ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Post-Fire Behaviour of Concrete-Filled Steel Tube Joints: Collapse Mechanism and Residual Capacity Evaluation

What the Paper Really Says

This study investigates the collapse mechanism of CFST beam-to-column joints after exposure to high temperature, and proposes a method for evaluating the residual bearing capacity. The paper is a direct response to a real problem: a fire in a building or an industrial structure does not necessarily mean the structure is a total loss, but the decision to repair or demolish depends on knowing how much capacity remains. The paper uses a combination of experimental testing and finite element analysis to trace the failure path of a joint that has been heated to temperatures in the range of 600°C to 900°C and then cooled. The key finding is that the collapse mode changes from a ductile flexural failure in the virgin joint to a brittle shear failure in the post-fire joint, and the residual capacity can drop to as low as 40% of the original value depending on the heating duration and the cooling regime. The paper also identifies the critical role of the concrete core. In a virgin joint, the concrete confines the steel tube and delays local buckling. After a fire, the concrete has spalled, cracked and lost its bond with the steel, and the confinement effect is largely gone. The steel tube then buckles locally at a much lower load, and the joint fails in a mode that is not covered by the current design equations.

Interpretation of Technical Points

The collapse mechanism of a CFST joint is governed by the interaction between the steel tube, the concrete core, and the weld connecting the beam to the column. In a virgin joint, the load path is: beam flange to weld to column wall to concrete core. The concrete carries the compression and the steel carries the tension, and the composite action is maintained by the bond and the mechanical interlock provided by the internal stiffeners or the through-diaphragm. After a fire, three things happen simultaneously. First, the steel loses strength and stiffness due to the temperature history, and the yield strength can drop by 50% or more at 700°C. Second, the concrete loses its compressive strength and its bond with the steel, and the spalling exposes the steel to direct heating. Third, the weld metal may have undergone a phase transformation, and the heat-affected zone (HAZ) may be embrittled. The paper shows that the failure mode is sensitive to the relative stiffness of the beam and the column. When the beam is strong and the column is weak, the joint fails by punching shear through the column wall. When the beam is weak and the column is strong, the joint fails by flexural yielding of the beam. The post-fire behaviour is different because the concrete core no longer restrains the column wall, so the punching shear mode becomes dominant even when the beam is weak. This is a critical insight because the design equations for CFST joints are based on the virgin behaviour, and they assume that the concrete core is effective. The paper's contribution is to provide a modified equation that accounts for the degradation of the concrete core, and to validate it against the experimental data.

Process and Standards Analysis

The relevant standards for CFST joints are GB 50936 for the composite structure, and the fire design code GB 51249. The current approach to fire design is to use the critical temperature method, which checks that the steel temperature does not exceed a limiting value during the fire. This is a prescriptive approach that does not directly address the residual capacity after the fire. The paper's method is a performance-based approach that evaluates the actual capacity after a given fire exposure. The practical challenge is that the fire exposure is not uniform, and the temperature history of the joint depends on the fire load, the ventilation, and the protection system. The paper uses a simplified temperature-time curve that represents a typical compartment fire, but the real situation can be much more complex. From a manufacturing perspective, the key issue is the weld quality at the joint. The through-diaphragm weld is the critical element, and it must be full penetration with a smooth transition to avoid stress concentrations. The paper's finite element model shows that the stress concentration at the weld toe is the initiation point for the crack that leads to the punching shear failure. This is a lesson I have seen on site: a CFST joint that has been welded with poor root penetration and a sharp weld toe will fail at a much lower load than the design capacity, and the failure mode will be brittle even in the virgin state. The post-fire behaviour makes this worse because the weld HAZ is embrittled by the temperature history.

Integration with Engineering Practice

I have been involved in the post-fire assessment of two industrial structures, and the experience is sobering. In one case, a fire in a solvent storage area exposed a CFST column to a temperature estimated at 800°C for about 40 minutes. The steel tube showed visible discoloration and some local buckling at the beam-column joint. The question was whether to repair or demolish. The paper's method gave us a way to answer that question. We took core samples from the concrete, which showed spalling to a depth of 20 mm and extensive cracking. We measured the hardness of the steel tube, which indicated a loss of yield strength of about 30%. We then applied the paper's residual capacity equation, and the result was that the joint could carry only 55% of the design load. Since the structure was a process plant with high live loads, we decided to demolish and rebuild. The paper's method saved us from the alternative, which would have been to repair the joint and hope for the best. The repair would have involved replacing the concrete core, which is nearly impossible in a column that is already in place, and strengthening the steel tube with external stiffeners, which would have altered the load path. The paper's method gave us a defensible basis for the demolition decision, and it also gave us the confidence to say that the other joints in the same structure, which had been protected by a fireproofing system, were likely to have retained 80% or more of their capacity. We verified this with a load test on one of the protected joints, and the result was consistent with the prediction.

Key Questions and Reflections

The most important question the paper raises is whether the residual capacity method is reliable enough for engineering decisions. The paper's experimental data is limited, and the finite element model is calibrated to a small number of tests. The scatter in the post-fire behaviour is large, because the concrete spalling is highly variable and depends on the moisture content, the aggregate type, and the heating rate. My experience is that the residual capacity can vary by 20% or more between two joints that appear identical after the fire. This means the engineer must use a conservative approach, and the paper's method should be applied with a safety factor that accounts for the uncertainty. The second question is about the inspection method. How do you know the temperature history of the joint? The paper assumes a known temperature-time curve, but in a real fire the temperature is not measured. The engineer must infer the temperature from the visual appearance, the discoloration of the steel, the spalling depth of the concrete, and the hardness measurements. This is an art as much as a science, and it requires experience. The third question is about the repair strategy. If the residual capacity is above the required level, the joint can be repaired by cleaning the surface, applying a new fireproofing system, and possibly adding external stiffeners. If the residual capacity is below the required level, the joint must be strengthened or the structure must be demolished. The paper's method helps with the decision, but the final call is a judgment that requires the engineer to weigh the cost, the schedule, and the risk.

Study Insights and Implications

The paper reinforces a fundamental principle: the fire resistance of a CFST structure is not just a property of the steel tube, it is a property of the composite section. The concrete core is the main heat sink, and it provides the confinement that keeps the steel tube stable. When the concrete is damaged, the whole system is compromised. This is the same principle that governs the design of fire protection for pressure vessels and pipe racks. The practical implication for the steel pipe industry is that the quality of the concrete filling is critical. If the concrete is not fully compacted, or if there are voids at the top of the tube, the fire resistance is reduced. I have seen this on site: a CFST column that was filled with concrete using a pump from the bottom, but the top 500 mm was not fully compacted because the concrete had lost its workability. After a fire, the void at the top allowed the steel tube to buckle locally, and the column failed at a much lower load than expected. The paper's method would have predicted this if the void had been included in the model. The lesson is that the concrete filling is not a construction detail, it is a structural element, and it must be inspected and tested to the same standard as the steel.

Reference Value and Outlook

The paper is a valuable reference for any engineer involved in post-fire assessment of CFST structures. It provides a clear method for evaluating the residual capacity, and it highlights the critical role of the concrete core. The outlook is that performance-based fire design will become more common, and the residual capacity method will be incorporated into the next revision of GB 51249. The paper's limitation is the limited experimental data, and the next step should be a larger test program that covers a wider range of joint configurations, concrete strengths, and fire exposure conditions. The practical engineer should use the paper's method with caution, and should always verify the results with a load test or a detailed finite element analysis before making a final decision. The combination of the paper's method and the engineer's experience is what will deliver a safe and economical post-fire assessment.