Project Sideboard II
Having fixed the basic parameters of the design, I had to develop it incorporating a series of constraints, some of them self imposed. To start with, I am just one boy living in a one bedroom fully carpeted flat in inner city London. I don't have many tools or access to a workshop. I don’t have a car to carry things and even if I had one, my building is on a raised plinth next to a football stadium: security is tight and no private car is allowed to get any closer than a few hundred meters to the main door. As a result all the elements of the sideboard need to fit in the back of an uber or, ideally, be easy enough to carry in public transport, because I'm also poor. The design must also be such that the CNC machine can do the maximum amount of work possible, so the pieces arrive to my flat almost ready for assembly. This is probably the biggest design constraint since it means all the joints and pieces will be manufactured by this one machine alone, when your standard IKEA piece of furniture is probably made using 20+ machines in an assembly line.
To the above constraints I'm adding a self-imposed one: I want all the joints to be timber-to-timber: no screws, no metal angles. I have inherited a fascination for Japanese woodworking from my dad and the bit that I admire the most is timber-to-timber joints. I think the beauty about them is that you replace metal elements engineered and manufactured by others with your own brain designing how different pieces come together. I not only want my timber to timber joints to be there but I also want to express them as part of the language of my design. I want the finished product to be transparent and readable so anyone knows at first sight how it is put together. That's an advantage of timber-to-timber joints: you don't have to hide them. They are already beautiful.
Traditional Japanese timber to timber joinery. Gif by @TheJoinery_jp
As a starting point I have already modelled in rhino the 'form' of my sideboard, but now I need to split this form into pieces that can be CNC cut off sheets of ply, ideally from as few different thicknesses as possible. Doing this is slightly more complex than just dividing the piece in several boards and working out the connection, you also need to consider the sequence of assembly. In order to do that I have to continually drag pieces across in 3D and check if they can be put into place without 'going through' other components.
Once each joint has been designed it is prototype time. I cannot afford a full size prototype, but I have access to a lasercut that can cut thin ply up to 6mm thick. Using 3mm ply means I can build a prototype at scale 1:6 which will be enough to test that the joints work but also that I have got the sequencing right. In my experience, in and outside the office, it doesn't matter how many times you have tested a design in the computer, something goes wrong the first time you go on and build it. And so it does. When lasercut the first 1:6 prototype I realised I have designed the joints in such a way that I cannot assemble the back panel of the sideboard, which is quite key so that you don't see the mess of cables behind it every time you need to pick up a piece of crockery. Iteration two of the 1:6 prototype is a success and even looks cute. This has all the components of the sideboard apart from the doors. It's time to get ready to CNC the actual thing. Plywood and machine time are both expensive so, like in a spaceship launch, I can only really afford one go.
