Structural steel has been used since the early 1970’s in Korea as primary structural members such as columns, beams, and trusses. The materials have much higher strength such as fast construction, high load bearing capacity, high construction quality but those have a fatal weakness as well. Load-bearing capacity is going down when the structural members are contained in fire condition. Therefore, to protect the structural members made of steels from the heat energy the fire resistance performance required. Generally, the fire resistance performance have evaluated from the exact fire tests in fire furnaces. But the evaluation method takes much more time and higher expenses so, the engineering method requires. The engineering method not only adopts a science but also an engineering experience. In this paper, to make various data-bases for evaluation of structural members such as columns(H-section, RHS), beams, loaded fire tests were conducted and derived not only each limiting temperature but also fire resistance respectively.
Explainable machine learning using real, synthetic and augmented fire tests to predict fire resistance and spalling of RC columns
