Experimental procedure for the characterization of cyclic behavior from very thin plate specimens

Many investigators, including those involved in the INTERNATIONAL BENCHMARK PROJECT ON SIMPLIFIED METHODS FOR ELEVATED TEMPERATURE DESIGN AND ANALYSIS - PROBLEM II, have tried to reproduce experimentally observed behavior by inelastic calculations. Unfortunately, the material characteristics used in the computer code were established from monotonic tensile tests performed with specimens extracted from the plate product itself (1.45 mm thick) employed to construct the ratchetting specimen. It now appears that the cyclic behavior of the material is much more relevant to the phenomenon observed. Hence the need to make this kind of characterization. Nevertheless, the practical problem is to produce cyclic stresses, i.e. tensile and compressive stresses, with very thin specimens. The main difficulty is to prevent the buckling effect. A new special device set up for this particular purpose is described here. The solution adopted was to create uniformly distributed alternative pure bending stresses in the thin plate specimen. Bending moments were produced by two end-grips fixed to the specimen, and these grips were mounted on a conventional test-machine which was displacement-controlled. To reduce tensile and compressive membrane stresses inside the specimen, the grips had two parallel axles of rotation. The forces produced by the machine and the displacements of a number of points of the specimen were continuously recorded during the test, so that cyclically stabilized, bending moments could be evaluated easily for each curvature variation imposed. The very first cyclic experimental data obtained, at room temperature, for the material of the sodium test specimen, a 316 type stainless steel, are presented. It may be noted that the simple specimens were very easy to prepare and hence inexpensive. (orig.)