Experimental investigation and analytical prediction of viscoelastic behavior of additively manufactured (AM) acrylonitrile butadiene styrene (ABS) parts

Purpose This study aims to characterize temperature (−130°C to 130°C) and strain-rate (0.00352–3520 s−1) dependent viscoelastic behavior of material extrusion (MEX)-fabricated acrylonitrile butadiene styrene, and develops hybrid decompose/shift/gap-reduction-ratio/reconstruct (DSGR) predictive model to capture storage modulus evolution and extend reliable predictions for demanding service conditions. Design/methodology/approach Thermal transitions are characterized using differential scanning calorimetry (DSC) to establish a processing-relevant thermal baseline. Viscoelastic behavior is measured through dynamic mechanical analysis at 0.1, 1 and 10 Hz (3.52 × 10−3, 3.52 × 10−2 and 3.52 × 10−1 s−1) to capture α- and β-relaxations with strain-rate-dependent shifts. Predictions beyond these ranges are made using decompose/shift/reconstruct (DSR) and new hybrid DSGR methods, which are validated against additional 100 Hz experiments. Findings Thermal characterization reveals glass transition at 107.5°C (heating) and 113.8°C (cooling), exhibiting hysteresis and enthalpic recovery from frozen-in stresses. Viscoelastic characterization identifies strain-rate-dependent shifts in α-transition (106.1°C → 117.8°C) and β-transition (−73.5°C → −65.0°C) with storage modulus drop by three orders of magnitude across α and by approximately 1.1 GPa across β. A hump/peak in storage modulus confirms stress-release consistent with DSC. DSR shows high accuracy outside Tg (R = 0.9977, AARE = 7.4%) while within Tg, newly developed GR model captures hump/peak (R = 0.9895, AARE = 3.62%). Hybrid DSGR framework integrates GR and DSR, which achieves high accuracy (R = 0.9954, AARE = 6.90%) across full temperature domain and extends predictions up to 3250 s¹. Originality/value DSGR provides a robust predictive framework that links thermal history with viscoelastic performance and allows accurate prediction across broad temperature and strain-rate regimes that can be further applied for design of MEX components from rigid load-bearing to impact-resistant applications.