Costing a POM Part: Resin, Cycle Time and Scrap in One Model

Polyoxymethylene (POM), also described as acetal resin, is a high-crystallinity engineering thermoplastic used for gears, fuel-system components, snap-fit housings, and precision wear parts. Costing a POM Part: Resin, Cycle Time and Scrap in One Model treats the accepted-part cost as a linked function of three operational variables rather than a single resin-price lookup. The material term is dominated by shot mass and grade density; the cycle-time term is dominated by cooling time, mold temperature, and part thickness; the scrap term is dominated by cold-runner mass, reject fraction, and regrind return. A formal cost-per-accepted-part expression is:

C_part = [m_shot × P_virgin] / (1 − s_reject) + (t_cycle / 3600) × R_machine + m_scrap × H_scrap

In this expression, m_shot includes part and cold-runner mass; s_reject is the fraction of total shots that are rejected; t_cycle is the complete cycle measured in seconds; R_machine is the machine hourly rate; m_scrap is the mass of regrind or purge; and H_scrap is the handling cost per kilogram of scrap. Hot-runner versions set runner mass to zero in m_shot but add a hot-runner maintenance and capital charge to the machine rate.

Density influences the resin term directly. Unfilled POM grades are typically reported at 1.41 g/cm³ to 1.42 g/cm³ according to ISO 1183-1:2019. A part designed with a 2.0 mm wall and a projected volume of 15 cm³ has an estimated cavity mass of 21.2 g at 1.41 g/cm³; shifting to a glass-reinforced grade with density above 1.55 g/cm³ may increase mass by more than 9% unless wall thickness is reduced. The selected melt-flow grade also changes the minimum feasible wall thickness and therefore the part mass. Medium-flow injection grades with an MFR near 9 g/10 min at 190 °C and 2.16 kg load are common for thick-wall gears, while high-flow grades above 22 g/10 min are used for thin-wall connectors and tight flow paths. Melt-flow values are determined by ISO 1133-1:2022.

Solving the three variables independently creates false optima. A resin-cost reduction that selects a lower-viscosity POM may fill a thinner wall and shorten cooling time, but if the grade has lower impact toughness and increases crack rejects at snap-fit gates, the accepted-part cost can rise because the reject divisor increases. Similarly, raising mold temperature improves surface gloss and dimensional stability but adds to cooling time. The combined model therefore compares alternatives only at equal accepted-part output and equal tolerance capability.

What Is the Function of Scrap Fraction in a Combined POM Cost Model?

The scrap fraction is not a flat waste multiplier. For a reject fraction s_reject = 0.02, the number of shot cycles required to produce 1000 accepted parts is 1000 / (1 − 0.02) = 1020.4. The resin-consumed factor is therefore 1.0204 before runner mass is counted. Cold-runner systems add a separate runner-mass ratio. If a sprue and runner weigh 20% of the part mass, the shot mass is 1.20 times the part mass, and the combined material factor becomes 1.20 / 0.98 = 1.2245. In hot-runner tools, the runner term approaches zero, but the hot-runner system adds capital, thermocouple, and tip maintenance costs that should be charged to the machine-rate term. The model therefore compares cold-runner and hot-runner options only after the scrap divisor and the hourly tool overhead are both entered.

Regrind of POM is viable within limits. Clean sprues, runners, and rejected parts can be ground and reintroduced if the melt-flow rate is tracked by ISO 1133-1:2022. Repeated processing of POM at melt temperatures above 230 °C reduces molecular weight and raises MFR, which can shift pressure requirements and increase flash. Supplier technical bulletins for standard unfilled POM grades typically permit regrind levels of 15 wt% to 20 wt% for non-critical parts, provided that the regrind is dry, segregated by grade, and limited to two or three thermal histories. Published data for specific post-consumer or mixed-color POM scrap is limited; therefore, the model should default to a lower regrind ceiling for dimensionally checked parts.

Within the cooling phase, the governing material parameter is effective thermal diffusivity. Unfilled POM has a thermal conductivity near 0.31 W/(m·K), a specific heat capacity near 1.5 kJ/(kg·K), and a density of 1.41 g/cm³, giving an effective thermal diffusivity of approximately 0.14 mm²/s. The ideal cooling time depends on the square of wall thickness. For a 4 mm flat wall with a melt temperature of 215 °C, mold temperature of 90 °C, and ejection temperature of 120 °C, the one-dimensional conduction estimate is approximately 14 s; adding injection, packing, screw recovery, mold open/close, and ejection brings the total cycle into published industrial ranges of 30 s to 40 s for similar wall sections. The exact total must be validated on the specific press because screw recovery, mold movement, and robot removal are machine-dependent.

Melt viscosity controls injection pressure and therefore clamp force. High-flow unfilled POM with MFR 22 g/10 min reduces injection pressure at identical flow length and part thickness compared with a medium-flow grade. On a 1000 kN electric injection molding machine with a 25 mm screw and L/D ratio of 20:1, processing a 2.0 mm wall gear of 35 g shot mass, barrel-temperature profile of 190–215 °C, and mold temperature of 80 °C, the high-flow grade may permit a 5–10% reduction in injection time and a lower holding-pressure peak. However, high-flow grades can exhibit reduced Charpy notched impact strength and may require gate location shifts because weld-line strength is affected. These changes alter the reject divisor, so the cycle-time gain must be compared at equal reject fraction.

