Extrudate die swell in acrylonitrile-butadiene-styrene filament production is the viscoelastic recovery of melt immediately downstream of the forming die. The response is generated by recoverable shear strain stored during converging flow through the die land, and in 1.75 mm and 2.85 mm diameter filament lines it acts as a direct disturbance to closed-loop diameter control. Uncontrolled swell changes the effective drawdown ratio between die exit and the laser micrometer, shifting final diameter even when screw speed, melt temperature, and puller speed remain nominally constant. Moisture uptake alters this response because water absorbed by the resin flashes to steam at melt processing temperatures, producing intermittent pressure pulses in the metering zone and at the die entry that modify the recoverable strain field. ABS grades equilibrate to moisture levels typically between 0.2 wt% and 0.6 wt% depending on relative humidity, rubber phase content, and compounding additives, with moisture measured by ISO 15512:2019 or ASTM D6869-17. The resulting diameter variation is therefore a coupled rheological, transport, and control problem that cannot be resolved by draw-speed correction alone if the resin moisture state lies outside the target window.

Why Does Unconditioned ABS Resin Produce Diameter Drift in Small-Diameter Filament Lines?

Unconditioned ABS resin produces diameter drift because moisture is not uniformly distributed within the pellet bed, and the transition from liquid water to steam at melt temperature creates a two-phase flow regime in the die. The equilibrium moisture uptake of general-purpose ABS at 23 °C and 50 % RH is commonly reported in the range 0.25 wt% to 0.45 wt%, although impact-modified grades with higher polybutadiene content can exceed 0.60 wt%. When pellets with this moisture content enter an extruder operating between 210 °C and 250 °C, water flashes to steam within the compression and metering zones. The resulting gas bubbles do not remain dispersed; they collapse and re-form under pressure, causing melt-pressure oscillations that are typically observed on the gear pump inlet or die pressure transducer with amplitudes of 0.2 MPa to 1.0 MPa at frequencies below 5 Hz. These pressure oscillations directly modulate first normal stress difference at the die wall, which controls die swell, and they alter local melt density. The filament diameter sensor, commonly a two-axis laser micrometer with a sampling rate of 100 Hz to 2400 Hz, detects the resulting periodic diameter variation after the water bath and air knife, but the feedback loop sees a phase lag equal to the transport time from die to sensor. Consequently, the corrective puller-speed command may reinforce the oscillation rather than suppress it. This is the basis for the common production observation that diameter variability increases when the dryer dew-point sensor reads above -30 °C or when virgin pellets are blended with regrind stored in open hoppers at ambient conditions.

Across single-screw filament lines, ABS is normally processed on extruders with screw diameters from 20 mm to 50 mm, L/D ratios between 24:1 and 30:1, and three-zone barrels with compression ratios from 2.5:1 to 3.5:1. A melt pump between the extruder and the die stabilizes throughput and reduces surge caused by pellet size variation and hopper bridging. The forming die for 1.75 mm filament is usually not identical to the final diameter; dies from 2.5 mm to 4.0 mm are common, with the melt drawn down in a controlled air gap before entering a 40 °C to 60 °C water bath. Drawdown ratio, defined as the square of die diameter divided by the square of final filament diameter, typically lies between 2:1 and 5:1 for ABS filament lines. Die swell acts as a positive offset to the effective draw ratio, reducing the amount of draw required to reach target diameter. If swell varies by 5 % in relative terms, the open-loop diameter error can exceed 0.05 mm on a 1.75 mm diameter line unless the puller speed is corrected. In-line diameter measurement must therefore be located after the filament reaches dimensional stability in the cooling bath, and the control loop must include a feed-forward term from die pressure rather than relying solely on feedback from the micrometer.

Die Land Length, Drawdown Ratio, and Apparent Shear Rate Effects

The apparent Newtonian shear rate at the die wall for a circular die is calculated as γ_a = 4Q/(πR³), where Q is volumetric flow rate and R is die radius. For a 2.5 mm diameter die at a volumetric output of 12 cm³/min, γ_a is approximately 130 s⁻¹. Increasing die land length at constant volumetric flow reduces die swell because the extended residence time under shear allows viscoelastic stress relaxation before the melt exits the die. A die land L/D ratio below 5:1 generally produces high recoverable strain and larger swell, while land L/D ratios above 10:1 are common for dimensionally stable filament extrusion. However, excessive land length raises pressure drop and melt temperature, especially in small-diameter filament dies, and can trigger degradation of the polybutadiene phase if melt temperature exceeds 250 °C at the die wall. Capillary rheometry data for commercial ABS grades at 230 °C typically show die swell ratios between 1.15 and 1.45 at apparent shear rates from 100 s⁻¹ to 1000 s⁻¹, with high-rubber impact grades at the upper end and low-rubber high-flow grades at the lower end. These values are usually generated with zero drawdown; production drawdown reduces final diameter but does not eliminate the underlying swell signal because the draw force acts on a melt that is still relaxing. The drawdown ratio should therefore be set so that the measured post-die swell does not consume more than 30 % of the available draw, otherwise small changes in melt strength from moisture or regrind will produce disproportionate final diameter variation.

