a stretch, this consisted of two screws driven into the benchtop.

Also, one other lesson I learned is being aware of the space around you, and what you put where.

The most upsetting moment of this build so far is putting my Veritas straight edge on a stool behind me, squatting to sight down my winding sticks, and knocking the straight edge off the stool onto the concrete floor. Twice. The corners got dinged and I am simply praying it isn’t thrown too far out of straight and completely useless. Nevertheless, I don’t want to spend the money to replace it so I use it and the results will speak for themselves.

Have a spot to rest the tools you’re picking up regularly – out of harm’s way. Harm, in this instance, being my arse.

With a reference face and edge, I squared off the bottom with a plane and chamfered the edges. Then I measured from the bottom to mark them the same height. They were close enough so it just took some finessing to plane them to the same height and square.

Now they’re ready for joinery.

Before that, I had to cut and prep the corresponding stretcher boards. I used 2×4 on the bottom stretchers and here is where I made a big mistake which came back to bite me later on. I used 2×8, the same material for the top, for the top stretchers on the legs. This would be fine, and really it is fine to do, but it does cause problems. I did not catch this in the build that Richard ripped down this board to make it narrower.

I prepped them the same as before with a face and edge, and got ready for joinery.

It makes a lot more sense in this project to cut the male portion and mark directly off that. It’s also better practice to mark out one component precisely, and mark the rest off of it – specifically the shoulder locations which would determine the internal length of the stretchers, and thus the width of the leg assemblies. It took some calculation here to figure out the final width of the leg assemblies when I took into consideration the actual maximum size of the top, allowing for material to be removed with planing, and allowing tolerances like overhanging the apron slightly, and overhanging the top to the apron and planing it down to meet.

After calculating the overall width I was going for in the leg assemblies, I offered the legs to the stretchers to get the approximate shoulder lines. 

I took for granted that all legs were the same width, meaning the same length of half-lap tenon. So I used one leg for all of them, which was wider than others. I didn’t precisely thickness the legs so some were thicker, some were thinner. The shoulders were exactly the same which meant that internally the trestles were dead parallel with each other but because of the discrepancy in leg thickness, the outsides were the slightest bit out of parallel. This gave a taper of about a quarter inch over the 5ft along the bench,

So after using the leg to mark the shoulder locations, I had the shoulder-to-shoulder length of the stretchers. Using a square and knife, only from my reference faces, I marked out the tenon and dovetail, and used a marking gauge to get the depth. I went halfway through the stretcher, and kept the marking gauge as is for marking the female portion on the legs.

At this point, you could technically mark the shoulder-to-shoulder width on your other stretcher boards too.

I chose to complete the full joints one by one. The first one was a learning process so I chose the far back joint to complete first because in the worst case scenario it would be completely hidden. This was a simple operation; saw down to the shoulder line and using a chisel and mallet split down close to your lines, then finesse from the lines. During this process I did change my approach. Maybe this is the first time I realised that you don’t have to strictly follow exactly how the “teacher” does it, you can find your own way to get to the same result. It’s not about mastering what they’re showing you, there’s no point in punishing yourself to do it exactly as how they do it. Specifically I struggled paring the layout lines flat across the tenon. Maybe because of my setup, but after the first one I knew I had to find a better way that worked for me. 

I find it much easier to get square across a face by sighting plumb rather than straight on (vertical instead of horizontal), and I can’t tell if I’m shooting up or down with the chisel. So when I split off most of the waste, I flipped the leg on its side, with the length running away from me, sat the chisel in the layout line and sighted plumb before knocking out the last piece of waste. I had much better success this way, even if it took slightly longer with the change in position. I did this in all directions. Once I had established the floor of the joint square to the side, even just a half inch, I could work from either side to remove the waste in the centre using the established outer edges as a reference. On end grain I didn’t prop it straight up but started it from the plumb position coming across the side, made some relief chops about a half inch from the end to prevent it splitting too far, then worked across the end grain in the layout line. Anywhere visible should be as exact as possible but I wasn’t too concerned about undercutting the floor of the joint just slightly. It’s better to undercut it so it fully seats, than leave humps that will hold it off and rock it around. However, it is obviously best to flatten and smooth it for 100% contact!

Getting back to the process, I split off the waste, worked across the layout lines, did that for both ends of the stretcher including marking and chopping out the dovetail, and was fairly pleased with myself. Then, I decided on the height of the bottom stretcher. I marked the height on both legs with a knifeline for accuracy and lined up the dovetail with this line. You have to press hard to seat the shoulder, keep it square, and flush against the leg face. When I was happy, I used the knife to mark the line of the dovetail, trying to just sneak under it to ensure a snug fit. 

