Sent 18-Feb-03In answer to a question on how to house a GPS payload Mike answered ...
If you've not seen my paper on foamcore construction on the EOSS website, I'd recommend you peek at it; I've included some tips on how to make foamcore housings quite robust for balloon payloads.
Re your question: I'm guessing that your GPS experiment would fit easily into a housing no larger than a 20 cm cube. And at that size, a simple box of 0.6 cm thick foamcore board with mitered and glued joints will be more than robust enough to survive many flights and rough landings.
The greatest stresses appear at the flight line attachment points, since the flight line carries the tension of all the lower payloads and can experience some significant G enhancement during post-burst chaos. So it's best not to allow that loading to be passed thru your payload. This can be effected easily by providing a central, vertical hollow spine through which the line can pass; simple knots in the line above and below the payload will serve to keep it in place relative to the other payloads. I use 3/8" IPT threaded pipe of the type used in common table lamps for that purpose, and this stuff is available in the lighting section of Home Depot and Ace Hardware, along with matching nuts, threaded couplings and plastic anti-chafe bushings (very important to avoid cutting the flight line!). This method, unlike others tried, has NEVER failed to retain the payload on the line at recovery, regardless of how severe the landing might be.
note: we did recently (EOSS-81) suffer a payload separation event using the pass through method. However, it appears that the structure of the string between two payloads failed and separated.
It is also possible to provide line attachment eyes on one side of your housing, but these MUST be well secured and their load spread over the inside using large fender washers, for example. We have lost two CU Gateways Balloonsats for failure to implement this technique properly, however.
Now, I have done some weird-shaped payloads, e.g hex and octagonal approximations to vertical cylinders (foamcore is not adaptable to curved forms!), with appropriately cut miters at the corners. But this was done for aerodynamic purposes rather than strength. I've even built a dodecahedron (geodesic sphere) with just 2 butt joints just to demo the flexibility of foamcore construction.
If you havn't handled foamcore, I'd recommend that you play with a piece to gain some insights into the kinds of stresses it can tolerate and those it can't. It's cheap and easily available at arts and crafts shops, such as Hobby Lobby, where its most common application is in matting for framed pictures.Further comments on Foam Core Housings in answer to another question:
I've used foamcore for payload structures for over 10 years and have not encountered any problems with structural instability or weakening due to expansion of trapped gas. I think the gas bubbles in the polystyrene foam used in foamcore may be too small to overstress the styrene bubble walls. Water is a greater threat, however, and that can come from passing thru a cloud, especially on descent. As a rule, I put a thin coat of clear Krylon paint to moisture-proof the paper shear layers, and this also seems to render some additional tear strength to resist abrasion during a post-landing parachute drag.
EOSS does have one payload housing formed entirely from high-density polyethylene (PE) foam, but none of its fit tolerances are so tight as to present a deformation risk. And cutting tight tolerances is near impossible in that stuff, to boot. I use foamcore inside the PE housing where tight tolerances are needed. PE foam is nearly indestructible, however, so I do use use a glob of it to absorb landing shock on the bottom of some of the payloads. Foamcore's not too shabby in that regard either, but it can get pretty beat up after 20 or more flights, including a few hard landings.
I've had some rotten experience with epoxy joints on payloads: it embrittles at low temps. And I'd sure question how well it'll bond to polyethylene. I have made some built-up PE structures (shock absorbers) using 3M #90 spray adhesive, and it seems to tolerate the environment well. Ordinary hot-melt adhesive works great on foamcore joints, and it'll do a pretty good job on metal and plastic as well, provided both of the pieces to be joined are burnished and heated above the melt temperature.
Don't rely on ANY adhesive to keep your payload attached to the flight string! This is just begging for a loss during post-burst chaos.
As a rule, I design my housings around a central vertical tubular spine thru which the payload line can pass. A figure-8 knot in the line just below and above the spine secures the paylaod to the line and isolates it from the line tension. Thin-wall 3/8" threaded pipe commonly used for table lamps works well for this, and the mating nuts and washers allow for easy access to the interior of a clamshell housing; and although it's made of steel, it weighs only about 1 oz per foot.