You may have noticed an absence of 6 long rectangular boards and 2 large wooden ladders from the wood shop, and that’s because the bookshelf is finally finished! I’m writing this to not only outline the process of going through this project for the other members, but also for myself as a journaling exercise to reflect on and reinforce the things I learned on this project. To start this post off, lets have some before and after pics, along with a detail of the joinery:
Before:
Thanks @DuncanK for the help getting the timber to the space!
After:
Detail:
I started this project intending to deepen my experience with a handful of processes/techniques that I’d familiarized myself with on the prior woodwork I’ve done here at SLMS (jig making, mortise and tenon joints, routing, use of certain tools, etc.). There were also a few new related techniques that I wanted to challenge myself with, mainly milling, the use of laser cut templates and 3D design software. The rest of this post will be broken up into the stages of this project, and will include my thoughts and lessons learned along the way.
Design
This was by far the longest stage of the project. I’d had thoughts on how I wanted the shelf to look for months, but eventually realized for a project of this complexity, I’d need to create accurate drawings to be able to refine the design and work off of. I’d been putting off learning Sketchup for a while, and eventually bit the bullet and played around with it for a weekend; it ended up being a lot easier to learn the basics of than expected. The final state of this project looked a lot different than my initial plan:
What I was designing to with this one was something that would be technically challenging as well as make use of contrasting wood colours and complex joinery. As I rent, I also wanted to be able to take the shelf apart, but not have visible bolts everywhere; a couple of Youtube videos had convinced me that wedged tenons were the way to go. I’d also settled on beech and sapele, as they’re a bit cheaper than other woods and matches the desk I made last year. After significant effort and refinement, the above picture was the result. After soliciting feedback from a few people before buying wood, I asked a certain outspoken member for his feedback, and after some prodding (I could tell he was holding back, and I’m so glad I prodded) he listed out all of the reasons he didn’t like it. Now, I didn’t really care that he he didn’t like the way it looked, but what I found particularly valuable were the reasons why, because many of those resonated with thoughts that I’d had in the back of my mind, but hadn’t yet put words to. A particular piece of deep wisdom that was dropped was something along the lines of “who are you trying to impress, people or woodworkers, there’s a time for each”. Something was also said about the dangers of “Youtube woodworking”, and the fine line between elegance and clunkiness.
Lessons learned: don’t get so caught up in and attached to your initial design/idea that you lose the big picture. Ask others for their thoughts, but seek actionable feedback - “I like/don’t like it” is a fine opinion for someone else to have, but you should ignore when it comes to your design. What you want to takeaway from feedback is the why. And don’t be afraid to hear constructive negative feedback, especially for something you’ve been hyper-focused and potentially lost in the weeds on. Also, Youtube is a fantastic resource for learning certain techniques and getting ideas, but like all social media, it can become a bit of an echo chamber. Be aware of that.