Now that I think about it, to avoid any fluctuation in where the top of the joint lands, you could use Paul Sellers’ approach and work on the top line first, chop down and chop out the waste, to establish that line to depth, then present the tenon to the line, squeeze it tight, and mark the dovetail from there.

Once this was marked, it was just a matter of deepening the knife lines, squaring them down the sides and using the preset marking gauge to define the depth. Then, saw down the lines, chop out the waste, clean up the floor and the sides, and test fit. All but one of the dovetails was too tight off the saw so I had to take a few light shavings off the top of the stretcher to seat it fully. I completed one dovetail before moving on to the next one, and once that was done I moved on to the upper stretcher. The process was the same but I had some difficulties with knots and wild grain so they weren’t so clean. Never mind – this will never be seen under the bench.

To mark the dovetails on to the legs I assembled the bottom stretcher with the legs and slid the top stretcher into place, squeezing tightly to pull the shoulders tight, flush the tenons against the leg faces, and position the height exactly before marking the dovetails with a knife line. 

The bottom stretcher came together well, the joint lines were tight and shoulders were looking well but the depth (or possibly thickness of stretcher and/or legs) was slightly uneven. One was slightly below the surface and one was slightly above. Again, not the end of the world once it’s all square.

At this point I should mention do not use stock the same width as your bench top stock for the stretcher. Use narrower stock to avoid my problems!

So after splitting off the housing for the top stretcher’s dovetails, they came together quite nicely but there was still a long way to go. I had overshot how proud to leave the top stretcher above the legs and would have to plane it down closer before attaching the top bearer. I also realised at this point that the 2×8 stretcher would cause problems.

It would fall lower than the apron, so I will have about an inch of end grain visible on the outside face of the leg where the apron meets the leg.

I knew the next step was to mark the shoulder for the housing to receive the apron. Unfortunately poor planning struck here again. I planned for the apron to slightly overshoot the top of the trestles, marked the horizontal line, and measured up an inch to mark the angled shoulder. I didn’t realise that I should have been taking into account the material that was already removed for the top leg stretcher, so the shoulder ended up being way too low, when the extra width of the stretcher was taken into account. I realised this too late. I cut the full housing and shoulder, and when I assembled the trestle I realised that inch of end grain was sticking out to the outer side of the leg. I knew I would have to trim it to the depth of the apron shoulder but I had to leave a little piece sticking out which would be visible, then mark and carry the angled shoulder across that piece. This pushed the apron shoulder lower. At first I thought about working with it, and planing the bearer on the top of the trestle down to the apron height, but instead of losing height off the top I opted to add little spacer blocks to lift the shoulder height for the apron. Hopefully a visual will make sense of this mess.

At this point I assembled the trestles. I bought 2” wrought square cut nails from Lee Valley. This was my first time using cut nails and a little nerve wracking considering the amount of work that went into these components and the possibility of things going horribly wrong here. To be honest they went about as bad as they could go but everything went together and was usable. The first assembly, I didn’t drill any holes and tried to just knock them in (like Richard does). I’m fairly certain my nails are a lot thicker than the ones he uses so I did actually have some trouble knocking them in deep, and some would just not pound below the surface with a nail set. Luckily this is the “back” trestle which has been experimental for the whole build. The bottom stretcher locked in tight, so did the top stretcher but being exposed on the one side meant that it did split to the end grain where I nailed it. It’s still glued on and I partially blame one nasty knot and wild grain. I was a little upset that the leg actually split beneath the stretcher as well but hopefully it’s not so bad to cause an issue in the future. 

Lesson learned for the second trestle: drill clearance holes. And countersink. The second trestle went much smoother with these two changes. I knocked them together and got the top bearer ready. I waited until this point to measure the length of the bearers and gave them a quick planing to make sure they weren’t super wavy or twisted. I glued and just screwed these on since they will be completely hidden below the benchtop, and why waste nails and risk more splitting unnecessarily.

Once the bearers were attached I checked the height from bottom to top on all four legs, and did have to finesse this a bit to bring them all into one plane. I figured it’s much easier to take down 1/8th from a corner now than later on from the bench top. Then I moved on to extending the shoulders that will receive the aprons.

I used ½ inch plywood offcuts, finessed the shoulders so the blocks would fit snugly, and glued them on. I left them over night then trimmed the ends, planed them flush with the legs and took about an eighth of height off them so the apron would fit to just above the trestle height, as originally planned.

I gave some surfaces a smooth planing if they were dinged or dirty but wasn’t too precious about every surface being perfectly glassy smooth.

Now it’s time to move on to the aprons.


Leave a Reply

Your email address will not be published. Required fields are marked *