Machine-rate allocation includes energy, labor, maintenance, and depreciation. The cycle-time term is the most sensitive after resin mass. For any positive machine rate, reducing t_cycle from 35 s to 32 s lowers the machine cost per part by 8.6%. If that change is achieved by a material substitution that increases the reject divisor from 1% to 2%, the net accepted-part cost may remain unchanged or rise. The integrated model therefore requires a paired comparison of cycle-time savings against the scrap fraction.

When High-Flow POM Shortens Holding Time, the Reject Rate May Offset the Resin Savings

A high-flow POM grade may allow a shorter hold time because cavity-pressure decay is faster after fill, but this is not a universal reduction. Semicrystalline POM solidifies rapidly, and the gate freeze time depends on gate diameter, mold temperature, and melt temperature. For a 2.0 mm wall with a 1.0 mm diameter tunnel gate, increasing mold temperature from 80 °C to 100 °C prolongs the time to gate seal by approximately 5–10% and can increase total cooling time by 3–5 s. A high-flow grade with a lower molecular weight may form a frozen skin that is less resistant to sink-mark growth, so the holding pressure must be held long enough to compensate for density loss. If the hold time is cut too aggressively, sink marks and post-mold shrinkage exceed the limits of ISO 294-4:2018, and the reject fraction increases.

Thermal degradation places an upper boundary on melt temperature and residence time. POM releases formaldehyde when overheated, and the degradation is autocatalytic if the melt is held above 230 °C or if acid or chlorine-containing residues are present. Processing equipment should use a vented barrel, a low-compression screw, and a shut-off nozzle. The melt-residence time at high barrel temperatures should be kept below 30 min; purging with polyethylene or polypropylene is standard at shutdown. If a high-mold-temperature strategy is used to improve surface appearance, the cost model should include the additional venting, purging, and potential scrap from black specks because degradation-related rejects add directly to the scrap divisor.

Across a multicavity tool, dimensional reject rates are governed by molding shrinkage and process capability. Unfilled POM exhibits mold shrinkage of 1.8% to 2.5% depending on wall thickness, flow direction, and mold temperature. Measurements are made according to ISO 294-4:2018 or ASTM D955-08(2014). A tolerance drawing that requires a ±0.10 mm dimension on a 50 mm length corresponds to a relative tolerance of ±0.20%. If the shrinkage window is 0.2% across the cavity, the process may be at or beyond the drawing limit. This makes dimensional capability a direct input to the scrap divisor. Process capability indices quantify the effect: a centered two-sided process with Cpk 1.33 produces an expected reject fraction near 63 parts per million, while Cpk 1.00 produces a two-sided expected reject fraction near 0.27%. The cost model uses these reject fractions to compute the number of cycles needed per accepted part.

The homopolymer values reflect higher crystallinity, higher tensile strength and flexural modulus, but a higher melt temperature and narrower processing window; the high-flow copolymer reduces viscosity but sacrifices some impact strength. The integrated cost model uses these values only as initial estimates. Final values must be taken from the actual lot certificate because POM batch-to-batch MFR variation can reach ±2 g/10 min in commercial supply, and that shift can alter fill pressure and the reject divisor.

Cooling-channel placement determines whether the cycle-time reduction is achieved without increasing scrap. On a multicavity POM gear tool, a conventional straight-drilled cooling circuit can produce a cavity-to-cavity mold-surface temperature variation of ±5 °C. For POM, a 5 °C mold-temperature shift can alter shrinkage in the range of 0.05% to 0.15%. This may be enough to move a tight gear dimension outside ISO 294-4:2018 tolerance. Conformal cooling or beryllium-copper inserts reduce the temperature spread but add tool cost. The combined model adds tooling amortization to the machine-rate term and compares the reduction in reject fraction against the higher tooling depreciation.

Screw recovery must be completed within the cooling window to avoid extending cycle time. For POM, screw speeds of 50–150 min⁻¹ and back pressures of 0.5–3.0 MPa are common; excessive back pressure above 5 MPa can overheat the melt and increase splay or black specks. If the screw recovery time is longer than the cooling time, the machine rate term is set by screw recovery, and the part cost model must substitute the recovery-limited cycle time. This occurs with large shot volumes, small screw diameter, or low-density regrind. The resin mass, cycle time, and scrap terms converge at this constraint: a larger screw may reduce shear heating and recovery time, but it can also increase residence time and degrade the melt if the shot is too small.

Purge and shutdown sequences constitute a hidden scrap cost that is specific to POM. When a barrel is held at processing temperature without cycling, formaldehyde-based degradation can accumulate. A standard shutdown procedure uses polypropylene or polyethylene purge at a lower temperature until the barrel is clean. The purge mass is not recoverable as POM regrind and must be entered in the scrap-handling term. For a 40 mm screw barrel, the purge volume may approach 1.5 to 2.0 times the shot capacity; the cost model should include this mass at the POM resin price because it is consumed without producing accepted parts.