Flashlight threads that do not gall: material pairing and anodize allowance

Threads are the part of a flashlight the customer interacts with most and the part most likely to come back. Three failure modes cover almost all of it: galling, dimensional interference after coating, and mismatch between parts cut on different setups.

Galling

Aluminium against aluminium is a poor sliding pair. Both surfaces are soft, the oxide layer is thin, and under load the two surfaces cold-weld locally and tear. That is galling, and once it starts the thread is destroyed.

Three fixes, in order of how often they are the right answer:

Hard anodize both halves. Type III produces a hard ceramic surface that will not cold-weld to itself. This is what most production flashlights do, and it is the reason a bare-aluminium prototype often threads worse than the finished product.

Change one material. Delrin, PEEK or a brass insert against aluminium slides cleanly. Common on premium hosts for the tail cap thread specifically.

Lubricate and accept it. A silicone grease on assembly is what every flashlight ships with anyway. It is not a substitute for the first two on a thread the user will operate daily.

Anodize allowance

This is the one that catches first-time brands. Anodizing is not a coating laid on top — it converts the surface, growing outward roughly half the layer thickness and inward the other half.

Type II adds perhaps 5 to 15 microns per surface. Type III adds 25 to 50. On a thread, that dimension change happens on both flanks of both parts, so a thread machined to nominal will be significantly tighter after coating, and may not assemble at all.

Tell the shop the part is getting anodized before it is cut, and specify which type. The programmer then relieves the thread — typically by machining to the low side of the tolerance band, or by specifying the thread class for the post-anodize condition. Say explicitly whether your dimensions are pre- or post-finish. Putting “all dimensions after finish” in the title block removes the ambiguity entirely.

The same applies to close-fitting bores, O-ring grooves and any sliding fit, not just threads.

Cut both halves on the same setup family

A head that threads onto its own body but binds on the next unit is usually a fixturing problem, not a tolerance one. If the male and female threads are produced on different machines, on different days, with different offsets, small differences stack.

Ask for the mating threads to be run as a matched operation where possible, and ask for a go/no-go check on the pair rather than on each part alone. On a 200-piece flashlight run this is the single most useful inspection requirement you can write.

Thread specs worth stating

• Form and pitch: metric fine is common on lights — M20x1, M22x1 and similar. State it rather than leaving the shop to scale it off a model.
• Class of fit: 6g/6H is the sensible default. Looser if the parts are hard anodized and you did not relieve.
• Thread relief and lead-in chamfer: a 30 degree lead-in on both parts makes assembly forgiving and prevents cross-threading in the customer’s hand.
• Depth of full thread: especially on blind holes, where the last turn is often incomplete.

Prototype the thread first

If the light is new geometry, ask for a thread-only test pair before the full first article: a short section of body and matching head, anodized to the production spec. It is cheap, it takes days rather than weeks, and it settles the question that otherwise gets settled by 200 finished parts that do not assemble.

Send the host drawing and say which anodize type you are planning — that is the detail that decides whether the threads work. Related: Type II vs Type III anodizing.

Square versus triangular threads

Most flashlight threads are standard 60 degree metric. Some premium lights use square or trapezoidal threads, and the reason is worth understanding.

A 60 degree thread form has a wedging action: axial load produces radial force against the flanks, which is what makes it good at staying tight and bad at sliding smoothly. A square thread has no wedge, so it operates with noticeably less friction and a smoother feel, and it is far more resistant to galling.

The cost is machining time — square threads are single-point turned rather than cut with a standard tap or die, which is slower and needs a lathe with the right capability. On a premium light where the twist action is part of the experience, it can be worth it. On a budget host it is not.

If you want square threads, say so explicitly and expect the quote to reflect the turning time.

Double-start and triple-start threads

A multi-start thread advances further per revolution, so the head comes off in fewer turns. A triple-start M20 thread can go from four turns to a little over one.

This is a genuine usability feature for a light operated one-handed, and it costs nothing in material. It does require a shop comfortable with the programming and it makes thread gauging slightly more involved. State the number of starts and the lead explicitly — “M20x1, 3-start, 3 mm lead” — because “M20x1” alone means single start.

O-ring grooves

Every threaded joint on a waterproof light has an O-ring, and the groove is a precision feature that often gets loose tolerances by default.

Specify groove diameter, width and depth, and name the O-ring size you are designing around. Standard metric or AS568 sizes are the sensible choice; a custom O-ring is a tooling cost and a spare-parts problem for your customers.

Remember the anodize allowance applies here too. A groove machined to nominal then Type III anodized is a tighter groove, and an O-ring that is over-compressed takes a set and stops sealing.

Anti-roll and knurling on the tube

Not a thread issue, but it shares the setup. Flats, facets or knurling that stop the light rolling off a table are done on the same turning operation if planned, or as a separate milling setup if not.

Say it on the drawing. Adding an anti-roll flat after the first article means a second setup that could have been free.

Testing the joint

Three checks worth asking for on the first article:

1. Assemble and disassemble twenty times, then inspect the threads under magnification for galling or burnishing
2. Check torque to seal — the point at which the O-ring compresses fully should be comfortable one-handed
3. Measure the thread after anodizing, not before, with a thread gauge rather than a caliper

The first one costs the shop ten minutes and it is the test most likely to reveal a problem before it reaches 200 units.

FAQ

Can I use the same thread spec for head and tail? You can, and it simplifies tooling and gauging. Many designs do. The only reason not to is if one joint is operated constantly and the other is opened once a year for a battery, in which case you might use square threads on the first and standard on the second.

Will hard anodizing alone stop galling? In practice, yes, for the overwhelming majority of aluminium lights. Both surfaces must be anodized — one bare half against one coated half is worse than neither.

What about titanium flashlight bodies? Titanium galls against titanium more readily than aluminium does, and titanium cannot be hard anodized in the aluminium sense. Options are a different material for one half, a dry film lubricant, or square threads. Titanium lights are genuinely harder to get right.

How much clearance should I add for Type III? Rather than adding a number yourself, state the anodize type and specify whether drawing dimensions are pre- or post-finish, and let the shop apply its own allowance. Shops know their own bath.

Should threads be cut or rolled? Cut, for these sizes and quantities. Thread rolling is a volume process with its own tooling and it is rarely the right answer at EDC batch sizes.

My prototype threads feel gritty. Is the batch ruined? Often not. Bare aluminium prototypes routinely feel worse than the anodized production part. Ask for one anodized sample pair before condemning the design. Related: Type II vs Type III anodizing.