| Summary: | 碩士 === 國立中興大學 === 土木工程學系所 === 101 === Effect of welding method on the fire behavior of concrete filled box columns
ABSTRACT
Keywords: complete penetration weld, partial penetration weld, concrete-filled box column, in fire
1. Background
Nowadays the requirements of structural fire protection contained in most building codes represent the minimum levels of fire safety deemed acceptable to their countries. In order to provide sufficient time for occupants on the floors above the fire floor to reach an area of safety; to support the fire separations necessary to control the overall size of the fire and prevent conflagration; and to minimize potential damage to adjacent properties, most codes normally require that load bearing elements and assemblies (walls, columns and beams) have a fire resistance rating at least equivalent to that required for the supported assembly (floor or roof). Steel reinforced concrete columns have been used on various building projects with great advantage throughout the world in recent years. To promote the applications of concrete filled box columns (CFBC) in the domestic building and construction industry, this study aims to examine the fire behavior of CFBC fabricated by complete penetration weld and partial penetration weld, respectively.
2. Research method and process
A series of two full-size experiments were carried out to consider the effect of welding method on the fire resistance of CFBC. CP1 specimen was fabricated by complete penetration weld while PP1 specimen was fabricated by partial penetration weld. The tested columns were all square section 500 mm × 500 mm with 4350 mm hige and were filled with plain concrete. All of the steel sections of the specimens were fabricated from steel plate of 22 mm wall thickness. These plates were joined by longitudinal fillet welds at the vertices. Each of the CFBC had end plates welded to them in order to transfer the load, and end stiffeners were also introduced to ensure that end conditions did not affect the failure resistance of thermal load. Besides, the temperature from the specimen''s surface to the inner central core was measured with type K thermocouples located at different depths in four sections of the column. The columns were subjected to a constant compressive load, during the whole test, of 4969 kN. This load was controlled by a load cell of 19.6 MN, located on the head of the piston of a jack. The applied load corresponded to 28% of the design value of buckleing resistance of the column at room temperature. Thermal load was applied on the columns in form of CNS 12514 time-temperature curve in a natural gas-fired large-scale laboratory furnace untill the set experiment termination condition was reached. On the other hand, a numerical model was established for predicting temperature distribution at surface and inner portions of the CFBC at elevated temperatures, and thus making a comparison with the experimental values.
3. Important discoveries
Two concrete-filled box columns with shear studs were tested to failure by exposing the columns to fire. No external fire protection was provided for the steel. During the test, the column was exposed to heating controlled in such a way that the average temperature in the furnace followed, as closely as possible, the standard time-temperature curve of CNS 12514. These two columns had similar characteristics except welding method and were subjected to similar load levels. As expected, the columns expand in the initial stages and then contract leading to failure. The deformation in these columns results from several factors such as load, thermal expansion and creep. The effect of load and thermal expansion is significant in the early stages, while the effect of creep becomes pronounced in the later stages. Results from the fire tests indicate that the fire resistance of CP1 specimen is about 55 minutes, as compared to about 51 minutes for PP1 specimen. As a result, it can be concluded that under a lower load ratio of 0.28, the effect of welding method on the fire resistance of CFBC is not significant. Inaddition, the established numerical model was able to reasonably predict the temperature distribution in time history on the specimen cross section.
|
|---|