Anyways, I went back to the drawing board and came up with a few more iterations. One key change that was made was actually using a calculator to find the deflection of beams under weight (https://woodbin.com/calcs/sagulator/), this influenced the removal of additional horizontal supports, which was something I’d been caught up on. From there I went on a couple of iterations, beginning to focus in on the idea of the shelves being simple horizontal planes, with the supports being differentiated in some way:
By the time I got to the final design, I’d dropped the breadboard ends and detailing from the shelves and dropped down to two vertical supports rather than 3. I’d also dropped the idea of using two types of wood, and settled on making it all out of ash:
Some other feedback I received that helped influence the final form of the project: A friend told me the prior angled look was too “hetero-normative” and needed to be less square. I’m pretty sure he was taking the piss, but that did play into the idea I’d been exploring on the differentiation between the shelves themselves and their support. What I ended up settling on was a simplified horizontal rectangular shelf, and (at least an attempt at) more subtly intricate rounded vertical supports. Also I hadn’t done “round stuff” with wood yet. Another friend mentioned to keep the context of the shelf in mind - it will eventually be covered in books and other objects that are colourful and shaped differently. As such, you don’t want the shelf to “overpower” the objects that are placed on it - I really liked this idea, it influenced the decision to remove the two tone wood and use ash. I removed the breadboard ends to reduce the complexity of the project, and removed the chamfers on the shelf ends. At this point, I started to consider the size of the wood that I’d be buying and playing with the proportions on the project. Having never bought rough sawn timber and milled it, there were a few unknowns that led to some excess deliberation and hesitation before actually getting the project started. After a call to the wood supplier, they said it typically comes in 3m lengths and is a bit oversized from the listed dimensions to account for warping and such. As wood is expensive, I settled on shorter 1.5m shelves to economize (two from each board). I also had the idea of having overhangs to allow space for other objects on the shelf. By making the shelves shorter and the supports closer together, I was able to drop the third support. I also worked out the detail I wanted on the joints, and thought through step by step how I was actually going to make the relevant cuts:
Lessons learned: ask a lot of people for advice! I don’t want to just do what other people like, but as above, I recognize that others have different perspectives that are very helpful in refining the design. Also, don’t be afraid to iterate. There were a lot of unknowns related to the purchasing of the wood and the finer details of the joinery that made it hard to start; I had some circular reasoning going on and found it hard to get going, but I had to start somewhere and eventually everything else came into place. Finally, its one thing to make something complicated in a drawing or software, and another to do so in real life. Simplification and repeatability were more important for me at my skill level, so it was worth it to simplify and think through each technique I’d use to arrive at the final shape.
Milling
This was a big unknown for me with this project, and I learned a ton going through the process of making my boards square. Before I could even get started, I needed to make some rollers to support the wood as it went through the planer as the boards were quite long. I also wanted to mill the boards while they were long rather than after I’d cut them to size. If I didn’t I’d have to spend a lot of time making parts the same size down the road. This was especially important for this project as I was going to need lots of components to line up pretty exactly:
I also made a set of winding sticks to be able to identify twist in my boards. These live behind the planer now:
I don’t have any pictures of the rest of the milling process, but I ended up getting pretty intimate with the No. 7 plane to flatten the boards as the planer was imparting a slight twist that usually wouldn’t really matter, but would over the 1.5-2mlenghts of my components for this project. I learned a bit about wood movement through the process, and realized after a certain point that they were never going to get perfectly flat, just flat enough. 4 of the shelves had a little bit of a bow, so I ripped and rejoined them to get a bit more thickness out of the boards. I ended up fussing with the boards a bit more than I probably should have, so they ended up around the thickness they probably would have had left the original bow in. Oh well. Here’s cutting one of the shelf halves to size, note the reuse of the roller:
Before moving ahead with cutting all of the milled pieces to size, I used resin to fill in and stabilize all of the small knots and cracks.
Cutting to size
Now that I had (functionally) squared boards, it was a matter of cutting them to size and preparing to make the mortises and tenons. Here, I’ve cut all of the short vertical supports. I didn’t use a proper stop to get them all exactly the same size, but I solved for that by adding shims(taped on squares of plastic) to the shorter ones to make them all the same length. This way, they’d all register against the router table fence in the same way so the distance between the shoulders is the same on each one:
I tried something new and cut the tenons on the router table rather than by hand or with the band saw. I did it this way as I’d need to make 24 tenons that were as similar to each other as possible; I wanted to do it in a very repeatable fashion. I had to do some setup with the router table fence and made a guide so everything was set to be as square as possible. Once set up, it worked like a charm:
The next step was to make corresponding mortises. to do this, I measured my tenons with a caliper and created a few templates on the laser cutter. I used some double sided tape to attach the template to the beam, then registered to the template on the mortiser to ensure they were all the same size. I did make a mistake here and cut on the wrong side of one of my lines - the second lowest shelf went a bit higher as a result. I think it’ll look fine, but here’s a reminder to mark which parts to remove with a little X in the future!
