Gaussian And Non-Gaussian Distributions Of Fracture Properties In Tensile Stretching Of High-Density Polyethylene

The statistical fracture behavior and microstructure of two high-density polyethylene samples with different molecular weights (HDPE-151K and HDPE-932K) were investigated as a function of thermal history conditions utilizing engineering stress-strain measurements, the small-angle X-ray scattering technique, and H-1 NMR T-2 relaxometry. The probability density functions of elongation at break and fracture toughness were used to characterize the scattered fracture properties. It was found that the distributions of elongation at break and fracture toughness for the quenched sample of HDPE-151K are well fitted with a Gaussian function, whereas both annealed and isothermally crystallized counterparts demonstrate a non-Gaussian distribution of such fracture parameters. In the case of HDPE-932K, a stochastic fracture process is observed for the isothermally crystallized sample. Furthermore, the tie molecule concentration, the phase composition, and the T-2 relaxation time were quantitatively evaluated for both systems. The results indicated that a structural defect dominates the fracture behavior in HDPE systems treated at high temperatures. For HDPE-151K, the structural defect originates from the low fraction of tie molecules and thus a low number of stress transmitters after thermal treatment. In the case of HDPE-932K, the structural defect is caused by the less mobile amorphous network as less mobility of the network chains means a less effective transfer of stress that is produced when the sample is elongated. Due to the globally interlocked network being built up by crystalline lamellae and amorphous chains, a portion of structural defects can be stabilized during tensile stretching, thus resulting finally in non-Gaussian distribution of fracture features.