Fused deposition modelling (FDM) is one of the most popular additive manufacturing (AM) methods, FDM can produce complex shapes at low cost in a short time, Polyethylene terephthalate glycol (PETG) is a popular FDM material which has increasingly been used due to its ease for printing, good chemical resistance, strong durability and odourless. Since material performance of the parts printed by FDM is lower that that using traditional manufacturing methods, mainly due to presence of voids and weak bonding between filaments. In addition, difference in the printing parameters also affects the FDM performance. It is therefore necessary to compare and characterize performance of PETG prepared using different manufacturing methods and, for the FDM method, using different printing parameters.

In this study, a newly developed test method, named multi-relaxation (MR) test, is used to characterize the relaxation behaviour of PETG specimens. Five types of PETG specimens were manufactured for the testing, including the conventionally extruded PETG, and 3D printed PETG specimens of [0°], [±30°]_s, [±45°]_s and [90°] using FDM. The MR test results were analyzed using Eyring’s model with four parameters to describe the relaxation behaviour at fixed strokes. These parameters are initial viscous stress at the onset of relaxation, σ_r(0), reference stress, σ₀, time-independent quasi-static (QS) stress, σ_st, and relaxation time, τ_r. For [90°] specimen, however, only the QS elastic modulus (E_x) could be determined due to its brittle behaviour, which was determined using both MR test and monotonic tensile test. For the other four types of PETG specimens, two types of specimens, i.e., with and without holes in the tab sections, were evaluated. The MR test results suggest that the holes in the tab sections can significantly reduce specimen performance, especially for [0°] and [±30°]_s specimens. Therefore, specimens without holes were used to evaluate the influence of raster angle on the relaxation behaviour of PETG.

Comparisons of MR test results among different types of specimens suggest that [0ᵒ] specimens are less stiff than the extruded specimens, but after removing the effect of voids in the former, these two types of specimens show very similar relaxation behaviour. Among 3D printed specimens, their stiffness is in the order of [0ᵒ] > [±30°]_s > [±45°]_s . Those 3D printed specimens also showed different fracture behaviours. In particular, [0ᵒ] specimens did not show any sign of fracture in the test, but [±30°]_s and [±45°]_s specimens started fracturing during the middle of the test.

For each type of PETG specimens, the relationship between total applied stress and its QS and viscous stress components and the variation of relaxation parameters (σ₀ and τ_r) with stroke were investigated. Three transition points were detected in the deformation process. Variation of applied stress, QS stress, σ_r (0), σ₀ and τ_r among PETG 3D printed specimens was investigated, which indicated that the [±30°]_s specimens showed a different profile of stress-stroke curve after the second transition point from the other 3D printed specimens. The corresponding values for σ_r (0), σ₀ and τ_r after the second transition point are also different from their values for the other 3D printed specimens.

The study also examined applicability of the classical laminate theory (CLT) to prediction of the QS elastic modulus (E_x) for 3D printed specimens. The results suggest that the CLT is applicable to the E_x prediction, but the prediction accuracy could be improved by consider the effect of voids on the mechanical performance.