Austenitic stainless steels are promising engineering materials demonstrating good corrosion resistance and good formability but they have also relative low yield strength. Among the different strengthening mechanisms, grain refinement is the only method to improve both strength and toughness simultaneously. Metastable austenitic stainless steels undergo a strain-induced martensitic transformation, where the metastable austenite phase is transformed to the thermodynamically more stable ??-martensite phase due to the plastic deformation. The cold reductions at -15°C and 20 °C of 10 - 90% were carried out with inter-pass cooling on AISI 304L stainless steel. In order to obtain homogeneous austenitic microstructures, 90% deformed samples were annealed at different temperatures (700 °C – 900 °C) and times (5 second – 8 hour). Mechanical testing was performed by means of Vickers Hardness. The experimental measurements using Ferritescopy, X-ray diffraction, scanning electron microscopy and optical metallography were performed. This thesis concentrated on the effects of the strain-induced martensitic transformation and also the effects of annealing time and temperature on the reversion of austenite to martensite in order to grain refinement producing nano/submicron austenitic structure. These results were used for simulation with artificial neural networks and through this attempt, 6 different networks for prediction of structure and mechanical properties designed. The resulting diagrams can be used as a guide maps to help finding appropriate heat treatment considering the structure phases, grain sizes and mechanical properties.
