Machine Inscriptions

Typically, CNC programmed toolpathing is aimed at removing material in the quickest or most efficient manner - aspects which are essential when trying to streamline workflow and produce parts for production. CNC machining also allows for extremely high levels of finish - on a range of complicated parts - even surfaces with complex curvature. Machine Inscriptions looks at toolpathing for other attributes - to uncover potentially innate qualities in various ‘paths’ or bring new ‘readings’ to parts through expressive patterns or textures. These studies looked at several techniques specifically to leverage the unique qualities of aluminum.

A single overall surface / composition provided the base for six separate test parts. A simple parametric technique generated the base surface (through deformation of a flat plane) and was also used to drive two of the test patterns. The specific patterns/tests were:

A_ point drilling

B_ streamlines

C_ staggered paths

D_ progressive topo

E_ mesh pillowing

F_ intersecting contours

TOOLS:

HAAS VF2-SS

Sandblaster

1x6 aluminum flat bar (6) x 6” pieces

A variety of end mills were used for roughing and finishing the aluminum panels. In all but the ‘mesh pillowing’ technique, the ‘finished’ parts were sandblasted prior to running the toolpath textures for added variation between the ‘stock’ and the machined ‘texture’.

A series of points (taken from the overall parametric surface) were projected down onto the part. These projected points became the center-points for a drilling routine. Unfortunately Inventor HSM won’t automatically select all the points for drilling (if the drilling procedure isn’t perpendicular to a surface - which these weren’t) so I had to select every point individually for the operation. A ball end mill was used to drill below the original surface by a specified depth. Depending on how closely spaced the points were, the intersections between the divots varied the overall pattern in unexpected ways.

This technique took the same points used to create the base surfaces peaks and valleys and used them as particle charges to create a vector field with a Grasshopper plugin called FlowLines. The resultant vector lines were then projected to the base part and used for a ‘tracing’ operation in Inventor HSM. These tracings were machined again using a ball end mill.

To created the staggered pattern paths, lines taken from the parametric surface were then further altered using another script to create dashes and spaces with the ability to alter the proportions of both. After spacing was tuned, the dashes were projected to the base part again for running a ‘tracing’ operation in Inventor HSM. These tracings were machined using a ball end mill. Depending on the spacing of the dashes the tracings could run into one another, creating some interesting variations in the otherwise straightforward pattern.

To generate a topography of a sloped terrain, a series of evenly spaced contours is taken through the terrain, giving a reading of slope - where closely spaced contours indicate a steep area and widely spaced contours indicate a gradual incline. A simple script generated an uneven (and parametrically controllable) set of contours. Having this control over the vertical placement allowed for either an amplification of the terrain or for confusing it (by inverting the usual understanding of closeness = steepness). Once the desired topo contours were generated, they were used for ‘tracing’ operations in Inventor HSM. These tracings were machined again using a ball end mill. In this particular sample the shallower areas of the part have a much closer spacing of contours and machining.

For this part, a script used to ‘pillow’ meshes was applied as the first step in the process. The parameters allowed for more pillowing in ‘deeper’ areas (or areas with further deviation from a baseline). Rather than simply leave it with a fine parallel finish (as was done in the one of my earlier CNC explorations - GEODE Series) a regularly spaced contour through the mesh was taken and again used for ‘tracing’ operations in Inventor HSM. These tracings were machined using a small ball end mill. Because the ‘pillowing’ varied between areas, the contouring was quite irregular in appearance and slightly unexpected.

This final technique was simply generated by taking two sets of evenly spaced contours through the part/surface at different angles, resulting in a pattern of intersecting contours. The contour lines were then used for ‘tracing’ operations in Inventor HSM. These tracings were machined using a ball end mill. Where lines converged more tightly, more material was removed, giving quite a variety of quality across the surface.

In general, all of the tests/techniques were successful to some degree, however some of the results were rather predictable. The more interesting aspects (converging paths, overlapping and distorted areas, various depths, etc.) seemed a result of unforeseen factors (not controlled in Rhino or Inventor HSM). These occurrences seemed to yield particularly rich areas in the aluminum - where texture and visual affect are amplified. Moving forward (as this is an ongoing interest) the results of this first round could be capitalized upon - perhaps using larger tools for more unintended overlap, other endmill types (v or chamfer for example) or using less conservative toolpath spacing for more intersections. Additionally, modifying depth-of-cut between or within toolpaths themselves could add more variation to the process.