Structural design for non-linear metallic materials

Abstract

The material stress-strain behaviour of structural carbon steel may be suitably accurately reflected for design purposes by an idealised elastic, perfectly-plastic material model; such material behaviour lends itself to the concept of section classification. There are, however, a number of structural materials, such as aluminium, stainless steel and some high strength, cold-worked steels, where this idealised model becomes inaccurate due to non-linearity of the stress-strain response below the yield point and considerable strain hardening beyond the yield point. Resulting design methods, developed on the basis of the idealised material behaviour, are necessarily overly conservative. A new method has been developed that utilises a more accurate material model and a continuous measure of cross-section deformation capacity (rather than the discretised system of section classification) to provide more rational and efficient designs. This paper describes the basis for the proposed design method and presents a comparison with results obtained from laboratory testing and those predicted by the current Eurocode approach. The proposed design method offers average increases in member resistances of around 20% over the current Eurocode approach, and a reduction in scatter of the prediction.

Flexural behaviour of stainless steel oval hollow sections

“…This has been highlighted in previous studies investigating the ultimate capacity of stainless steel cross-sections and members and an alternative design method, termed the continuous strength method (CSM) has been developed [22,37,50] and statistically validated [51]. The CSM has also been applied successfully to structural carbon steel [52].…”

“…The symbols E, σ0.2, σ 1.0, σ u, εf, n and n'0.2,1.0 used in Tables 1 and 2 refer to Young's modulus, 0.2% proof stress, 1% proof stress, ultimate tensile stress, strain at fracture, Ramberg-Osgood strain-hardening parameters [22] below and above the 0.2% proof stress, respectively. These results are subsequently utilised during the analysis of the three-point bending tests and in the development of numerical models.…”

“…The adopted material model is a compound Ramberg-Osgood [33,34] formulation, with the second stage of the model utilising the 1.0% proof stress σ1.0. Two-stage material models for stainless steel have been developed and studied by a number of authors [22,[35][36][37][38]. The material parameters given in Tables 1 and 2 were averaged for each nominal cross-section size and applied uniformly to each model in the true stress true -log plastic strain pl ln  format, as required by ABAQUS [32] and defined by Eqs.…”

Description of stress–strain curves for stainless steel alloys

“…Ashraf [9] in order to improve the accuracy of the model at low strains (less than approximately 10%) and to allow the model to be applied also to the description of compressive stress-strain behaviour. The modifications involved use of the 1% proof stress instead of the ultimate stress in the second stage of the model, leading to Eq.…”

“…This requirement led to the development of three stage versions of the Ramberg-Osgood formulation: Quach et al [11] proposed a material model that uses the basic Ramberg-Osgood curve (Eq. (1)) for the first stage, covering stresses up to the 0.2% proof stress, the Gardner and Ashraf [9] model (Eq. (13)) for the second stage covering stresses up to the 2% proof stress and a straight line from the 2% proof stress to the ultimate strength for the third stage.…”