04.03: Buckling behaviour of axially compressed cylindrical shells made of stainless steel

Abstract

ABSTRACTCylindrical shells made of stainless steels are widely used, e.g. in tanks and biogas plants [1]. Whereas austenitic stainless steels were commonly used in the past, austenitic‐ferritic and lean duplex stainless steels are more frequently used nowadays. By contrast with structural steels, stainless steels exhibit a significantly nonlinear stress‐strain behaviour which must be considered in the buckling analysis of these shells because the reduced material stiffness below the proof stress may cause premature buckling. For this reason, EN 1993‐1‐6 [2] indicates the use of either a reduced value of elastic modulus Ered or the secant modulus at the 0.2% proof stress when assessing the buckling resistance of shells with nonlinear stress‐strain curves. These provisions may produce over‐conservative low buckling strengths for stainless steel shells at both low and high slendernesses. Only stainless steel shells with medium slenderness are covered by this rule.For this reason, additional temperature‐ and slenderness‐dependent buckling strength correction factors for austenitic stainless steel axially loaded cylindrical shells were developed by Hautala and Schmidt (1999) [3,4] which were described in the ECCS Design Recommendations [5] for buckling of steel shells but are not yet implemented into EN 1993‐1‐6. Recent experimental, numerical and theoretical research, carried out within the European RFCS research project BiogaSS on austenitic‐ferritic and ferritic stainless steels has investigated the potential to apply these buckling strength correction factors to austenitic‐ferritic and ferritic stainless steels. This paper shows that shells made of austenitic‐ferritic and ferritic stainless steels have a less serious deviation from the reference mild steel than austenitic stainless steel, permitting less significant buckling strength correction factors.The results obtained for all families of stainless steels should be transformed into the generalized design verification format of EN 1993‐1‐6, in which a parameterized capacity curve [6,7] is used. This paper also shows how the correction factors devised for axially loaded shells made of austenitic stainless steel can be better represented by an appropriate characterisation using the capacity curve alone. The outcome should be relevant to austenitic, austenitic‐ferritic and ferritic stainless steels when sufficient data is available. The paper also illustrates how behaviours of considerable complexity in some shell buckling situations can still be represented by the unaltered universal capacity curve [6,8].