Comfort and versatility (to make use of otherwise impossible locations) are the primary reasons that I use a hammock (rather than a similarly lightweight/compact ground shelter).
I started with a Backpacker Ultralight Asymmetrical from Hennessy Hammocks. My experience using that hammock inspired the development of a new system to replace it as documented here.
Hennessy exercises several patents on various basic design features which are conspicuously absent from other hammocks in the mainstream market. However, if I were to buy a hammock again, I might instead choose the Blackbird from Warbonnet Outdoors. In any case, my Hennessy hammock was useful/inspiring enough to keep me going and it served me well, while I developed these caveats:
This design is informed by some basic criteria (without strict precedence):
The hammock is just one part of a modular, suspended shelter system which includes:
Both tarp and hammock can be attached to the shared suspension line independently (each without the other). The tarp can be put up first and taken down last (to help keep things dry). The same suspension line can be used with different interchangeable hammocks and tarps to suit the season.
I avoid documenting specific dimensions because I’m still experimenting, and most are based on body dimensions anyway (i.e. they’re different for different bodies).
A single, continuous suspension line is used to which the other components can be easily attached. The cord remains in one long piece, ready to be used for other purposes, and also enabling the hammock to be biased toward either end (in case conditions require off-centre placement with maximal span).
Toggles are used at attachments for simplicity, ease of attachment, and to reduce stress on the cordage. To prevent attachment loops from coming off the toggles when load is removed, a simple safety lock can be made using elastic cord to wrap around the toggles, binding the attachment loops in place.
Webbing slings are used around trees:
Polyester webbing is easy to find (and it doesn’t stretch as much or absorb water like nylon). HMPE (Dyneema/Spectra) has a much greater strength:weight ratio but it’s slippery, expensive, and only saves a few grams. Nylon stretches and absorbs water, but I prefer it for durability and elasticity (as it conforms to irregular surfaces to distribute load and provides some shock absorption in the otherwise inelastic suspension line).
The basket hitch is used for strength efficiency (deliberately avoiding the much weaker choker hitch). The slings are made in two different lengths for versatility. The length of webbing is limited to avoid unnecessary material while providing enough contact area to distribute the load. The remaining length (which has little or no direct contact with the tree) is made up by a replaceable piece of cord (either a loop or a single strand with an eye at each end) attached to each end of the webbing.
The cord:webbing ratio is a compromise between two sling arrangements:
I usually avoid more than one round turn because of difficulty in balancing the load to each leg. To limit loss of strength efficiency, maximum diameter for each arrangement is limited by a sling angle of 90° between the legs (i.e. 45° between the reach and each leg).
The ends of the webbing are terminated with sewn eyes, using parallel longitudinal stitches (in preference to the weaker “box X” or bar tack patterns). A symmetrical Albright bend can be used instead but eyes in the webbing enable direct attachment of the suspension line (bypassing the cord parts) to reduce minimum hang span.
An adjustable constrictor splice is used on each end of the suspension line. Each constrictor is terminated with a spliced loop which attaches by toggle to the tree slings and enables quick and easy adjustment of the suspension line as necessary. If a shorter span is needed, the constrictors can simply be slid out of the way and the suspension line tied directly to the slings.
Alternatively, the suspension line can be run around the attachment at the ends of the tree sling and tied back to itself with an adjustable grip hitch (which also enables the suspension line to be adjusted without untying).
Adjustable or not, attachments are prone to high stress and abrasion. The slings can be augmented to mitigate abrasion and increase radius (to reduce stress on the suspension line). The eyes (or areas of contact) can be served with ringbolt hitching or whipping, and further covered with a short section of curved, low-friction tubing (e.g. HDPE, &c.).
The use of cord loops also enables a special attachment: The loops can be attached to a shared toggle using nested girth hitches, straddling (between their turns) the suspension line which runs around the toggle.
Apart from support, the hammock also provides an enclosed space, sealed from all but microscopic bugs. Symmetrical construction enables one to lie in either diagonal (or other orientations). A rectangular piece of fabric is gathered and bound at each end, and noseeum netting is fitted to the top, with zippered side entry.
Breathable, high-tenacity, continuous-filament nylon or polyester fabric is used for the hammock. Nylon stretches more than polyester (though both have similar strength); personal preference may favour one or the other.
Special attention is given to the gathering, to improve comfort by increasing the effective flatness and width, without increasing size and weight. Subtle changes in the gathering technique can significantly affect the lay. Each gathered end is bound by a cord loop which also serves as attachment.
Making the middle and edges slightly shorter (and thus higher when hung) produces a wave-like cross section. I use a W-fold on both ends; pulled flush on the foot end, and with the three peaks (the corners and the middle fold) of the W pulled out slightly to increase the effect at the head end.
Each side of the W (each V) is progressively folded in half, inward and up into the middle of the bottom.
The hammock attachments are continuous loops (joined by an end-to-end splice) made of HMPE cord. Each loop is attached to the hammock with a modified Albright bend (i.e. tied symmetrically) which also serves to bind the gathered end and remains easy to release for adjustments, repairs, &c.
