Planning the Package


A well-designed package will minimize unfolded joints and place them where they don’t carry much load, or add reinforcement if they do.  The floor of the package is likely to bear the greatest load, such as batteries.  The floor should be joined to the walls of the package with at least reinforced butt joints; this was the method used for the ATV package, but that floor bore only the weight of the camera. 

A sturdier floor-to-wall joint will result if the walls are formed as bent-up extensions of the floor.  The walls can be joined as mitered reinforced butt joints.  This approach might be best for shallow packages with wall seams less than 4 or 5 inches high. 

If the wall structure is complex, the bottoms of the walls can be bent inward to form ½” or so wide flanges; the ends of these flanges should be cut at the proper angle to form a tight miter when they’re all folded inwards.  After the floor is cut to size, it can be pushed in from the top and glued to the flanges to form lap joints.  The exposed edges of the flanges can be protected by cutting a smaller second floor which just fits inside the area circumscribed by the flange edges and gluing it in place; this will add weight, but it more than doubles floor strength and R-value.  It’s probably an excellent approach to use if the floor is over a foot square. 

Reinforcement strips should be located inside the package to present the smoothest possible surface to air and to grabby objects on the ground. 

During a rough drag over the ground, battery mounts are heavily stressed.  They should be designed to take it, and to prevent shorting and damage to other components if the mount fails.  The EOSS Shuttle is provided with an internal foamcore battery box with a lid.  The box is located near the center of the floor to control CG, and its bottom is glued to the package floor.  We’ve encountered no structural problems so far with this method. 

Distributing the cells over a wide area and supporting them near the inside corners would be structurally sounder than concentrating them all in the center of the floor.  Some weight savings may be realized by eliminating reinforcement.  Battery replacement would be more tedious, however. 

Access to the interior of the package is best made from the top via a removable lid.  The ceiling is unlikely to carry much load, and it’s the least likely surface to contact the ground during landing.  Thus its joints don’t require much strength.  The ATV module lid is a flat sheet which carries alignment strips on the inside.  These strips fit snugly inside the top edges of the walls.  The strips not only keep the lid aligned, but they also add to the package’s torsion resistance and prevent a straight shot for water and EMI penetration.  The raw edges of the lid and walls are separately sealed with space tape before aluminum foil shielding is glued on the external surface.  A few space tape tags hold the lid in place during flight. 

A stronger lid might be formed as a cap which fits over the tops of the walls, but one is then one is faced with a bigger snagging target. 



Printed circuit cards may be mounted in the conventional fashion using standoffs and threaded hardware through the package walls.  It’s quite easy to form card guides out strips of foamcore glued to the package walls, however.  This makes access and preflight replacement a snap.  If the there’s not at least 1/8” free edge on the card to clear the slot, then the card can be mounted to a separate sheet of thin foamcore which slides into its own slots.

Captive machine threads may be included for ease of disassembly using T-nuts, commonly used in cabinet carpentry.  Ordinary machine nuts may also be glued or space-taped in place on the foamcore surface, where screw tension pulls the nut into the surface.  Large-diameter flat washers will prevent the nut from pulling through if the screw is over tightened or subject to high stress.

Low-stress openable lap joints, such as for an enclosure lid, can be secured using sheet metal or deck screws; the threads engage OK into the paper.  The needle tool makes a suitable pilot hole. A stronger joint will result if the screw passes through a double layer; this can be implemented by gluing a ½” – 1” square patch of foamcore on the inside.

External connectors for ground, power, antenna feedlines, etc. may be hot-melted into snug-fitting holes in the package wall.  Connectors which are expected to be heavily stressed may be secured to patches of printed circuit board material before being glued into the package.  Connectors which project out from the sides of the package are subject to drag damage.  They should be avoided or shrouded; suitable shrouds may be made from foamcore.  These shrouds may also be designed to mount strain reliefs for cables or their messenger lines. 

Load-bearing attachment points, such as for payload support lines, are best made on vertical walls near the outer corners.  It’s also possible to cradle the package in macramé fashion, but this can create a challenging tangle for the post-flight crew.  Single-point attachment to the center of a flat lid is not recommended except for very small packages.  A steeple-topped package will tolerate this well, but a pyramid of nylon line is much lighter.  We have had some success with dowel-backed holes in the sides which accept loops at the line ends.  After the loops are attached, the holes are sealed with RTV.  This method avoids projecting snag targets.

If your payload comprises a number of packages strung together, design your support system to avoid transferring the weight of the lower packages through the upper package structure.  It’s far better to let the support lines carry that tension past the structure than it is to burden it with heavy reinforcements.  The support lines may be tied together at single points above and below the package.  Compressive stress is reduced significantly by placing these points at least 3 package widths away.

Recently, we have used threaded vertical through bushings to provide both support line attachment and to secure the lid. This method keeps all of the flight line tension off the payload structure.   Threaded 3/8 NPT pipe intended for lamp fixtures, along with mating nuts, washers and bushings, may be found at the hardware store.  A nylon dress bushing with a hole just large enough to pass your flight line will provide a safe mating surface to a figure-8 knot in the line.  Plastic tubing is lighter, but it is tedious to thread.

Package spin during flight is probably unavoidable.  Its adverse effects are windup of cables running between strung-together parts of the flight system and motion sickness by those who watch the ATV image of the ground for too long.  Spin has a slow, basal rate intrinsic to balloon dynamics, so I’m told, plus a faster oscillatory component.  Wind shear causes suspended parts of the system to sway horizontally, creating a horizontal component of airflow around the package.  Asymmetrical projections will weathervane downwind, resulting in an oscillatory yaw rates which may be much faster than the basal spin rate.  Oscillatory spin might be reduced by making the package nearly cylindrical to reduce the horizontal drag coefficient, and large asymmetrical projections should be avoided. 



Foamcore material has proven to be an attractive material for high altitude balloon payload package construction.  The design and construction techniques described have been used successfully on over 40 EOSS flights.  They were developed over just a few projects, however, and are by no means the pinnacle of the art.  Others of you will develop significant improvements in materials and techniques.  In the meantime, it is hoped that some of you in the high altitude balloon community will be encouraged to fly innovative new payloads using what insights might be gained from our experience.