§1Tool-free by design
A hand fastener is a deliberate design statement: this joint will be opened often, by unknown hands, with no tool present — so the tool is built into the fastener.
Every other page in this section assumes a spanner, a key or a driver. This one removes the assumption. Inspection covers, battery boxes, guards, jigs and fixtures changed every batch, clamps adjusted a dozen times a shift, equipment serviced in the field — all share the property that the cost of access dominates the joint, and a fastener needing a tool multiplies that cost or, worse, invites the wrong tool (§6). The wing nut and its relatives answer by growing their own lever: wings, a knurled disc, a flatted head — surfaces fingers can grip and turn. What is gained is obvious. What is given up is the subject of §3: the lever fingers can turn is short and the force they can apply is small, so a hand fastener can never reach the preloads the torque-and-tension page is built on. The whole discipline of using them is knowing that, and designing the joint so it does not matter (§4–5).
Contents§2The family
Four basic forms cover the field, differing mainly in which member carries the grip and how much torque the grip geometry allows.
The wing nut is the classic: an ordinary nut with two wings cast or pressed on, run down a standing stud or bolt — the largest hand lever of the family, and the natural choice where the male thread is fixed. The wing screw inverts it, putting the wings on the screw itself so the whole fastener comes away with the cover it secures — nothing left standing to snag, one part to lose instead of two. The thumb screw trades the wings for a compact head, either knurled — a cylinder with the knurling pages’ diamond or straight pattern giving fingers their grip — or flatted, a flat blade of head pinched between finger and thumb; both are slimmer than wings, kinder in cramped panels, and correspondingly lower in torque, since their effective radius is smaller. Around the basic four cluster the practical variants: captive thumb screws retained in their cover so they cannot be dropped into the machine (§5), spring-loaded and quarter-turn types for the fastest access, and plastic-headed hand knobs — star, tri-lobe, tee — which are simply thumb screws with the ergonomics enlarged, and which mark the transition from fastener to clamping handle.
Contents§3What fingers can do
Put numbers on finger-tightening and the result is humbling: a wing nut delivers about a newton-metre, and on its thread that is a few per cent of what the fastener could carry.
Two fingers on the wings apply a couple: roughly 40 N each at an effective radius of 12 mm gives T = 2 × 40 × 0.012 = 0.96 N·m — call it one newton-metre, the working ceiling of firm finger effort. Feed that into the torque–tension relation on an M6 (K = 0.2): F = T/(K·d) = 0.96 ÷ (0.2 × 0.006) = 800 N of preload. Yet the M6 class 8.8 itself yields at 640 × 20.1 = 12.9 kN — so the wing nut has developed just 6.2% of its own fastener’s capacity. The geometry offers only one lever to pull: span. Double the wing radius to 24 mm and the same fingers give 1.92 N·m and 1600 N — 12.4% — which is why hand knobs grow with the duty, and why a big tri-lobe knob is not styling but torque. But no plausible span reaches the 75%-of-proof preload the bolted-joint pages target. The conclusion is not that hand fasteners are inadequate; it is that they are a different device: a closure, not a preloaded joint — and §4 draws out what that means.
§4Location, not preload
With one newton-metre available, everything the preloaded joint achieves — fatigue resistance, friction-carried shear, vibration security — is off the table, and the joint must be designed to need none of it.
Recall what preload buys, from the torque-and-tension page: the clamped members carry the fluctuating load, friction between faces carries the shear, and thread friction resists loosening. At 6% of proof load a hand-tightened joint has essentially none of these. So the surrounding design must supply them structurally: the cover locates on its own features — a rebate, a lip, dowels from the pins page — so the fastener positions nothing; shear never passes through the hand fastener, which merely holds the cover against its seat; fluctuating loads are reacted by the structure, not the thread; and sealing, if needed, comes from a compliant gasket that seals at low and uneven clamping force, not a hard joint face demanding tonnes. Done properly, the wing nut’s only duty is the one thing fingers can genuinely do — keep a located, unloaded closure closed — and the arithmetic of §3 is ample for it. Done improperly — a hand fastener asked to preload a structural or vibrating joint — the result is the commonest failure of the species: the joint that “keeps coming loose”, which was never tight in any engineering sense to begin with.
Contents§5Designing for hands
A hand fastener is a human–machine interface, and the design rules are ergonomic before they are mechanical.
Size the grip to the duty: §3’s arithmetic scales with radius, so the head grows with the required closing force — thumb screw, wing, then knob — and knurling or lobes are chosen so a cold or gloved hand can still turn it; a smooth head that needs dry fingers fails its whole purpose. Keep the thread coarse and short: coarse pitch means fewer turns per opening (the machine screws page’s advance-per-turn, used in reverse), and a fine thread on a hand fastener wears patience and invites cross-threading by feel. Make it captive: a loose wing nut above an open gearbox is a failure waiting its turn — captive screws, retained washers, or a lanyard keep the fastener with its cover, and the pins page’s clevis-and-lynch arrangement does the same for hitch duty. Protect the hand: deburred wings, generous radii, clearance for knuckles at full turn. And signal the intent: a wing nut announces “open me by hand” as clearly as a socket head announces the opposite — mixed signals, like a hand knob on a joint that actually needs a torqued bolt, are design lies that someone eventually believes.
Contents§6Keeping them closed — and honest
The hand fastener’s two chronic complaints — it vibrates loose, and someone put pliers on it — are the same complaint, and both are answered in the design, not the tightening.
At §3’s preload there is little thread friction holding the nut, so under vibration a plain wing nut will walk — the machine screws page’s loosening logic at its most exposed. The remedies are the low-preload members of that page’s locking family: a spring or wave washer under the wings to keep some live axial force as the joint relaxes; a nyloc-style insert in the wing nut for prevailing torque that fingers can still overcome; a detent or serrated seat; or, where the duty is genuinely vibratory, the honest admission that a hand fastener is the wrong device. The pliers problem is the same truth from the other side: when a wing nut must be “persuaded” with a tool, the joint has outgrown finger torque — and pliers promptly bend the wings, wreck the grip, and hide the overload instead of fixing it. Corroded or gritty threads produce the identical symptom by stealing the one newton-metre in friction before any of it becomes clamp; on outdoor and field equipment, thread condition is the fastener’s capacity. The closing rule of the page: if hands are no longer enough, the answer is never a tool on a hand fastener — it is a redesign of either the joint or the fastener choice.
Contents§7Quick reference
The working core of the page on one card rack.
Purpose
frequent, tool-free access
the lever is the fastener
Family
wing nut · wing screw
knurled · flatted thumb
Ceiling
T = 2Fr ≈ 0.96 N·m
→ 800 N on an M6 = 6.2%
Doctrine
closure, not preload
locate and load elsewhere
Security
spring washer · captive
pliers mean redesign