Prediction of SENB Fracture Toughness From Charpy Data Using the Beremin Model in the Lower Transition Region

Abstract

This work focuses on the application of a mechanistic local approach model to describe the statistical distribution of experimental Charpy (CVN) impact test data obtained at several temperatures in the ductile to brittle transition temperature range. The current objective is to develop a correlation in the lower transition regime between quasi-static CVN absorbed energy (CVE) and the J-integral fracture toughness (Jc) obtained from deeply pre-cracked Charpy (PCCVN) specimens tested quasi-statically to laboratory test standards. The Beremin model for cleavage fracture has been applied to a ferritic steel which has been comprehensively tested using standard CVN, shallow U-notched and PCCVN specimen types in the lower ductile to brittle transition. This has enabled a prediction to be made of the absorbed CVE at cleavage fracture initiation for a Charpy specimen tested quasi-statically in the lower part of the CVN transition curve. By applying the Beremin model to PCCVN single edge notch bend specimens at quasi-static rates it was possible to use the Weibull stress, to achieve a reliable correlation between CVE and Jc in the lower ductile to brittle transition region. The results from this work indicate that the Beremin model can provide a theoretically based correlation for CVE to Jc fracture toughness for a ferritic steel under quasi-static loading conditions. The overall objective of the project remains to predict dynamic CVN absorbed energy using micromechanical modelling and which is valid for all ferritic steels.

Numerical and Experimental Results on Charpy Tests for Blends Polypropylene + Polyamide + Ethylene Propylene Diene Monomer (PP + PA + EPDM)

“…Reviewing the available literature on this type of tests [ 7 , 8 , 9 , 10 ] and on Charpy impact modeling [ 11 , 12 , 13 , 14 , 15 , 16 , 17 ] pointed out the following aspects: models are using very different mesh web rougher or finer (especially near the notch) [ 18 , 19 , 20 ], on a quarter or half of the sample, from 2D to 3D [ 21 ], but new achievement in hardware and dedicated software allows for using finer mesh, all the sample, investigation on strain and stress, simulation in actual time etc. ; first models did not take into account friction, but recent ones have included condition for friction, usually with a constant friction coefficient; at the beginning, constitutive models were simpler, implying only a failure criterion, usually the strength limit in tensile and there has been interest in performant materials like steel, titanium and aluminum alloys [ 22 , 23 ]; for polymers and especially for blends fewer research reports being [ 24 , 25 , 26 , 27 ]; and each model has simplifying hypotheses that could alter the sample behavior as compared to the actual one.…”