Usually the Albright bend is tied asymmetrically (i.e. one end wraps around the other) because only one end is under tension, and both ends exit the same side of the loop; however, it can be made symmetrical, which is a better arrangement when tied on a bight or loop (to balance the forces when both legs are under tension). The gathered end is folded back over and enclosed in the standing part of the fabric which increases security by reducing direct contact between the cord and the tail end of the fabric (otherwise, the tail may pull out).
Alternatively, the attachment loop can be joined with a multiple fisherman’s bend (with the knot used as a thimble at the smallest radius of the girth hitch) though this is likely a weaker arrangement.
A light ridge line is used to keep the bug netting elevated and to provide attachments to hang things inside the hammock. An alpine butterfly loop is tied near each end. Movable loops/hooks/clips can also be attached.
To provide maximum height, the ridge line does not attach to the gathered ends of the hammock; instead, it is terminated with loops passed over each toggle (which elevates it to the same height as the suspension line). One end is terminated with an adjustable eye splice enabling easy adjustment of length.
A pocket made of bug netting can be slid along the ridge line (which runs through a channel in the top seam of the pocket). It has a full-size pocket on one side and a divided pocket (i.e. two pockets) on the other.
The stuff sacks for the hammock and tarp can be attached, one at each end inside the hammock (either to the loop on the ridge line, or to a loop/toggle/hook integrated with the gathering).
Noseeum netting is sewn to the hammock along one side and a #3 coil zipper (which weighs less by length than ½” velcro) is used on the other side to provide an entrance. The opening is long enough that the bug net can be folded back for a clear view. The zipper is sewn to the outside, leaving the inside seam smoother and more comfortable, and also less likely to catch debris when the hammock is being cleaned.
The insulation hammock supports insulation and also provides protection from wind and mosquitoes (which can bite through the single layer of hammock fabric). Constructed as a contoured fabric shell (like an empty under quilt), it provides five longitudinal baffled chambers which can be stuffed with improvised insulation.
I avoid carrying a dedicated under-quilt (unless conditions demand it), preferring to use what is already available instead (i.e. clothing, &c., even the pack itself). Local materials (leaves, grass, &c.) can be gathered as needed, and insulation requirements can be reduced by choosing sites carefully (for wind protection, &c.). A space blanket laid, loosely folded (several layers) in the insulation hammock, is enough for warmer temperatures (> 10 °C), and augments other insulation in colder conditions.
The insulation hammock is narrower than mainstream under quilts and designed to be held in place on either diagonal enabling one to moderate temperature by changing positions in the hammock without manipulating the insulation.
A loop at each end of the insulation hammock attaches by clove hitch around each bound end of the hammock. The insulation hammock can be left in place (though it remains easy to remove) enabling the combination to attach as a unit to the suspension line.
The space between the two hammocks is adjustable, as are the perimeter lines which must be in contact with the hammock (to reduce convective heat loss).
Many Hennessy users complain about insufficient coverage of the stock (parallelogram) tarp but it has always kept me dry (in combination with natural windbreak); however, sometimes one wants to be both out of the hammock and out of the rain, and in that case the coverage is awkwardly small.
Six-sided “hex” tarps are popular but I prefer a five-sided tarp for reduced weight and versatility. As I often hang in constrained locations, the tarp is asymmetrical in both axes (i.e. with four different contours), and hence more likely fit into tight places (because it has four different configurations instead of one). Optional guy line attachments enable the tarp to be adapted to an even greater range of shapes and configurations.
Triangular sides are useful for capturing rain water because there is a single point of exit, with little spill along the sloping edges. Reducing exit points and drip lines also makes it easier to mitigate splatter (from water running off the tarp and hitting the ground).
The two sides are joined with a seam along the ridge line. The edges are folded once and bound with binding tape to distribute force along the perimeter of the tarp, supporting attachments without additional reinforcement, and also serving as rain gutters. Separate lengths of binding tape are used for each side, overlapping at the corners for extra strength at primary attachment locations.
The edges are cut with curves to distribute tension (reducing wrinkling, &c.). Curved edges are popular and many tarps have very deep curves with a ratio of 1:12 or more but I prefer no more than 1:25. The catenary curve may be ideal but differences diminish as curves approach a flat line. In any case, using an adjustable bow made of dimensionally stable material is easier than trying to trace a hanging chain.
Soft loops are used for attachments in preference to hardware. Loops of binding tape are sewn on at primary attachments, but loops of cord can simply be threaded though the taped edge (and relocated as necessary). The attachment loops on the ridge fit over toggles on the suspension line.
Each toggle is held by a multi-overhand slip knot (for increased diameter at attachment), with an adjustable loop (made by passing the tail back through the knot) enclosing the suspension line (to keep the toggle held close to the suspension line). The other end is terminated with a perfection loop which is used to make a prusik hitch on the suspension line.
