An improved Johnson–Cook constitutive model for flow stress prediction of 92W–5Co

An improved Johnson–Cook constitutive model for flow stress prediction of 92W–5Co–3Ni alloy

Poluru, Suswanth / Kotkunde, Nitin / Singh, Swadesh Kumar / Panchal, Ashutosh / Gnanasambandam, Prabhu

A trustworthy prediction of flow stress behaviour is essential to optimise the hot working process parameters. It also helps accurately capture the finite element simulations of many complex processes. In this work, modification in the Johnson–Cook (JC) model has been proposed for better prediction of the flow stress behaviour of the 92W–5Co–3Ni alloy. Initially, uniaxial compression tests were conducted at different strain rates (1 s⁻¹, 25 s⁻¹, 50 s⁻¹, 75 s⁻¹, and 100 s⁻¹) and temperatures (323 K, 473 K, 673 K, 873 K) using Gleeble-3800 thermo-mechanical simulator. It was confirmed that flow stress variation is sensitive to both strain rate and temperature change. Subsequently, various microstructural parameters were evaluated, such as grain size, tungsten–tungsten contiguity (W/W contiguity), tungsten–tungsten connectivity (W/W connectivity), dihedral angle, neck length, solid volume fraction, and matrix volume fraction. Afterwards, the phenomenological-based constitutive models, namely, Johnson–Cook (JC) and modified Johnson–Cook (m-JC), were initially established. The analysis of flow stress prediction based on various statistical parameters revealed that both models demonstrate poor flow stress prediction capabilities with correlation coefficient (R) of 0.7715 and 0.7925, respectively. An improved Johnson–Cook model (i-JC) was proposed, replacing the strain term with the Ludwigson hardening equation and varying the coefficient of strain rate hardening term with plastic strain and strain rate. The i-JC model significantly improved the accuracy of flow stress prediction with a correlation coefficient (R) of 0.9891, average absolute relative error (AARE) of 1.35%, and standard deviation of 1.33%.