ABS Thermal Conductivity and Specific Heat: What Matters in Mould Cooling

Acrylonitrile butadiene styrene (ABS) is an amorphous matrix of styrene-acrylonitrile copolymer with dispersed polybutadiene rubber domains. In injection moulding, the cooling phase generally controls cycle time because heat is conducted through a low-conductivity polymer shell before reaching the tool steel. The relevant material inputs are not limited to steady-state thermal conductivity; specific heat and thermal diffusivity define the transient heat-extraction path from melt to ejection. Representative unfilled ABS values at 23 °C are 0.17–0.25 W·m⁻¹·K⁻¹ for thermal conductivity, 1.25–1.55 J·g⁻¹·K⁻¹ for specific heat, and 1030–1070 kg·m⁻³ for density. These ranges are grade-dependent, with rubber content, SAN copolymer composition, pigments, and processing orientation shifting the measured values. Thermal diffusivity α = k/(ρ·C_p) then falls between roughly 0.11 mm²·s⁻¹ and 0.15 mm²·s⁻¹ at room temperature. At the melt-processing range of 220–260 °C, specific heat rises toward 2.00–2.40 J·g⁻¹·K⁻¹ while thermal conductivity remains relatively flat, so diffusivity drops to approximately 0.08–0.10 mm²·s⁻¹. Because ABS is amorphous, no crystalline latent-heat term is added; enthalpy removal is dominated by specific heat over the cooling interval.

Why Does Solid-State Thermal Conductivity Fail to Predict Mould-Cooling Time?

For a slab-like ABS part, transient conduction limits heat removal rather than steady-state resistance through the steel. The conduction-limited cooling time can be approximated by t_c = (s² / (π²·α)) × ln[(8/π²)·((T_melt − T_mould) / (T_eject − T_mould))]. The equation shows that cooling time scales with the square of wall thickness and inversely with thermal diffusivity, not with k alone. An unfilled ABS grade with k of 0.20 W·m⁻¹·K⁻¹, C_p of 1.45 J·g⁻¹·K⁻¹ at 23 °C, and density 1050 kg·m⁻³ yields α ≈ 0.131 mm²·s⁻¹. For a 2.0 mm thick slab, melt temperature 230 °C, mould temperature 50 °C, and ejection temperature 85 °C, the conduction term is 3.38 s and the logarithmic temperature correction is 1.43, producing a conduction-limited floor near 4.8 s. Production cooling time is longer because packing, hold, gate freeze, and coolant-circuit response add to the thermal path. The same calculation for a 3.0 mm wall gives 10.8 s conduction-limited cooling, showing the quadratic penalty of thickness.

As the melt contacts the tool wall, a solidified skin forms and grows inward. The heat flux across this skin is roughly k_solid × (T_{inside skin} − T_mould)/δ(t), where δ(t) is the local frozen-layer thickness. Because δ(t) grows with the square root of time, the instantaneous heat flux decreases. This is why transient diffusivity, rather than steady-state mould-to-coolant resistance, controls local solidification. A grade that exhibits only a 0.02 W·m⁻¹·K⁻¹ shift in solid-state conductivity may change predicted cooling time by less than the repeatability of the thermal-property method unless the shift is accompanied by a density or specific-heat change. Higher rubber grades tend to sit at the lower end of the published conductivity range because the polybutadiene phase is less conductive than the SAN phase; this can shift simulated cycle time by several percentage points when grade-specific ISO 22007-2 data replace generic datasheet inputs.

Specific Heat, Shot Mass, and Thermolator Duty in Repetitive Cooling

Specific heat governs the amount of enthalpy removed per unit mass. For an illustrative production case with a combined component and cold-runner shot mass of 156 g, melt temperature 240 °C, and ejection temperature 90 °C, an average C_p of 1.9 kJ·kg⁻¹·K⁻¹ over the cooling interval gives an energy removal of 44.5 kJ per shot. At a 30 s cycle, this corresponds to an average heat-removal rate of 1.48 kW. A four-cavity tool under the same conditions would reject approximately 5.9 kW to the cooling circuits, excluding machine nozzle and sprue losses. This calculation is not a universal ABS value; it is linearly dependent on the grade-specific C_p measured by ISO 11357-4 or ASTM E1269 and on the selected ejection criterion.