A tensioner can be used if necessary. I prefer elastic slings (a length of elastic tubing connected in a loop by a short cord) which can also be used for a fishing spear, sling bow. &c. The short cord which connects the elastic tube can be attached to a line by prusik or other means.
HMPE fishing line is used for guy lines; however, it’s either too thin (and apt to cut things) or too strong (and apt to damage the tarp). A small piece of weaker material is used in each line (between tarp and line) to avoid potential damage to the tarp (as it simply breaks when too much force is applied to the guy line).
The adjustable grip hitch works well for guy lines but I prefer the convenience of constrictor splices (like those on the suspension line), located at the tarp (rather than the far ends) for adjustment while remaining sheltered. Instead of using stakes, guy lines can be tied to resilient branches or saplings to provide spring-loaded suspension (with a round turn or two to avoid damaging the bark).
Three separate stuff sacks are used to isolate contaminants (e.g. dirt, resin, algae, moisture, &c.):
Though separate and removable, the sack for tree slings can be attached to the containing sack to serve as a reminder while packing (i.e. don’t forget to pack the tree slings as I once did!).
I deliberately make my stuff sacks somewhat larger than usual to allow the contents to compress/deform/adjust when loaded into the pack (much like organs in an abdomen) avoiding air spaces and other problems caused by rigid forms.
Nylon is avoided because it absorbs water and stretches too much. Polyester is good but HMPE (Dyneema/Spectra) cordage is preferred for its excellent strength:weight ratio (about 30% of the weight of equally strong nylon or polyester). Unfortunately, it’s also slippery and has a very low melting point, both of which greatly compromise (but do not entirely preclude) the use of friction hitches. HMPE cordage is especially weak in knots; however, adjustable bury splices work well, though they require significant length to achieve sufficient friction.
Single-braid, small-diameter, HMPE cordage is used for high-performance sailing applications and usually available in sizes down to 2.5 mm (7/64”). AmSteel-Blue from Samson is popular (and inexpensive) but I was turned off by Samson’s sleazy marketing, failure to provide requested technical information, and inferior coating. Cortland Cable will manufacture to custom specifications. I’m pleased with Marlow.
HMPE is also used for fishing line, but most sizes and strengths are too small; however, TUF-Line is available in a range of sizes up to 800 pounds breaking strength. Such fishing line is often available with a dark olive-green coating.
Most kite surfing line is made of HMPE but it usually comes in bright colours which I prefer to avoid.
Strong, lightweight fabric is made for parachutes, kites, spinnaker sails, &c. It can be found in nylon or polyester but the construction probably affects stretch more than material. Two different fabrics of a given material each weighing the same (e.g. 1.1 oz ripstop nylon) can have significant differences in physical properties based on the quality of material and construction.
Titanium and aluminum may be tempting but bamboo has a similar strength:weight ratio; it’s also easy to find and work with, and much less costly (in many ways). Aluminum tubing can be salvaged from discarded ski/hiking poles, tent poles, arrow shafts, &c. In any case, whichever material one chooses, it should be thoroughly tested to ensure sufficient strength.
The strength of a system reflects that of its materials, and perhaps more subtly and importantly, the way in which they are connected/used.
The strength of cordage is most efficient when the cord is straight; i.e. it supports the greatest load when force is evenly distributed among its strands and fibres. The more a cord is bent, the more unbalanced the force (and effectively, the weaker the cord). Knots, pins, toggles – anything a cord passes over or around – should be carefully considered. For a given cord diameter bent over a cylinder, the ratio of bend diameter to breaking strength is not proportional, as can be seen in this plot of D/d.
Knots can reduce cord strength efficiency by more than 50% (and should be avoided if more efficient alternatives are available, such as splicing), but sometimes knots are necessary. With careful attention to the application and the forces involved, losses can be reduced by choosing an appropriate knot.
The Ashley Book of Knots remains the most comprehensive knot reference ever published, and it ought to be in every library and school (because it addresses such fundamental technology of timeless utility). Some useful websites on knots:
The angle of a cord relative to the force of its load also affects strength efficiency. Cord supports maximum load when it runs parallel to the force of the load (i.e. with an angle of 0°). As the angle increases, the force on the cord is increasingly multiplied. At 60° from parallel, the load is effectively doubled (approaching infinity at 90° from parallel), as described in this document about rigging with slings.
Abrasion reduces the strength of cordage permanently. Both abrasion and bend radius can be mitigated by using thimbles when appropriate. These need not be off-the-shelf thimbles; an eye can be augmented with smaller diameter cord (ringbolt hitching, &c.), or a tubular sheath, &c.
Material properties vary significantly. Some materials, especially expensive synthetics (Technora, Kevlar, Vectran, Dyneema/Spectra, &c.) which have extraordinary tensile strength, have very low transverse strength (which is important to knot strength). Even a small amount of flex cycling can permanently reduce strength by 40% or more. Consider: melting point (friction), elasticity (shock loading), water absorption (may change properties), tolerance to potentially damaging environmental conditions (UV light, chemicals, temperature, &c.). Naturally, construction (twist, tightness of braid, &c.) also affects strength.