Augmenting Material Characterisation via Numerical Analysis to Improve Robustness in Mechanical Property Evaluation

• Published in: Applied mechanics and materials Vol. 921; pp. 27 - 38
• Main Authors: Mathew, Samuel P., Venugopal, B., Mishra, R.K.
• Format: Journal Article
• Language: English
• Published: Zurich Trans Tech Publications Ltd 26.04.2024
• Subjects: Axial loads; Bias; Defects; Fatigue life assessment; Material properties; Mechanical properties; Medium carbon steels; Misalignment; Numerical analysis; Random errors; Strain; Stress-strain relationships; Tensile strength; Twisting; ASTM Tensile Test; Mechanical Property Evaluation; Design Allowances; Misalignment; Imperfections
• DOI: 10.4028/p-ZhMC5n
• ISSN: 1660-9336; 1662-7482

Experimental evaluation of mechanical properties of materials is often standardised by pre-defining the test specimen in terms of its geometrical features and random errors if any need to be screened out during experimentation. However, eliminating the systematic biases is a formidable task. This study presents the efforts taken to address three important systematic biases possible in the case of mechanical property evaluation of materials. For a simple tensile strength characterization, misalignment of the specimen with respect to the axis of loading, turning effect in the load application system and geometrical imperfection are considered in the test specimen. These concepts are illustrated using numerical analysis for SAE 1045 steel material. Sheet specimens as per the ASTM standards are modelled and the stress-strain behaviour of the material is bench marked with experimental results. Subsequently, pre-defined cases of (a) specimen misalignment, (b) twisting and (c) geometric imperfections are introduced to study the variations in stress-strain behaviour. It is observed that an inadvertent twisting force coupled with an axial load increases the Von-Mises stress at the mid-section of the specimen increased by about 33% and reduces its fatigue life by 96%. The study clearly brings out the implications of such inadvertent systematic biases occurring in a typical experimental or usage scenario on the component life. It also shows how the numerical computations can offer a robust methodology to assess the bounds of possible deviations.