We cut our test-pieces from 4mm plywood from the local timber merchant. Now we’ve moved beyond tests and into the realms of cutting the real-thing, we need to look at what material to use.
The floor that they’ll be laid on is new, and flat – so we don’t need super-thick tiles. We eventually settled on 6mm MDF, with a veneer of oak on the top. For our tests we had been using Plywood, but we found that it had a tendency to expand (only slightly) when cut, and round-off the edges. MDF, on the other hand, kept its shape and definition extremely well – and was a tiny bit cheaper too.
The wood is sold in 8 foot by 4 foot sheets – the timber merchants cut each sheet up into five pieces, of dimensions that would fit onto the CNC. We can get two lizards on each piece, and we need 200 lizards. So: that’s ten lizards to a full sheet – twenty sheets of wood in total, at a cost of about 550 pounds.
Credit should go to my Dad here, who did all the hard work cutting it all. He apparently got into a routine quite quickly – lightly sanding the burrs off a cut lizard whilst waiting for the next one to be cut. But it was still a full-time job – so you can understand why a professional cutting company wanted to charge 2,000 pounds for the labour. I’m afraid my Dad will have to be content with a meal at a posh restaurant …
It’s worth mentioning our experiments in shaving off extra wood, to increase tolerance. When cutting “final” lizards, we found that the shape was so accurate that fitting them together was extremely tight. It required quite a lot of violence, and both pieces had to be completely horizontal to go in.
So it seemed worth finding a convenient way to buff a bit extra off the edges, just to make laying it easier, and to add more of a border around each lizard. We achieved this by modifying the radius of the drill bit that the G-Code generating software thought we were using. We were using a 1.5mm drill, but we told the software that it was only 1.45mm – so the path generated was slightly too close, so a little bit extra was removed (consistently) around the shape.
After a few iterations, we found that setting the drill radius to about 1.2mm had a pleasing effect. Which sounds like a big deviation from the 1.5mm drill, until you consider that this is still only one third of a millimetre!
For me, this is yet another reason why getting the CNC and doing it ourselves was better than paying a company to do it. They would have just cut the lizards to our specifications (which were unproven) and we would have spent nearly 3000 pounds on lizards that didn’t tessellate! So either I would have spent a lot of evenings buffing the edges on 200 lizards, or I would now be burning 3000 pounds’ worth of firewood!
Here are all the cut lizards – stacked in my dining room, being left for a fortnight whilst they acclimatise to the room:
[Part 1] [Part 2] [Part 3] [Part 4] [Part 5] [Part 6] [Part 8]