Seismic Performance of Fully Bolted Through-Diaphragm Connections in Square CFST Column-Steel Beam Joints

Literature Overview

This experimental study by Li Zilin, Ding Hongyi, and Xue Jiang, published in the China Civil Engineering Journal in 2014, investigates the seismic performance of fully bolted through-diaphragm connections between square concrete-filled steel tube (CFST) columns and steel beams. The research was supported by the Tianjin Natural Science Foundation (Grant No. 13JCBJC19600) and the National Basic Research Program of China (973 Program, Grant No. 2011CB013603). Four full-scale specimens were tested under low-cycle reversed loading to evaluate the ultimate bearing capacity, failure modes, ductility, and energy dissipation capacity of the connection.

Core Technical Content

The fully bolted through-diaphragm connection is a moment-resisting connection that uses a through-diaphragm plate welded to the CFST column to transfer bending moments and shear forces from the steel beam to the column. The connection is designed to be fully bolted, meaning that all load transfer is accomplished through bolted connections rather than field welding, which simplifies construction and reduces the risk of welding defects.

Test Specimen Configuration

Four full-scale specimens were tested, with variations in three parameters:

The test specimens consisted of square CFST columns with through-diaphragm plates connected to steel beams via bolted connections. The CFST columns were filled with concrete of specified strength grade, and the steel beams were fabricated from structural steel with specified yield strength.

Test Results and Key Findings

The experimental results revealed several important characteristics of the connection behavior:

Failure Modes: The failure mode was found to be highly dependent on the presence of beam eccentricity:

• For specimens with beam eccentricity: Failure occurred at the corner bolt holes of the diaphragm bolt group due to tensile tearing. This is a brittle failure mode that is undesirable in seismic design.
• For specimens without beam eccentricity: Failure occurred by tensile fracture of the diaphragm plate itself. This is also a brittle failure mode but indicates that the diaphragm plate is the weakest link in the connection.

Bearing Capacity: The bearing capacity of the connection was found to be significantly affected by the beam cross-section size:

• Increasing the beam height from 600 mm to 700 mm increased the bearing capacity by approximately 50%.
• The presence or absence of diaphragm welding had no significant effect on bearing capacity.

Ductility: The ductility of the connection was evaluated using the ductility coefficient, defined as the ratio of displacement at maximum load to displacement at yield:

• Specimens with larger beam cross-sections exhibited better ductility.
• Beam eccentricity reduced the ductility coefficient to approximately 85% of that of the concentric specimen.
• Diaphragm welding had a minor effect on ductility.

Energy Dissipation: The energy dissipation capacity was evaluated using the energy dissipation coefficient:

• Welded diaphragm specimens exhibited energy dissipation coefficients 16-37% lower than non-welded diaphragm specimens.
• This indicates that welding defects can significantly reduce the energy dissipation capacity of the connection.

Hysteresis Loop Characteristics

The hysteresis loops of all specimens were relatively full, indicating good energy dissipation capacity. However, the loop shape was affected by the beam eccentricity and diaphragm welding:

• Concentric specimens exhibited symmetric hysteresis loops with stable load-displacement response.
• Eccentric specimens exhibited asymmetric hysteresis loops with reduced post-peak strength degradation.
• Welded diaphragm specimens exhibited more pinched hysteresis loops, indicating reduced stiffness and energy dissipation.

Welding Quality and Connection Performance

From a welding engineering perspective, this study provides critical insights into the effect of welding quality on connection seismic performance. The finding that welded diaphragm specimens exhibited 16-37% lower energy dissipation coefficients is particularly significant because it demonstrates that welding defects can have a substantial impact on the seismic performance of connections.

Welding Defects and Their Effects

The following welding defects can affect connection performance:

The study suggests that even minor welding defects can reduce the energy dissipation capacity of the connection by up to 37%. This is a significant finding because it implies that the connection may not meet the intended seismic performance objectives if welding quality is not carefully controlled.

Welding Process Considerations

For the through-diaphragm connection, the following welding considerations are essential:

1. Welding sequence: The welding sequence should be planned to minimize residual stresses and distortion. Symmetrical welding from the center outward is recommended to reduce angular distortion.
2. Heat input control: The heat input should be controlled to prevent excessive grain growth in the heat-affected zone (HAZ) and to minimize the risk of hydrogen-induced cracking. For structural steels with yield strength above 355 MPa, heat input should be limited to 20-25 kJ/mm.
3. Preheat and interpass temperature: Preheat temperatures of 80-100°C and interpass temperatures of 150-200°C are recommended to prevent hydrogen-induced cracking and to control the cooling rate in the HAZ.
4. Post-weld heat treatment (PWHT): PWHT at 600-650°C for a duration of 1 hour per 25 mm of thickness is recommended to relieve residual stresses and to improve the toughness of the HAZ.
5. Non-destructive testing (NDT): 100% ultrasonic testing (UT) of the diaphragm-to-column weld is mandatory, with acceptance criteria per ISO 17637 or equivalent standards. Magnetic particle testing (MT) should be performed on the weld surface to detect surface-breaking defects.

Seismic Design Implications

The findings of this study have important implications for the seismic design of CFST column-steel beam connections:

• Ductility requirements: The seismic design of connections should target a ductility coefficient of at least 3.0 to ensure adequate energy dissipation capacity. The study shows that beam eccentricity can reduce the ductility coefficient to 85% of the concentric case, which may not meet this requirement.
• Welding quality: The seismic design of connections should include provisions for welding quality control, including 100% NDT, PWHT, and acceptance criteria that are more stringent than those for non-seismic connections.
• Bolted connection design: The seismic design of bolted connections should account for the cyclic loading that occurs during seismic events. The bolted connection should be designed to prevent bolt pull-out and bolt hole tearing, which are brittle failure modes that can lead to sudden connection failure.

Key Questions and Reflections

A significant question arising from this study is whether the fully bolted through-diaphragm connection can achieve the ductility required for seismic design. The study shows that the connection can achieve a ductility coefficient of approximately 3.0 for concentric specimens with larger beam cross-sections, but the ductility is reduced for eccentric specimens and for specimens with welded diaphragms. This suggests that the connection may not be suitable for all seismic applications, and that additional research is needed to develop connection details that can achieve higher ductility.

Another reflection is that the study highlights the importance of welding quality in seismic design. The finding that welding defects can reduce the energy dissipation capacity by up to 37% is a strong argument for stringent welding quality control in seismic construction. Engineers should advocate for the development of welding procedures that are specifically designed for seismic connections, with tighter acceptance criteria and more rigorous quality control.

Study Insights and Implications for Practice

This study provides valuable experimental data on the seismic performance of fully bolted through-diaphragm connections in square CFST columns. The identification of failure modes, bearing capacity, ductility, and energy dissipation characteristics gives engineers a comprehensive understanding of the connection behavior under seismic loading. The findings on the effect of beam eccentricity, beam cross-section size, and diaphragm welding provide guidance for connection design and detail selection.

For welding engineers, the key takeaway is that welding quality is not merely a compliance issue but directly affects the seismic performance of connections. The finding that welding defects can reduce energy dissipation capacity by up to 37% is a powerful argument for stringent welding quality control. Engineers should ensure that welding procedures are qualified, welders are certified, and NDT is performed to the highest standards.

The study also underscores the importance of full-scale testing for seismic connection design. While analytical models can provide useful insights, they cannot fully capture the complex behavior of connections under cyclic loading. Full-scale testing is essential for validating connection designs and for identifying failure modes that may not be predicted by analytical models.