Thermal ratcheting and progressive buckling

Pure elastic buckling is not a frequent mode of failure and plastic deformations often occurs before buckling - like instability does. Elastic-plastic buckling is very difficult to analyse. The most important difficulty is the material modeling. In the elastic plastic buckling phenomena, small modifications of the material constitutive equation used are of great influence on the final result. When buckling cannot occurs, it is well known that distortion due to applied loads is greatly amplified when there is also some cyclic straining (like thermal stresses). This effect is called ratcheting - and thermal ratcheting when caused by cyclic thermal transients. As cyclic thermal stresses can be applied in addition of load able to cause buckling failure of a component, the question arise of the effect of cyclic thermal stresses on the critical buckling load. The aim of the work presented here is to answer that question: 'Is the critical buckling load reduced when cyclic straining is added'. It seems sensible to avoid premature computation based only on arbitrary assumptions and to prefer obtaining a sound experimental basis for analysis. Sufficient experimental knowledge is needed in order to check the validity of the material modeling (and imperfections) used in analysis. Experimental tests on buckling of compressed columns subjected to cyclic straining have been performed. These experiments are described and results are given. The most important result is cyclic straining reduces the critical buckling load. It appears that distortion can be increasing progressively during cyclic straining and that buckling can happen at last at compressive loads too small to cause buckling in the absence of cyclic straining. (orig./RW)