Injection-moulded tablet stand: sink, warp and draft before the tool is cut
A Hong Kong consumer-goods manufacturer is tooling a 90 mm × 60 mm ABS desk stand for a tablet, at 60,000 per month on a 90-tonne machine. The nominal wall is 3.0 mm. A 6 mm-diameter screw boss stands at each of the four corners, each with a 2.0 mm-thick rib running 25 mm toward the centre, and the gate is a single 1.2 mm side gate at one end of the 90 mm edge. First shots show: a sink mark about 0.15 mm deep on the outer face directly opposite each boss; the base warps 1.2 mm across its 90 mm length, bowing away from the cavity side; short shots along the far edge; a cycle of 28 s against a target of 20 s; and 0.8° draft on the 30 mm-deep side walls. The customer requires flatness within 0.3 mm, no visible sink, and a Class B cosmetic finish on the outer face. Tool steel is P20, and the cooling circuit is a single straight line 12 mm from the cavity surface.
1. Explain the sink marks and propose design changes. Explain why machine settings alone will not solve the problem.
Read the worked answer
Sink marks opposite the bosses are a local shrinkage problem caused by a thick section. Each boss is a solid 6 mm mass against a 3 mm nominal wall, so it freezes much later than the surrounding plate. The plate sets first, the boss continues to shrink, and there is no melt nearby to compensate, so the outer surface is pulled in. The ribs at 2.0 mm are 0.67 t, above the usual 0.5–0.6 t guideline, adding further local mass.
Why packing pressure cannot fix it: the gate is only 1.2 mm and sits at one end of a 90 mm flow path, so it freezes before the bosses can be packed. Short shots along the far edge confirm that the pressure at the end of fill is already marginal. Raising pack pressure would flash the gate and increase residual stress without reaching the bosses.
Design changes:
- Boss: keep the 6 mm outside diameter but reduce the wall to 1.8 mm (0.6 t), with 2° taper and a 0.5 mm base radius.
- Ribs: reduce to 1.5 mm (0.5 t), height no more than three times thickness, 0.5 mm fillets.
- Use short gussets instead of long ribs where possible.
- Move the gate to the centre, or add a second gate, so all four bosses see packing pressure.
- Add a baffle to cool the boss from the inside.
Acceptance: sink ≤ 0.02 mm on the Class B face. The examinable judgement is that geometry must be corrected before the tool is cut; a 2:1 thickness ratio cannot be processed away.
2. Diagnose the 1.2 mm warp and propose remedies in order of effectiveness.
Read the worked answer
Bowing away from the cavity side across 90 mm indicates differential shrinkage between the two faces of the moulding.
Evidence-based causes:
- Gate at one end. Melt pressure and temperature fall along the 90 mm flow, so the far end packs less and shrinks more, pulling the part into a bow.
- Single cooling line 12 mm from the cavity surface. The cavity side cools faster than the core side, so the cavity skin sets first and the core side then shrinks more, warping the part toward the core - consistent with a bow away from the cavity.
- Incomplete cooling at 28 s. Ejecting while the core is still hot allows post-ejection shrinkage and adds to the bow.
Remedies, in order:
- Balance the cooling: add a baffle or bubbler to the core, or a second circuit, so the two halves are within 5 °C and cool at similar rates across the part.
- Centre the gate or add a second gate to equalise flow length and packing.
- Increase pack pressure and pack time until the gate freezes, and use a pack profile.
- Only if flatness is still out of specification, plan a cooling fixture or a 0.5 mm counter-curvature in the tool. These hide the symptom and should be the last resort.
Measurement: CMM on an unrestrained part after 24 hours, not immediately after ejection. The examinable reasoning is that warp is a differential-shrinkage problem and is solved by thermal balance, not by packing harder.
3. The cycle is 28 s against a 20 s target and the side-wall draft is 0.8°. Evaluate both, and advise what should change before the tool is cut.
Read the worked answer
Cycle time. Cooling normally dominates an injection cycle, at around 60–70% of it, and a 3 mm ABS wall needs substantial heat removal. A single 12 mm cooling line is inadequate for a 90 mm part: the flow path is long and the boss regions are far from the line. Practical levers are increasing heat-transfer rate rather than cutting hold time blindly - baffled or conformal cooling near the bosses, mould temperature at the lower end of the ABS range (about 40–50 °C), and checking whether the 90-tonne machine is actually plasticising-limited, since screw recovery can set the floor on cycle time regardless of cooling.
Draft. 0.8° over 30 mm gives about 0.42 mm of clearance per side. That is workable on a polished wall, but Class B cosmetic finish often implies a light matte texture. As a rule, allow about 1° of extra draft per 0.025 mm of texture depth: a 0.05 mm texture would then need roughly 2.8°, not 0.8°. Under-drafted textured walls scuff, drag and can crack at ejection.
Recommendation before cutting: specify at least 1.5° on walls that will be textured and 1° on polished walls, and confirm the texture depth with the customer. Adding draft after hardening means EDM time, possible weld-line repair, or a new insert. The examinable judgement is that draft and wall geometry are cheap to change on the drawing and expensive to change in hardened steel; process settings should be optimised only after the part geometry is sound.