Mixed numerical‐experimental method for generation of energy‐life fatigue master curves (original) (raw)

2019, Material Design & Processing Communications

Advanced high-strength steels are outstanding materials to meet the current demands of automotive industry, including the reduction of vehicle weight. These high value-added products, which operate generally under adverse service conditions, are highly susceptible to fatigue failure. Modern fatigue design approaches, namely, those based on the strain energy density concept, require detailed information about the energy-life relationship. This paper proposes a straightforward methodology to generate energy-life fatigue master curves on the basis of only two uniaxial low-cycle fatigue tests and two elastic-plastic finite-element models. The methodology is successfully tested in an advanced high-strength steel. KEYWORDS fatigue master curve, low-cycle fatigue, strain energy density 1 | INTRODUCTION Advanced high-strength steels are outstanding materials to meet the current demands of automotive industry, namely, the reduction of vehicle weight and also reduction of fuel consumption. These steels, which pose superior properties in terms of strength and ductility, have been developed to compete with various lightweight materials on the basis of cost, performance, and manufacturability. 1 However, due to the combination of stress concentration phenomena and timevarying loading, these high value-added products are susceptible to fatigue failure. 2-4 Therefore, detailed information on the fatigue response of these materials is pivotal to ensure a safe long-term life. Modern fatigue design models are generally formulated based on local stress-based, strain-based, and energy-based approaches. 5,6 Within the energy-based formulations, those based on the strain energy density concept have gained increased popularity. 7-9 This can be explained by the fact that they can handle different materials, loading scenarios, and geometries and because they require a balanced number of material properties. 9-14 Despite this, the determination of such properties is time-consuming and expensive. In this sense, the development of faster and low-cost approaches would be of great importance to simplify the procedures and to make these energy-based prediction models even more attractive. In this paper, it is proposed a straightforward methodology to generate energy-life fatigue master curves from only two uniaxial low-cycle fatigue tests. Here, two fully reversed strain-controlled tests, with strain amplitudes of 0.6% and 2.0%, were performed using standard cylindrical specimens made of a DIN 34CrNiMo6 high-strength steel. From these two tests, the Basquin and Coffin-Manson relationships were determined by linear regression, aiming at obtaining a strain-life relationship. Parallelly, two independent elastic-plastic numerical models were developed to describe the cyclic behaviour of the material for each strain-amplitude. These two models were then used to generate stable