Moisture analysis on ABS pellets is typically conducted by coulometric Karl Fischer titration following ASTM D6869-17 or ISO 15512:2019. Loss-on-drying moisture balances can overestimate volatile content because low-molecular-weight styrenic oligomers and mineral oil process aids volatilize along with water, producing readings that may deviate from Karl Fischer values by 0.05 wt% to 0.15 wt%. Desiccant wheel dryers for ABS are specified with a process-air dew point of -40 °C to -50 °C, a hopper inlet air temperature of 80 °C to 85 °C, and a residence time of 3 h to 4 h for virgin pellets fed from sealed containers. Pellets that have been exposed to ambient air at 60 % RH for more than 24 h may require 6 h or longer to reach a residual moisture of 0.03 wt% or less. The dryer itself becomes a process bottleneck when regrind is introduced because bulk density of ground ABS can be 30 % to 40 % lower than virgin pellets, reducing residence time in the hopper at the same hopper load. A dew-point transmitter installed in the return air line and a hopper discharge temperature probe are the minimum instrumentation needed to verify that the dryer is actively removing water rather than merely heating pellets.

If Regrind Content Exceeds 20 wt%, Melt-Pressure Fluctuations Amplify Die Swell Variability

Adding in-line regrind to virgin ABS alters particle shape, bulk density, and moisture re-uptake kinetics, and these changes interact with die swell through melt-pressure stability. Regrind particles possess high surface area and fractured edges, so they can adsorb surface moisture rapidly during storage even when the particle core remains dry. At regrind contents above 20 wt%, the non-uniform pellet size distribution promotes hopper segregation and irregular feed, which in turn produces screw-speed torque fluctuations and melt-pressure pulses that are independent of moisture. A gear pump can filter the low-frequency component of these fluctuations, but the high-frequency component caused by the melting behavior of regrind can pass through the pump and reach the die. The resulting melt-pressure variation modulates die swell at frequencies between 1 Hz and 10 Hz, overlapping the resonance band of the diameter control loop. Three practical control measures are standard in filament extrusion: regrind must be dried with the virgin polymer under the same dew-point and residence-time conditions; the regrind fraction should be metered gravimetrically rather than volumetrically to avoid density-driven composition shifts; and the die pressure signal should be monitored with a melt-pressure transducer mounted immediately before the breaker plate or screen pack, with the signal used as a feed-forward input to the puller-speed controller. Published data for the precise die swell amplification of specific ABS regrind blends is limited; however, production experience consistently identifies regrind moisture and particle variability as larger contributors to diameter variation than regrind content alone up to a threshold of approximately 30 wt%.

The following comparative matrix summarizes the typical response of ABS filament diameter control to three drying conditions. The values represent process ranges reported for general-purpose ABS rather than a single grade-specific data set; published data for specific configurations is limited when the die geometry and drawdown ratio differ from standard capillary rheometry conditions.

Diameter variation in ABS filament is measured at the finished spool by a two-axis laser gauge with reference standard calibration and off-line by digital micrometer applied to coiled sections conditioned at 23 °C and 50 % RH following ISO 23529 procedures. The in-line gauge is usually positioned 150 mm to 300 mm after the air knife or cooling trough exit, where surface temperature is below 60 °C and diameter is stable enough for feedback. Closed-loop control algorithms typically combine a slow screw-speed loop for throughput correction and a fast puller-speed loop for diameter correction, with the puller-speed loop update time between 10 ms and 100 ms. Die pressure is not typically used as a direct control variable, but it provides an early warning of moisture-induced instability because pressure fluctuations precede diameter fluctuations by the transport time from die to sensor. In practice, maintaining ABS pellet moisture below 0.03 wt% and dew point below -40 °C is more effective than increasing control-loop gain, because moisture-induced die swell variation is a distributed upstream disturbance that cannot be fully rejected by downstream draw-speed correction.

The following matrix identifies the test methods used to establish material and process compliance in ABS filament extrusion. These standards apply to both incoming resin qualification and annual in-line gauge calibration.

The operational boundary for stable ABS filament diameter control is therefore defined by three interrelated limits: residual moisture must remain below 0.03 wt%, dryer dew point must remain below -40 °C, and the die land L/D ratio must remain above 5:1 to reduce recoverable strain. If any one of these limits is violated, the resulting die swell variability cannot be compensated by routine closed-loop diameter correction alone. Incompatibility is also observed when ABS is combined with hygroscopic regrind streams that have not been dried under the same conditions as virgin resin, or when melt temperature exceeds 250 °C at the die wall and causes polybutadiene degradation. The diameter control system must therefore be treated as a subsystem of drying and die design, not as an independent corrective layer.