Next up was to dry fit the “ladders” and do a bit of fine tuning with a chisel. Here I took the opportunity to do some grain matching and labeled the pieces in the orientation I wanted them in:
After disassembling, I drilled the holes for the bolts on the pillar drill, marking the points to drill using another template. If I were to do this over, I’d have set up a jig on the pillar drill - I think it might have been a tad more accurate and repeatable.
I then rough cut the bottom sides of the horizontal supports on the band saw. This was done at this point as to make the eventual router template cutting process easier. I used the same template to mark where to cut to as I would use when actually routing:
Before the glue-up, I routed the outside curved edged of the ladders on the router table using a roundover bit. Here’s everything before the glue-up:
Finally, here’s a shot of the glue-up. If you look in front of the blue roll, you can see another of the templates:
Lessons learned: Try to make as few absolute measurements as possible. For example, it was more important that the support beams were exactly the same as each other, not necessarily 254mm each - so in that case I set up the router table once, then cut all of them in one go on the same setup and checked along the way. Or to mark where the mortises should go, I aligned all of the vertical beams together, measured one point per shelf, then used a square to transfer the line to all of the other beams. Also for complex glue ups, practice and do as much set up as possible ahead of the actual event.
Routing
Once the ladders were glued together, it was time to get going with the routers. The first cut I had to make was to squarely trim away the material that wasn’t removed by the rough cutting mentioned earlier.
Once this was done, I needed to add the interior roundover to the ladders. I made some test cuts and made some stops that I taped into each frame to make sure I ended the cut at the same location for each one.
Next was to cut notches into the shelves so they’d slot into the ladders. I was going to do this with a laser cut jig, but @ryanf offered the sage advice to rout them all at once by clamping them together:
After this, I dry fit the shelf, finalized the shelf orientation, marked and drilled the holes that the inserts would fit in, and did a bit more fine tuning the the chisel. When dry fitting, I used some shims to centre the shelves to allow for wood movement:
Next was to rout shallow ~1.5mm channels in the shelves so they’d slot into the ladders, hide the meeting surface, and get the tops of the shelves to align with the bottoms of the routed edges on the ladders:
Lessons learned: move fast and confidently on the router to prevent burn. Also make sure to lower the speed for wider bits. Take more passes - the time it takes to make another pass is less than the time it takes to sand out a burn. If there isn’t enough to register the faceplate (fence?) of the router against, make another one.
Final assembly and finishing:
I did another dry fit and made this little jig to help get the threaded inserts aligned a bit more squarely:
Sanding was eternal on this one, but what I’ve learned from my other projects is to
- use dust extraction
- work your way through the grits
to save time and end up with a nicer finished product. I found some of the burn marks to be difficult to remove - I experimented and reapplied a trick I learned from @giles to help with easier burn removal. Spray the burned area with a bit of water - the wood will absorb the water and raise slightly, allowing for easier removal of the burn via sanding. For the finish, I used a product called Osmo Raw. It’s an oil/wax that has a bit of white pigment in it. I arrived at this one due to some research leading me to believe that using a normal oil would yellow the ash a bit more than I’d like over time. Here’s the final dry fit in the space before @Brendon_Hatcher and @Ben_Mahon helped me take it home (thanks again guys!):
Another tip when using metal hardware - make sure to use the same metal and finish for different components that will be touching. For example, don’t use a stainless steel insert, a zinc plated washer, and raw iron bolt together - there’s a potential for them to facilitate the oxidation of each other.
To end, here’s the project information I engraved onto one of the support beams:
I’d like to give a huge thanks to the many members at SLMS who offered help and advise along the way. (@DuncanK @joeatkin2, @ryanf, @kyle, @slimshelly @lewisss, @chrispyduck, @giles and many more) I would not have executed this project as well as I feel I did without your support. I also know that I would not have even attempted to do this project without the experience gained from the other projects I’ve done over the past year at SLMS.
In summary:
MAKE JIGS
SHIMS AND LIGHT MALLET TAPS ARE YOUR FRIENDS
THINK THROUGH YOUR CUTS
DON’T BE AFRAID TO START
