The Finite Element Model Updating (FEMU) technique is an inverse method that enables to arrive at a complete solution to the problem of diffuse necking of a thick tensile specimen. Conventionally, FEMU relies on the identification of a phenomenological strain hardening law that inherently limits the accuracy of the method due to the predefined character of the adopted strain hardening law. A high-resolution multi-linear post-necking strain hardening model enables to describe more generically the actual strain hardening behaviour. A numerical concept study is used to scrutinise the identification of such a model using FEMU. It is shown that, unlike progressive identification strategies, a global identification strategy followed by a smoothing operation based on area conservation yields sufficiently accurate results. To study the experimental feasibility, the latter strategy is used to identify the post-necking strain hardening behaviour of a thick S690QL high-strength steel. To this purpose, a notched tensile specimen was loaded up to fracture, while the elongation was measured using Digital Image Correlation (DIC). It is shown that the global identification strategy suffers from experimental noise associated with DIC and the load signal.