While waiting in queue for the high-demand machines, I took advantage of the digital workshop tech’s absence by running some tests with the laser cutter. I ran a raster-etch several times until it was deep enough to create a groove in my wood. I then cut the outline and tested how well it would hold up to bending; the result was good.
I tested wide, shallow grooves, with copper shim – slightly thicker and stronger than the copper tape I’d used beforehand. I also tested thin, deep grooves for wire cables. The challenge would be to bend the wood with the conductive material glued inside with neither the wood or the circuit breaking.
Both tests were successful in terms of flex and conductivity, though the cables were so close to the outer face of the wood, that they showed through like veins. I decided this was not an aesthetic I’d like to embrace, so opted for the copper. I would pre-tin the ends of the copper and solder some wires to the ends to make final connection easier.
Having confirmed the etched groove size, I cut all of my stem pieces and spent about 12 hours sanding them. First, I removed the burn marks, then I thinned out the horizontal sections that would need to squeeze tightly into the base.
When the water-jet cutter was available, I was able to prototype my new sprung collar design. Because the material was so thin, I had to stick it to a board with double-sided tape. The speed that the cutter sliced through it was astonishing.
The copper was too pliable to work with the open-leaf design. My next plan was to adhere it tightly to the steel. Because I was working with 5 thin sheets, glueing them together in consistent alignment was difficult.
Before getting too ahead of myself with the aesthetic properties, I was excited to test the mechanism to see if it functioned. As you can see in the video below, it did not perform as hoped, and while it was usable, it did not improve on the usability of the very first flat-collar prototype.
By now, I had to make the decision to move forward with what worked. I’d spent a lot of time experimenting, but was not able to improve on the original mechanism. To that end, I made a layered compound of copper and birch and cut two simple flat collars from it.
While the water-jet cutter was available to me, I was able to finally start cutting the terrazzo for my final lamp. Because the button needed to be much shallower than the rest of the base, I tried slicing into it, longways. There are no ‘terrazzo’ settings on the water-jet cutter, so we’d been guessing between granite and marble. On shallower cuts, this had worked fine, but for such a deep cut, the discrepancies really showed through.
Rather than taking up more time on the water-jet or waste more terrazzo, I found my way upstairs to the glass and ceramics level of the school and was inducted on the use of the wet sanding discs. I sanded the terrazzo til it was flat and even.
Meanwhile, the base insert had been cut by the CNC router. I was prepared to fix any problems manually, because I’d never prepared an Enroute file and seen it cut before, but it came out perfectly.
I had to build up the height of my button switch, because the springs were stronger than anticipated. When I was happy, I soldered the switch to the base wires. I was also able to test that the side-wrapping veneer fit and was very happy with the result. I was glad that all my planning was working out.
I laser-cut some cork for the base, to allow it to slide across a table surface with ease and no scratching. Because the largest sheet I could find was too small for the base, I had to use two pieces. I didn’t want to have a straight half-way line, so I made it into a sort of yin-yang pattern. It’s much harder to clean glue and dust from cork. Luckily, this is the least likely part of my lamp to be seen.
I pre-bent the base-sections of my birch, so that when joined with the terrazzo, it would provide least resistance and possibility of breaking.
I slid the collars onto the stems with great difficulty. I’d made them just tight enough to create friction along the stems, but just loose enough to fit over the splayed tops. I tried testing their efficacy, but without the bases glued firmly in place, it was difficult to gauge a true result. With blind hope, I started the process of bending the top half of my lamp, in the same way that I had with my earlier prototype.
I measured the section for twisting against my 1:1 drawing and over-twisted to get a 90 degree turn. I had to try this several times to get it to bounce back to a right angle.
The head mould was larger and more contoured than my original prototype, and I hadn’t allowed enough length for the ends to wrap all the way around. I didn’t feel great about the bob-haircut I’d given it, but rather than re-bend, thought I might be able to fix the issue once I’d inserted the lighting.
I soldered the three sets of fairy lights together in a series and applied heat-shrink tubing to protect the connections. I soldered connector socket and pins on both the insert as well as the main lamp body. I then tested the entire circuit from the power source, through the switch, up the stems, into the the lights and back again. I was stoked when it worked. I fixed the head insert to the body using mounting tape and tidied up the haircut in the process.
I was really stressed about cutting the main pattern for my base. I would only get one chance at this. Miraculously, it worked, with no issues from the water-jet cutter.
I’d sanded the birch to fit into the 1mm cuts created by the water jet stream, but it was still too tight. Because the cutter left the stone jagged, there was a lot of friction for the birch to deal with.
I took all the pieces up to the stone sanding room and sanded down the outer curves of each. I also had to trim a piece of the birch that was causing interference. After a lot of patience and a few broken pieces of terrazzo, I was able to make all the pieces of birch fit inside. The fit still wasn’t perfect, so with the advice of a peer, I used a ratchet strap to tighten the overall form. This worked really well and gave me the confidence to start gluing.
I wrapped a bunch of masking tape around the outer edge of the central button to help press the birch into place while allowing enough clearance for the button to move up and down freely once the rest was set. I glued all the birch pieces, except the outer veneer, using a 24hr strong-setting epoxy glue. This gave me plenty of time to work with it. It also gave me the time I needed to take the prototype home and clean the glue off the top of the terrazzo with acetone. This would save me having to sand it off later and keep that clean, smooth face I started with.
When the glue had set, i took the whole piece over to the stone sanding room to smooth off the edges and get it to fit within the birch veneer wrap. This was really satisfying.
Once it all fit together, I soldered the cords from the stem to the cords in the base, taped the power socket in place, and aligned the whole thing with the button sitting easily in the centre. I then applied my slow-set glue to join the base, terrazzo, and veneer wrap. I tightened it together with the ratchet strap, a couple of clamps and some pieces of dense foam – the trouble in keeping the wrap tight while allowing the power socket to protrude.
Finally, I sanded down the top edges of all the birch at the base, sanded the scuff-marks off the veneer wrap, applied wax to the birch, polish to the terrazzo and sanded it all down to a smooth-feel finish.
This was the first chance I had to properly test the collar mechanism. It worked, but wasn’t stiff enough to last long periods of time. As it was too late to take off the collars for correcting, I applied some more carnuba wax to the inner stems. This time I didn’t sand it back smooth, but let it set hard and rough. I think this helped a little, but it’s still not perfect. I’m at a loss as to how to correct this problem in a low volume scenario. I imagine a thin coating of clear silicone to the inner stems might be a possible solution.
I turned my living room into a studio for a day and spent a very long time trying to get the lighting right. It’s very difficult to make it look like the main light source is the lamp itself, while also casting enough light on all the details of the lamp. I also had to move my lighting for each articulation of the lamp, to maintain this facade.
I was fairly happy with the resulting photos. It was difficult to show off the detail of the base and the detail of the head in the same shot. Shooting the terrazzo in the same lighting style of the main photos was very difficult. I did a lot of editing on the lighting around the base to make it look more natural. I also created an artificial background. Some photos work much better than others, but they all show the lamps true form, which is all I really needed to do with them.
I’m almost finished with this assignment. The final step is to work out manufacturing and environmental implications and present the design as a poster. I’m planning to auction this prototype off to family and friends because, as wonderful a learning experience it has been and as proud as I am of it, I will always be hyper-aware of what I could have done better.