17-4 PH alloy is a precipitation hardening stainless steel with a desirable combination of high corrosion resistance and good mechanical properties, achieved through appropriate age hardening treatment. The aim of this research was to obtain fatigue curves for two common age hardening cycles, known as H900 and H1150, and to study the conditions for optimum heat treatment of this alloy. For this purpose, 17-4PH stainless steel in the form of 9.8 mm diameter rods was used as the experimental material. To ensure the uniformity of heat treatment condition, the as-received alloy was solution annealed at 1050 ?C for 30 min followed by oil quenching at room temperature. To study the age hardening behavior of the alloy, the solution annealed specimens were aged in the range 400 to 600 ?C up to 240 min and age hardening curves were obtained. To study the variation of hardness in the ageing process, Artificial Neural Network (ANN) was applied. Ageing at 464 ?C for 129 min was determined by a combination of ANN and Genetic Algorithm (GA) to achieve maximum hardness value; this heat treatment cycle was called A460-2. Ageing processes were applied to 17-4PH stainless steel specimens based on A460-2, A480-1 (H900), A620-4 (H1150), and SA (solution annealed) conditions. Microstructural examinations and hardness measurements were carried out and, in addition, other mechanical properties including tension and fatigue tests were performed. Results of tensile tests showed that A460-2 condition with 44 Rockwell C hardness and Ultimate tensile strength of 1467 MPa produced the optimum hardness and strength in 17-4PH stainless steel. Data obtained from S-N curves revealed endurance limits between 525 to 700 MPa in the investigated specimens. It was concluded that the fatigue limit of 17-4PH stainless steel increased as the toughness, thus, optimum ageing cycle must be determined, accordingly. Key words: 17-4PH stainless steel, age hardening, Artificial Neural Network, Genetic Algorithm
