Photogrammetry Lighting Dome:

The light dome was needed to improve the scan and stitch quality of the booths software pipeline. Coworkers had previously determined the amount of lux needed to hit targeted scan quality by using stage lighting on C-Stands. With a deadline two months out, I planned to spend half that time prototyping and the second half building the final design. However, the scanner’s event schedule shifted up over a month, leaving me with just enough time to finish a single prototype with the small assortment of power tools I store in my grandfather’s garage.

Design Constraints

The initial request was for a portable dome that could sit on top of the scanner without external support.  Unfortunately, the scanner pillar material is too weak to support much more than its own weight would not be considered though rebuilding these pillars out of 6061 channel aluminum was suggested.

Light Requirements

The required lux was over 10,000 measured from the bottom center-point of the scanner’s footprint. In addition to this high value, the light also needed to be uniform in color and exceptionally diffuse to prevent color-grading and specular highlights from interfering with the stitching software. This requires a very high lumen value for a light source over 10ft away. High lumen garage lights and exterior E26 light socket strands were selected for the prototype due to quick shipping.

CAD Work

 

Using Rhino 5, I was able to design a 10 foot diameter geodesic dome and accurately measure necessary lengths and vertex angles for fabrication. The vertex angles were particularly challenging, as each vertex in software represented a bolt that must pass through a hole in 3 or 4 struts. The distance between the two holes in a strut, and the angle the strut is bent is critical to strength and alignment. Being off by a 1/32 of an inch is enough to prevent assembly. Based on strut length and tab angles on either side, there were several types to be made.

Geodesic Dome Design

It would have been simpler to create a flat lighting panel, but other recent upgrades to the scanner added hardware centered above the scanning volume. In order to work around these components, a dome was needed. The dome is based on a 2-frequency octahedron. This means that there are two unique triangles tessellated inwards from an octagonal perimeter. Most geodesic domes are based on an icosohedron due to their superior strength, but a compromise was made to reduce setup time and construction cost by using a shape with fewer edges. Additionally, the general hemispherical design of geodesic domes was deviated from to reduce the overall height of the fully upgraded scanner. The 10 foot diameter dome was squished downwards into a dish shape with a height of 2 feet. Despite these diversions from what makes these shapes so strong, the light dome has repeatedly help up to 300-pound compression tests on the center vertex and shearing on the bottom ring of struts.

Fabrication

With the approximate weight of lights and cable under 100lbs, I decided to build the struts out of schedule 40 steel EMT conduit. Taking my measurements from CAD, I labeled, carefully plotted where the holes were to be drilled, and tubing cut. Once cut to approximate length, the conduit ends were shaped with an arbor press. The intent was to have at least an inch of flat space on the conduit on either side of the hole. This would allow for fasteners and other struts to align without intersecting tubular sections. Once the conduits were shaped in the arbor press, I brought them to the drill press to create the bolt holes. Afterwards I bent the tabs to the correct angle using reference guides printed from Rhino. Lastly, excess material beyond the holes on each strut were removed with an angle grinder.

Touring

My deadline for construction of the final dome was moved from a month to a week away. Thankfully, the struts, fasteners, and cabling fit in an over-sized duffel-bag. The light bulbs are placed in a separate hard-shell tote box for safe transport. The prototype has been setup and torn down a dozen times, transported around the country, and lit thousands of people without any wear.

Planned Version

The final version of the dome was to replace the open-air frame with a closed dome of interlocking lighting panels. These flat-pack panels would have data and power pass-through for the center mounted sensors and lighting powered through daisy-chained quick connectors.