Think big. No, Bigger

Makes something big. No small task. Apologies for the pun.

In addition to merely making something of size, this week focussed on an introdcution to CNC machining on a macro scale, as opposed to the circuit boards that were made previously. In this regard, I had a fair bit of experience going in. I'd taken 2.008 which is a class that focusses on CNC machining of aluminum blocks to make tooling for injection moulding. I had never before worked with wood on a CNC though, and the Onsrud was a new beast when compared to the protrack I'd used before.

Inspiration for a project came in the form of a youtube video sent to me by a friend. Designed by Sebastian Errazuriz is a highly impracticle piece of furniture that is absolutely gorgeous in its function. Absurdly complex, I decided it would be a nice way to test how much I could (or couldin't) do.

First thing's first, I had to figure out how the cabinet worked. By studying the following images taken from the video and online, I was able to determine that a peg and groove system was used to create the staggared tag-along. The bottom appeared to be a long rod though a loose finger joint.

Assemblies would be an important means of evaluating whether or not the components were working together as designed. Solidworks, again, was my choice of cad program. Below is an image of the partial assembly as the cabinet was intednded to look. A variety of materials have been selected to emphasize where components meet, though colouring would be homogenous in the actual cabinet.

I exctracted the parts from Solidworks using a DXF, ran them through Rhino breifly to create layer separation. This layer separation was used in Mastercam (shown below) to select for the different tool paths. This was a lesson learned the hard way, as the first draft was selected by individual toolpaths, which ended up numbering greater than 400.

The Onsrud is an intimidating machine. Observing it while it cut the other pieces, I was amazed by the rate at which it accelerated the build plate. It was unfortunately at this time that I learned one of the big differences between routing would and milling metal. Namely defining the stock.

I wrote my Mastecam code referencing the top of stock as the "zero" expecting there to be a touch-off where the zero was defined, and then the rest of the process would proceed. This was not the case, and the Onsrud proceeded to drive the router bit into its self as it plunged well below the stock and into the sacrificial layer.

It was unfortuante that this was not caught earlier, but I redifiend the toolpaths, and this mitigated the issue when I came to cut again the next day.

The Onsrud holds the board to the build plate using a vaccum system. The vaccum system relies on most of the sacrificial layer being covered by the plywood, so an onion skin system was used to hold the pieces to each other. Nominally, this would miantain the vaccum while the perimiters were cut, and hold the pieces in place.

Unfortunately, this system didnt work with how much material I was removing and the poor quality of the plywood that was used. The vaccum was lost, several pieces vibrated free, and eventually the entire board was jarred from the bed and I cancelled the job about half way.

There were not enough peices to do an assembly, and not enough time remaining after 3 days in the shop to do any further cuts. I must therefore sign off with many lessons learned, but nothing big to show for it.