Machining Acrylic Rod: Cutting, Drilling and Polishing Without Crazing

Acrylic is not a difficult material — it is an unforgiving one

Machinists who come from metal find acrylic easy: it cuts fast, it needs no coolant on most operations, and the surface finish can be optical. Machinists who come from other plastics find it frustrating, because the failure modes are delayed. A part that looks perfect on the bench can craze three weeks later, and the crack always starts where the tool was.

The reason is that acrylic (PMMA) is amorphous and brittle. It has no crystalline phase to absorb energy, so any stress concentration stays concentrated. It is also notch-sensitive: a scratch, a tool mark or a sharp internal corner raises the local stress by a factor of three to five, and combined with residual stress from machining it is enough to propagate a crack.

Cutting to length

The first operation sets up everything downstream. A bad cut leaves a stressed, chipped end that will crack during the next operation or in service.

Table 1 — Cutting methods for acrylic rod. Laser gives a beautiful edge and a stress field that cracks later; it is a sheet process, not a rod process.

For a circular saw the parameters matter more than the blade. A triple-chip carbide blade with 60–80 teeth, running at 2,000–3,000 rpm with a feed of 0.05–0.1 mm per tooth, gives a clean edge. Too slow a feed and the blade rubs, generating heat and melting the surface; too fast and the teeth chip the exit edge.

Tournage

Turning is the most common operation on rod and the one where tool geometry decides the outcome. Acrylic needs a positive rake and a sharp edge — a tool that would be acceptable on aluminium will rub on acrylic and leave a dull, stressed surface.

Table 2 — Turning parameters for cast acrylic rod. The finishing pass should remove at least 0.2 mm, because a lighter pass rubs rather than cuts.

The most common turning mistake is taking a very light finishing pass to "be safe". Below about 0.1 mm depth the tool tends to rub, which generates heat and work-hardens the surface layer into a stressed state. A 0.3 mm pass at higher speed gives a better surface and lower stress than a 0.05 mm pass.

Tool material is usually high-speed steel or polished carbide. Carbide lasts longer but must be sharp — a worn carbide insert will rub and heat the surface. Diamond-tipped tools are used for optical work where the surface must be clear without polishing.

Forage

Drilling is where acrylic cracks most often, because the drill generates heat at the bottom of a hole it cannot escape from, and the flutes can grab and screw the bit into the material.

Table 3 — Drilling acrylic. The slow spiral and relieved lands are the two features that separate an acrylic drill from a general-purpose one.

Deep holes are a separate problem. Beyond about three diameters, chips cannot clear and the drill overheats. The practical approach is to drill from both ends, or to use a coolant-fed drill. A deep hole that has been drilled hot will craze from the inside out over weeks, and there is no way to inspect for it.

Polishing and the edge that has to be clear

Acrylic rod is usually bought for its clarity, so the machined surfaces have to be restored to transparency. That is a sequence, not a single step.

Table 4 — Polishing sequence. Every stage must remove the scratches of the previous one completely, or the final buff will reveal them.

Flame polishing is faster and gives a glass-clear edge in seconds, but it introduces a heat-affected zone with high residual stress. On a rod that will be subject to load or thermal cycling, flame polishing is the starting point for later crazing. It is fine for display parts and decorative edges, and it is not fine for structural or outdoor use.

Annealing: the step that is usually skipped

Machining leaves residual stress in acrylic. The stress is concentrated at the machined surface and at any sharp corner, and it is invisible. Annealing relaxes it, and it is the single most effective thing you can do to prevent delayed cracking.

Table 5 — Annealing schedule for cast acrylic. The cooling rate matters as much as the soak — cooling too fast reintroduces the stress you just removed.

The annealing temperature must stay below the point where the rod distorts. For cast acrylic that is around 80–85 °C; extruded acrylic softens lower and needs 70–75 °C. Above that the part will sag under its own weight, which on a long rod is visible as a bow.

The test for whether annealing worked is a solvent test: wipe a machined surface with a solvent (isopropyl alcohol or a proper stress-crazing agent), let it evaporate, and look for fine cracks. An unannealed part will craze within seconds. An annealed part will not. It is a destructive test but it can be done on a test piece from the same batch.

Faults and causes

Table 6 — Machining faults in acrylic. Haze that will not polish out is not a polishing problem — the polymer has been thermally damaged and has to be cut away.

Asymmetric removal is worth understanding on rod specifically. If you machine a flat on one side of a rod, you release stress on that side only, and the rod bends towards the cut. On cast rod the internal stress is usually low enough that this is minor; on extruded rod it can be significant. If the part needs a flat, machine small amounts from both sides in alternation where possible.

Cast and extruded rod machine differently

The two production routes give different machining behaviour, and the difference is not always stated on the quote.

Cast rod has higher molecular weight, better optical clarity, lower internal stress and machines cleanly. It also costs more and comes in a narrower size range.

Extruded rod has lower molecular weight, slightly lower clarity, and higher internal stress — which means it is more prone to craze after machining and more prone to bend when material is removed asymmetrically. It is cheaper and available in continuous lengths.

For a machined part with a critical clear surface or a tight tolerance, cast is the right choice. For a non-critical spacer or a decorative part, extruded is fine. The failure mode to watch for is specifying extruded to save money on a part that then crazes in service — the saving is a fraction of the part cost and the failure is total.