Coolant temperature rise can be calculated from the heat load and volumetric flow rate. At 10 L·min⁻¹ water flow per circuit, the thermal capacity flow is approximately 0.697 kW·K⁻¹, and the illustrative 1.48 kW load produces a circuit ΔT of about 2.1 K. At 5 L·min⁻¹, the same load produces about 4.2 K. Closed-loop thermolators with proportional-integral-derivative control typically maintain setpoint within ±1 K when flow is turbulent and circuits are balanced, but the instrument response depends on heat-exchanger size and reservoir volume. For ABS, an oversized heat-load spike at the moment of injection can appear if the cooling circuit contains dead zones or if flow drops below turbulent conditions.

In a high-volume line producing thin-wall ABS components, cooling capacity must be matched to the shot rate rather than to the average cycle time alone. If screw recovery time is shorter than cooling time, the heat exchanger cannot remove accumulated heat and mould temperature drifts upward. Published production data for specific ABS grades and tool layouts are limited, so cold-water injection tests and pressure-drop measurements across each circuit are normally performed before establishing the cooling-critical cycle window.

Mould-side heat transfer rarely limits the cooling rate of unfilled ABS because water-side coefficients are orders of magnitude above polymer-side values. For an 8 mm diameter water channel with flow velocity 3 m·s⁻¹ at 30 °C, the Reynolds number is approximately 30,000, and the water-side heat-transfer coefficient is roughly 12,000–13,000 W·m⁻²·K⁻¹ by the Dittus-Boelter correlation. A tool-steel cooling channel at 25–30 W·m⁻¹·K⁻¹ adds only modest conductive resistance. The controlling resistances are the solidified ABS layer and the polymer-tool interface. Effective polymer-side conductance is typically estimated in the range 1,000–3,000 W·m⁻²·K⁻¹, although published ABS-specific interface data are limited. The overall effective heat-transfer coefficient from ABS melt to coolant therefore does not scale with water flow once turbulent flow is achieved; increasing water-line flow beyond the turbulent threshold may improve thermal uniformity but not the conduction rate through the frozen layer.

The first priority for ABS cooling design is placing channels close to the cavity surface with uniform pitch, because the dominant resistance is the part thickness. The second priority is balanced turbulent flow in every circuit. A circuit at 4 L·min⁻¹ in an 8 mm line produces Re around 10,000; below that, laminar coefficients can fall to a few hundred W·m⁻²·K⁻¹ and create hot-spot asymmetry. Pressure drop per circuit can be estimated from the Darcy-Weisbach relation; for a 1.5 m smooth line at 3 m·s⁻¹, the pressure drop is below 0.2 bar, but bends, fittings, and ceramic inserts can increase it. Published data for production ABS moulds with conformal cooling channels are sparse; published data for this specific configuration is limited to tool-specific measurements.

When Glass Fibre Loading Reaches 20 wt%, Thermal Anisotropy Ceases to Be Negligible

Short-glass-fibre-reinforced ABS is used where structural stiffness and heat-distortion resistance are required. At glass loadings near 20 wt%, the compounded thermal behaviour becomes anisotropic because oriented fibres increase in-plane conductivity more than through-plane conductivity. Published data for specific glass-fibre-reinforced ABS grades are limited; the following are indicative ranges from compounded-material technical bulletins. In-plane thermal conductivity is typically near 0.28–0.35 W·m⁻¹·K⁻¹, while through-plane conductivity remains closer to 0.22–0.27 W·m⁻¹·K⁻¹; the anisotropy ratio is often 1.2–1.5. Specific heat decreases relative to unfilled resin because glass fibre has lower specific heat than the polymer. Depending on fibre content and coupling agent, room-temperature C_p may fall toward 1.15–1.35 J·g⁻¹·K⁻¹, and density rises to approximately 1.15–1.25 g·cm⁻³. The result is that in-plane diffusivity increases, but through-plane diffusivity remains close to unfilled ABS or improves only modestly.

