Oxidation Damage Evolution in Low-Cycle Fatigue Life of Niobium-Stabilized Austenitic Stainless Steel.

Austenitic stainless steel is a vital material in various industries, with excellent heat and corrosion resistance, and is widely used in high-temperature environments as a component for internal combustion engines of transportation vehicles or power plant piping. These components or structures are required to be durable against severe load conditions and oxidation damage in high-temperature environments during their service life. In this regard, in particular, oxidation damage and fatigue life are very important influencing factors, while existing studies have focused on materials and fracture behavior. In order to ensure the fatigue life of austenitic stainless steel, therefore, it is necessary to understand the characteristics of the fracture process with microstructural change including oxidation damage according to the temperature condition. In this work, low-cycle fatigue tests were performed at various temperatures to determine the oxidation damage together with the fatigue life of austenitic stainless steel containing niobium. The characteristics of oxidation damage were analyzed through microstructure observations including scanning electron microscope, energy-dispersive X-ray spectroscopy, and the X-ray diffraction patterns. In addition, a unified low-cycle fatigue life model coupled with the fracture mechanism-based lifetime and the Neu-Sehitoglu model for considering the influence of damage by oxidation was proposed. After the low-cycle fatigue tests at temperatures of 200-800 °C and strain amplitudes of 0.4% and 0.5%, the accuracy of the proposed model was verified by comparing the test results with the predicted fatigue life, and the validity by using the oxidation damage parameters for Mar-M247 was confirmed through sensitivity analysis of the parameters applied in the oxidation damage model. As a result, the average thickness of the oxide layer and the penetration length of the oxide intrusion were predicted with a mean error range of 14.7% and 13%, respectively, and the low-cycle fatigue life was predicted with a ±2 factor accuracy at the measurement temperatures under all experimental conditions.

Restored Action Generative Adversarial Imitation Learning from observation for robot manipulator.

In this paper, a new imitation learning algorithm is proposed based on the Restored Action Generative Adversarial Imitation Learning (RAGAIL) from observation. An action policy is trained to move a robot manipulator similar to a demonstrator's behavior by using the restored action from state-only demonstration. To imitate the demonstrator, the trajectory is generated by Recurrent Generative Adversarial Networks (RGAN), and the action is restored from the output of the tracking controller constructed by the state and the generated target trajectory. The proposed imitation learning algorithm is not required to access the demonstrator's action (internal control signal such as force/torque command) and provides better learning performances. The effectiveness of the proposed method is validated through the experimental results of the robot manipulator.