Assessment of fracture toughness for polyethersulphone injection mouldings

• Main Author: Whitehead, Richard Donaldson
• Published: University of Surrey 1988
• Subjects: 668.4; Thermoplastic fracture testing
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.382969

This project assessed fracture toughness (K[IC] and G[IC]) for thin-section injection-moulded PES and for which there are no standard fracture-mechanics tests. Standard methods of testing are based on thick specimens that are easily available for metals but unusual for injection mouldings. Tests were, therefore, made which adapted the standard Mode-I (crack-opening) tests to the thin sections of injection-moulded discs. Then the question of the tests' validity arose. Were the elastic assumptions being met? If both critical Mode-I stress-intensity factor and strain-energy release rate were found then a Young's modulus value could be derived. If this derived modulus when compared to the directly measured Young's modulus was found to be identical then the assumption of linear elasticity would be upheld. Two different styles of fracture-mechanics test were investigated: ones with explosive fracturing and one with controlled crack growth. In explosive fracture-mechanics tests, the fracturing of the testpiece was so rapid that the minimum data was collected, so many such tests had to be aggregated to sample the range of notch-depth possible. This was true for both the three-point-bend (TPB) and tensile tests made on single-edge-notched (SEN) bars cut from the discs. However, when the derived modulus for the explosive tests was compared with Young's modulus it was found to be smaller, so these tests did not uphold the assumptions of elasticity. The reason for this failure was found to be that the notch did not act like the sharp propagating crack required for successful tests. In the controlled crack-growth test sequence, a crack was grown across the diameter of an injection-moulded disc and measurements were made, as the crack progressed, of the crack's position, corresponding loading and work done. From this sequence of measurements, independent values of stress-intensity factor and strain-energy release rate, under quasistatic conditions, were calculated that produced values of derived modulus which were, at least, equal to the Young's modulus, thus upholding the elastic integrity of the test. Moreover, initial surges of crack-growth (similar to explosive fracturing) were shown by linear regression analysis to give larger stress-intensity factors and to be significantly different from the data collected only during controlled crack-growth. Thus, the use of controlled crack-growth is Justified as a more reliable method of fracture-mechanics testing. Yield stress measurements were also made and with the valid stress-intensity factor predicted plastic zone sizes entirely compatible with yielded regions observed as crazes whose lengths were reported by Hine.