12/13/2015 at 05:54 •
It was cold and raining, and took a long time. But it was well worth it.
Here is the mount I came up with to hold Explorer 1 up at a 45 degree angle:
The main piece of white plastic is simply a section of 6 inch PVC pipe coupler, cut down to the shape needed. Note the holes for mounting screws to the body.
A few weeks later, the display is still standing, despite wind and weather. So far, so good!
12/13/2015 at 05:26 •
The aft hemisphere was bolted to the front hemisphere. They were spaced a distance apart to provide a spherical shape, as the two hemispheres were not quite full hemispheres. (These were inexpensive security mirror ceiling tiles made of PMMA). The spacing worked out well, because the original had a wide equatorial feature. This was represented by a band of plastic automotive molding, which helped to cover the mechanical support structure. The equatorial feature was aft of center in the original, and the hemispheres were cut slightly unequally in order to evoke this configuration.
The mating structures were screwed and clamped in place to ensure repeatable positioning.
Other decorative embellishments on the aft hemisphere were round features, one on the main axis and one offset. The main axis feature was a support lug, and the offset feature was an electrical connection. These were assembled from PVC parts, as well as aluminum rod, a sawed and sanded-down RF connector, screws, glue, and paint. The aft hemisphere was quite thin in the area where these features mounted, so I reinforced the attachment area with fiberglass.
Both hemispheres were primed, blacked, then chrome painted, before assembly.
12/13/2015 at 04:56 •
The main body of Explorer 1 was a 6-inch nominal ID PVC pipe. It had numerous decorative embellishments, including many screws inserted to look like the rivets and screws originally used, a paint job designed to look like the original stainless steel and white coloring, as well as a chrome-painted coupling ring, and a section representing the original aft fiberglass slot antenna, the Juno 1 rocket booster attachment, and meteoroid sensor. Four stainless steel rods were threaded into the body to represent the four whip antennae used on the original. These had rubber tips added for safety - this was, after all, a display piece.
Afterward, the nose cone and tail cone were inserted and attached with screws.
12/13/2015 at 04:39 •
The forward hemisphere housed the main load bearing structure of the satellite. The main weight and load bearing stress were due to the four antennae, each an 8 foot long section of half inch diameter aluminum pipe.
(Yes, I know, not metric parts - but again, timeline and price were dominant here, and these were pretty close.)
The structure was made of aluminum extrusion, and provided the main support for the antennae, which were bolted through the front hemisphere. The front hemisphere was slightly reinforced with fiberglass, but really only to prevent cracking, not for load bearing purposes. The outer perimeter was reinforced with an aluminum strip, to help maintain the circular edge, and allow for some depth and stronger material for attachment screws for the rear hemisphere to bolt through. The perimeter screw holes were drilled by hand, at precisely marked spots, and using a gauge block, in order to get a consistent offset and precise radial positioning.
All other screw holes were marked and drilled and countersinked. Some cracks had to be repaired prior to chrome painting.
Antenna mounts were made of 1 inch diameter aluminum rod, turned on a lathe to create the taper of the antenna mount insulators (and later painted to look that way), then cut, drilled, and tapped to provide mounting to the frame, and receptacles for the antenna pipes, which were set-screwed in place. Antenna springs and mounts were cut on a Shapeoko 2 and assembled with stainless M3 screws.
12/13/2015 at 03:52 •
The antenna mount shields were cast in epoxy resin and chrome painted.
The original form was modeled in balsa wood and automotive body filler, then finished. The resulting form was molded in silicone, then the silicone mold was used to create epoxy resin copies. These copies were filled, sanded, primed, and chrome painted.
The most important aspect of the shield was that it was fit to a 58 cm sphere. This curve was allowed in the original model by aligning the balsa wood grain so that the wood could flex along a curve of the proper radius. The wood form was glued into shape and then filled and finished.
12/13/2015 at 03:21 •
The tail cone of Explorer 1 was created with a similar plastic doll cone form as the nose. Similarly, the cone was forced into a circular projection and then fiberglassed for structural integrity. The top of the cone was inserted into an air duct adapter having approximately the right mating angle, prior to fiberglassing. The final cone was filled with automotive body filler, then sanded, primed, and painted.
The tail cone mounting was created from a different air duct adapter, which was cut off and hollowed out to accept the composite tail cone. Outside of this was fitted a cake ring, which had been drilled to create the right number of holes as the feature on the real Explorer 1. The cake ring was epoxy coated to provide thickness and reduce sharp edges.
This was fit inside a section of PVC pipe that had been tapered to match the original profile.
The final assembly was able to be fit to the body via the same kind of fiberglass coupling made for the nose cone. However, the tail coupling does engage fully into the body, so therefore did not need to be painted.
12/13/2015 at 03:04 •
The Explorer 1 nose cone was fabricated from a plastic doll cone form, with the top cut off and half of a DIY Christmas ornament of the right size glued into the top. The flexible cone was fitted onto a pipe section to force the cone into a circular configuration. This was very stable when coupled with the circular joint between the cone and the Christmas ornament half. This allowed the nearly final shape to be stable enough to be fiberglassed in place, giving the nose cone structural integrity. This was joined to a small section of standard 6 inch PVC pipe at the bottom. The nose cone was finished with automotive body filler, then sanded, primed, and painted.
A mounting collar was made by fiberglassing the interior of a section of 6 inch PVC pipe, prepared with mold release. The resulting fitted cylinder was then driven out and became an insert of exactly the right size. It was attached to the inside of the cone, in the lower standard PVC pipe section, and allowed mounting to the body, also made of the same standard PVC pipe.
Painting required a mask to produce the patterned white and stainless steel coloring of the fore section of the satellite.
Holes were drilled and tapped to take screws in the locations shown in historical photos. The screws help to secure the mounting collar, though that is also affixed with adhesive. The mounting collar does not go fully into the body; rather it is left with a gap to represent the fore fiberglass slot antenna of the original. It is therefore painted with a color evoking the Bakelite of the era.
12/13/2015 at 02:05 •
These display replicas needed to be made on such a short timeline, that the techniques employed required hacking.
Had there been more time, designs could have been drawn up, things could have been fabricated exactly to scale, etc. However, the timeline dictated a more advanced starting point and strategy: find things that are close in shape and size, and cut and glue them together with minimal fabrication of major components.
Ultimately, Explorer 1 was made from a 6 inch (nominal ID) PVC pipe, some plastic doll forms, a Christmas ornament, two air duct adapters, a cake ring, screws, stainless steel rod, fiberglass, and plenty of epoxy, glue, and paint.
Sputnik PS-1 was made from two Plexiglas (PMMA) security dome mirror (near-)hemispheres, aluminum rod, aluminum tubing, aluminum extrusion, aluminum strips, plastic automotive molding, PVC pipe, fiberglass, epoxy resin, screws, and chrome paint.
Regarding the chrome paint, I've tried a lot of them over the years with varying results. There was no time to do a lot of testing for this project, but fortunately I found someone who had done the testing already and published it. I credit this video for the final choice of chrome paint (and related surface preparations).