This weeks assignment is to make something Big. We will Learn about using the Computer controlled full size shop Milling machines. At our lab these include the 3 axis - ShopBot and the - Precix 11100. The aim is to design something for construction in sheet wood.
The order bill for material went out early for this project and the board size was 2500 x 1200mm at first the options were water-resistant OSB orientated strand board - or flakeboard in 15mm thickness or the more expensive Plywood board in the same dimensions
I had to think fast what to make and as I had been modelling some similar things when practicing 3D software, I began to be drawn toward Dome type shapes and structures. I was also looking into tensile structures with wires and taught flexible materials at the time but conceived that for this project I could make a rigid skeleton of interlocking splines or a mesh. I thought a scale model in a waffle type construction of an unorthadox 3D dome would be fun and interesting thing to build. After a few sketches of what this might look like I thought it could easily be something to transform a large interior space or an exterior area by creating a unique inner sanctuary or energy zone for those who enter as well as a thing of beauty to behold both inside and out. So I began to come up with some models in Blender and Rhino, and then also tried transfering my ideas to grasshopper to give me more parametric control over my design.
My final dome design didn't come together as I had hoped it might, due to the joints and connections and it's scale. Rather than continue with this it may now make a sculptural lamp shade or wall hanging! I may get round to painting it and modifying it into something useful. However due to the thickness of this board it was not exactly utilising the shopbot and milling processes neccessary for completing this assignment. So I got to work and came up with another design to adequately fulfill the brief. This design was to make a child's baby walker in the shape of a tortoise. For some extra complexity I wanted to add the feature of a nodding head, cam powered to bob up and down as the cart is pushed! This was more difficult than it first seemed. These are the designs I came up with:
The first thing to do when setting up to use the table milling machine is to make sure your files are ready to output setting up the toolpaths and cutting order For this purpose we would use RhinoCam software. A test cut is also essential to ensure tolerances on joints and slot sizes are correct and matched for the material and tooling we are going to use. In this case I would use 15mm plywood, with a 4mm flat end-mill. I made the test part in Rhino 5 with varied slots and holes sizes increasing in 0.02mm increments.
This was then imported into RhinoCam software for setting up cutting tool paths and generating the g-code to send to the machine.
The first cut made drills holes to screw the working board material fixing it flat to the bed of the machine. Then, after screwing the board down, the next cut involves the dog-bones - these are small holes at the corners of all interior cuts.
Then the pockets - or interior holes are cut, ensuring that the tool path is set to be cutting on the inside of the lines.
Then the outer paths are cut with the tooling set to cut outside the lines. After machining three operations the parts can then be removed at their tags points with the help of a chisel and mallet (Tags are the small noggins that are left along the cutting part to keep the work from moving when being machined) any rough edges can then be cleaned and sanded.
I converted my 3D model into 2D flat parts. Then exported to RhinoCam to generate the tooling and cutting paths as described above. Then to send the g-code and select the feeds and speeds for the milling bit and material we are using. The following images and video link show the process of cutting out my parts in 15mm Plywood using the shopbot CNC mill.
Unfortunately it wasn't until about half way through cutting the outline that I noticed that the inner pockets hadn't cut after checking I confirmed this layer hadn't been sent with the g-code files.
Frustratingly it was too late as many parts had already been dislodged or cut out completely. I had to allow the machine to finish, remove the parts and do an entire re-cut. An unfortunate waste of material and steep learning curve!
After the second cut, more disappointing news was just around the corner. After I took the pieces home and spent a good hour sanding and cleaning up the edges on them, when it came to assembly I found that only the slots on the edges and not the interior pockets were of a big enough tolerance. 0.02 had not been enough for the interior holes. Even though on my test piece they both seemed a good tight fit, and this is what I wanted for the design. However, I tried some gentle chiseling and using some applied force and ended up breaking one of my parts. I think if I were to now go back and spend some time carefully with a dremel style rotary sander or a nibbler or something if I can get my hands on one! and try to widen these holes I may get this model assembled, but there are 22 holes in total and the quality of the edges of the wood I used are not really going to hold up to manual widening. Ultimately it will not be the best machined finish item that I wanted for this assignment. Having used now most of my allocated full sheet of wood in cutting twice, and also spent so much time for set up, test cut, and then cut of the parts. For now I might have to accept defeat and move on with the next assignment. I hope documentation suffices for this effort, and hope to produce something of value on the milling machines soon. If time permits I will try again with a fresh cut adjusting my hole sizes accordingly. I have learnt a lot and now know the process and set up procedure, but I am quite sad not to have come out with a finished nodding tortoise assembly.