Heat bending
Both defects are process problems with known causes: white haze from overheating, bubbles from moisture. The fix is a temperature window and a drying step, neither of which is expensive.
White haze and bubbles look like the same problem and are not.
White haze along the bend is thermal damage. The sheet was taken above the forming window and the polymer degraded. It may appear immediately or develop over weeks as the damaged zone stress-cracks.
Bubbles are moisture. Cast acrylic absorbs water from the air, and when the sheet is heated quickly the water turns to steam inside the material before it can escape. Thick sheet holds more moisture and needs longer drying.
There is a third defect worth separating: crazing, fine surface cracks that appear later. That is residual stress, usually from bending too cold or cooling too fast, and it interacts with chemical exposure.
Acrylic softens enough to form between roughly 150 °C and 170 °C. Below the window the sheet resists, and forcing it produces stress and often a hazy bend. Above it, the material degrades and the bend goes white or yellowish.
| Temperature | What happens |
|---|---|
| Below 140 °C | Too stiff; forcing produces stress, uneven radius, springback |
| 150–160 °C | Good forming range for most thicknesses; slower but controllable |
| 160–170 °C | Upper end; faster forming, less margin for error |
| Above 175 °C | Degradation; white haze, yellowing, bubbling |
The heater type matters as much as the number. Infrared heaters penetrate and can overheat the surface before the core is ready; contact and oven heating are slower but more uniform. For sheet above about 8 mm, even heating through the thickness becomes the dominant problem.
Cast acrylic absorbs moisture, up to about 0.3–0.4% by weight in humid conditions. That sounds trivial and is enough to ruin a bend.
When the sheet is heated, the absorbed water turns to steam. In thin sheet the steam escapes through the surface relatively easily. In 15 mm sheet it cannot, so it collects at the softened interface and forms bubbles along the bend.
The fix is to dry the sheet before forming: a controlled oven cycle at around 80 °C, held long enough for the moisture to diffuse out, then formed while still warm or cooled in a dry environment.
| Thickness | Drying at 80 °C | Notes |
|---|---|---|
| Under 5 mm | Usually not required | Steam escapes readily; risk is low |
| 5–10 mm | 2–4 hours | Recommended for critical bends |
| 10–15 mm | 4–8 hours | Necessary; bubbles otherwise likely |
| 15–20 mm | 8–16 hours | Necessary; also slow the heating ramp |
The times are indicative and depend on the oven and the ambient humidity. What matters is that the step happens at all — suppliers who skip it produce bends that look fine on the bench.
Acrylic cannot be bent to a sharp internal corner. The outer surface stretches and, beyond a limit, cracks or crazes. The limit scales with thickness.
| Thickness | Minimum inside radius | Practical radius |
|---|---|---|
| 2 mm | 1 mm | 2 mm |
| 3 mm | 1.5 mm | 3 mm |
| 5 mm | 2.5 mm | 5 mm |
| 8 mm | 4 mm | 8 mm |
| 10 mm | 5 mm | 10 mm |
| 15 mm | 8 mm | 15 mm |
| 20 mm | 10 mm | 20 mm |
A practical rule: for thickness above 8 mm, keep the inside radius at least equal to the thickness. It looks better, it stresses the material less, and it gives the bend a chance to be annealed successfully.
Line bending heats a narrow strip along the bend line and folds the sheet. Fast, accurate for straight bends, and the standard method for anything rectangular. It leaves a slightly visible heat line across the bend unless the strip width is controlled.
Drape forming heats the whole sheet and lets it sag over a form or into a mould. Better for gentle curves and for shapes with more than one bend. Slower, and requires a form for each profile.
Free-form bending for large architectural curves, often using a template and controlled local heating. The most skill-dependent of the three.
For thick sheet, annealing after forming is standard practice on anything load-bearing or chemically exposed. It adds a day and it is what prevents the crazing that appears three months later.
Case one. A 20 mm guard bent to an 8 mm radius cracked at the outer surface during cooling. The radius was below the practical limit for that thickness. Remade with a 20 mm radius; the design changed slightly and the part has been fine since.
Case two. Bubbles along the bend line in 12 mm sheet, discovered after polishing. No drying step. The whole batch was scrapped — the defect is not repairable.
Case three. White haze on 5 mm bends. Heater set too high to speed up the cycle. Lowering the temperature and increasing dwell fixed it, at the cost of about 20% longer cycle time.
Case four. No visible defect at delivery, crazing at the bends after four months in a coolant atmosphere. No annealing, and the residual stress plus the chemical exposure did the rest.
Three of these four would have been prevented by a slower, cooler, drier process. None of them required better material.
Five things to check on a bent part, in order:
1. Bend angle. Measure after the part has cooled fully — acrylic springs back slightly and a part measured hot will read wrong.
2. Radius. Check against the drawing, on the inside.
3. Haze. Inspect the bend under grazing light against a dark background. Faint haze is the early sign of thermal damage.
4. Bubbles. Backlight the bend. Any bubble in the formed zone is a reject, because it cannot be repaired.
5. Crazing. Inspect the outer surface at the bend under grazing light, ideally after 24 hours. Crazing that appears this early will get worse in service.
No. Once bubbles form inside the material the defect is permanent — polishing removes surface material but the bubbles are internal. The part has to be remade. This is why the drying step matters more than the forming step for thick sheet.
About 5 mm as an absolute minimum, but 10 mm is the sensible practical value. Bending tighter than the practical radius produces high outer-surface stress that cracks either immediately or later in service.
A faint heat line is normal in line bending, because the heated strip has a slightly different optical history. It can be minimised by controlling strip width and temperature, and it is usually invisible on anything other than clear sheet under direct light. If it is pronounced, the temperature was too high.
Above about 10 mm thickness, or where the part will be exposed to solvents or vibration, yes. Annealing releases the forming stress that otherwise causes crazing months later. It adds roughly a day and 8–15% to the part price.
Radius, thickness and bend angle are enough for us to tell you whether the bend is feasible and whether it needs annealing. Better to know before the tooling is built.