For a flat plaque cooled from both surfaces, heat must travel through the thickness, so through-plane conductivity controls solidification. A moulding engineer who observes the elevated in-plane datasheet value and shortens cooling time can produce under-cooled parts with sink marks, gate stringing, and poor dimensional stability. The same anisotropy influences warpage prediction: shrinkage calculated from thermal gradients in simulation must use direction-dependent conductivity and pressure-volume-temperature data, not a single isotropic k. In addition, glass-filled grades have higher melt viscosity and require higher injection pressures, which can raise polymer-tool contact conductance by improving surface conformity during packing. The net effect on cooling is therefore not a linear function of filler loading.

Carbon black and mineral fillers modify conductivity differently. Carbon black at conductive-grade loading raises both electrical and thermal conductivity through particle contacts, but the thermal percolation threshold is not identical to the electrical percolation threshold. With carbon black, the increase in thermal conductivity is usually smaller than the increase in electrical conductivity because the contrast between filler and polymer thermal conductivity is lower than the contrast in electrical resistivity. Published data for specific carbon-black-containing ABS grades should be obtained from ISO 22007-2 measurements on injection-moulded plaques, preferably with in-plane and through-plane sensor orientations.

Mould-cooling simulations require temperature-dependent k and C_p as input. Specific heat is commonly measured by differential scanning calorimetry under ISO 11357-4 using a sapphire calibration standard, a heating rate of 10 K·min⁻¹ or 20 K·min⁻¹, and controlled purge gas. Thermal conductivity for unfilled ABS is usually measured by transient plane source according to ISO 22007-2 on plaques thick enough to avoid sensor boundary penetration; for anisotropic glass-filled material, the instrument must be configured for anisotropic modelling. Guarded hot plate methods such as ASTM C177 and steady-state heat-flow meters such as ASTM C518 give through-thickness conductivity, but they require larger specimens and longer equilibration. Thermal diffusivity can be measured directly by laser flash per ISO 22007-4, although semi-transparent ABS specimens require an opaque coating and the measured half-rise time must be corrected for finite-pulse effects.

These ranges are not a substitute for grade-specific moulding trials. The discrepancy between thermocouple-instrumented tools and simulation often arises because the material file uses room-temperature values or because the selected C_p is not averaged over the actual cooling interval. For ABS, the glass-rubber transition between 90 °C and 130 °C can produce a shoulder in the specific-heat curve that affects enthalpy removal at ejection temperatures near the transition. If a part is ejected at 75 °C, the material may still be in or near the transition region; thermal capacity remains high, and the surface can reheat after demoulding. Tool instrumentation with cavity-surface thermocouples and pressure sensors is commonly used to validate simulation boundary conditions.

Cooling-Time Penalty for a 10 °C Lower Ejection Temperature

Reducing ejection temperature is sometimes used to improve part stability, but the conduction-limited cooling time responds nonlinearly because the logarithmic term contains the ratio between melt-mould and ejection-mould temperature differences. Using the same 2.0 mm slab, α of 0.12 mm²·s⁻¹, melt temperature 230 °C, and mould temperature 50 °C, lowering ejection temperature from 85 °C to 75 °C increases the logarithmic correction factor from 1.43 to 1.76, and the conduction-limited time rises from approximately 4.8 s to 5.9 s. That is an increase of about 23 %. At 3.0 mm wall, the same ejection change raises conduction-limited cooling from 10.8 s to 13.4 s.

The non-linearity is important when a processor seeks lower ejection temperature to meet dimensional tolerances. A mould temperature change has a smaller logarithmic effect than an ejection-temperature change when the melt temperature is held fixed, but it changes the shrinkage envelope. If the mould surface is below the dew point of the production hall, condensation rapidly destabilises surface gloss and can produce splay on an ABS moulding; mould coolers operating below 15–18 °C often require closed-loop dew-point control or dry air around the clamp. For unfilled ABS, a controlled mould temperature near the lower end of the supplier range, typically 40–60 °C, is common. Published data for a specific ABS grade and a given multiphase mould should be verified by grade-specific DSC and transient plane source measurements before tooling is committed.