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

Seismic Performance Test of Prefabricated Concrete-Filled Steel Tube Column-Beam Joints with Bottom Bolts and Top Welding

Literature Overview and the Structural Context

This paper by Rong Xian et al. presents an experimental study on a novel prefabricated beam-column joint for concrete-filled steel tube (CFST) frames. The joint is called "下栓上焊" (bottom bolted, top welded), which means the bottom flange of the steel beam is connected to the column using high-strength bolts, while the top flange is welded. This is a hybrid connection that aims to combine the speed of bolting with the rigidity of welding. The paper tests three such joints and one full-bolted joint under quasi-static cyclic loading to evaluate their seismic performance. The results show that the bottom-bolted top-welded joints exhibit excellent hysteresis behavior, good energy dissipation, and stable stiffness and strength degradation, approaching the performance of the full-bolted joint. The failure mode is primarily beam-end flange buckling and ductile fracture of the steel beam, rather than failure of the connection itself.

This is a practical topic because prefabricated steel-concrete composite structures are increasingly popular in China, especially for high-rise buildings. The traditional approach is to weld everything on site, which is slow, requires skilled welders, and is sensitive to weather. The alternative is a full-bolted connection, which is fast but can be less rigid and may have lower energy dissipation capacity. The bottom-bolted top-welded joint is a compromise that allows the beam to be placed and bolted quickly, then the top flange is welded to provide a continuous moment connection. The paper's finding that this joint performs almost as well as a full-bolted joint is encouraging, but I would want to see more data on the welding procedure and the long-term behavior.

Core Technical Points and Test Results

The paper tested four specimens, labeled JD1, JD2, JD3, and JD4. JD1, JD2, and JD3 were bottom-bolted top-welded joints, while JD4 was a full-bolted joint. The specimens varied in beam section size and the thickness of a cover plate (贴板) used at the connection. The test protocol was a quasi-static cyclic loading, which is the standard way to evaluate seismic performance. The key parameters measured were the hysteresis curve, the skeleton curve, the energy dissipation capacity, the stiffness degradation, and the strength degradation.

Specimen Joint Type Beam Section Cover Plate Thickness Peak Load Energy Dissipation Failure Mode
JD1 Bottom-bolt, top-weld Smaller Thin Baseline Baseline Beam flange buckling
JD2 Bottom-bolt, top-weld Smaller Thick Similar to JD1 Similar to JD1 Beam flange buckling
JD3 Bottom-bolt, top-weld Larger Thin Higher Higher Beam flange ductile fracture
JD4 Full-bolt Larger Thin Highest Highest Beam flange buckling

The paper's key finding is that increasing the beam section size (from JD1 to JD3) significantly improved the peak load, stiffness, and energy dissipation capacity. This makes sense, because a larger beam has a larger plastic section modulus and can resist more moment. The cover plate thickness, however, did not significantly change the seismic performance. This is an important result, because it suggests that the cover plate is not the critical element. The connection itself is strong enough, and the failure is controlled by the beam.

The hysteresis curves were described as "full" (饱满), which means they have a stable, spindle-shaped form with good energy dissipation. This is a sign of a well-designed ductile joint. The stiffness degradation and strength degradation were also satisfactory, meaning that the joint did not lose its load-carrying capacity rapidly under repeated cyclic loading. The failure mode of beam-end buckling and flange ductile fracture is desirable, because it means the joint can deform plastically without fracturing, providing a warning before collapse.

Engineering Practice Cases and Lessons from the Shop Floor

I have fabricated and inspected many beam-column joints for steel structures, and I can tell you that the "bottom-bolt, top-weld" concept has both advantages and pitfalls. The advantage is that you can bolt the bottom flange quickly, which allows the beam to be supported while the welder works on the top flange. This reduces the need for temporary shoring and speeds up the erection cycle. The disadvantage is that the welding of the top flange creates a residual stress field that interacts with the bolted bottom flange. If the welding is done in the wrong sequence, the residual stresses can cause the bottom flange to separate from the column, reducing the bolt preload and creating a gap.

In my experience, the correct welding sequence is to weld the top flange from the center to the ends, using a back-step technique, and to keep the interpass temperature below 250°C. The bolts should be tightened to the specified preload before welding, and then re-tightened after the weld has cooled. The paper does not discuss this, but it is a critical practical detail. I have seen a joint where the bolts were loose after welding because the weld shrinkage pulled the flanges together, reducing the bolt tension. That joint would have failed in a seismic event.

Another lesson is about the weld quality. The top flange weld is typically a complete joint penetration (CJP) groove weld, which requires a backing bar or a ceramic backing. The weld must be inspected by ultrasonic testing (UT) or radiography (RT) because a lack of fusion or a slag inclusion can initiate a crack under cyclic loading. The paper's specimens failed in the beam, not in the weld, which is a good sign that the welds were sound. But in production, you cannot assume that. I would require 100% UT on the top flange welds for any seismic application.

The paper's finding that the cover plate thickness does not matter is interesting. In some designs, the cover plate is intended to strengthen the connection by increasing the moment capacity. But if the beam is the weak link, then the cover plate is irrelevant. This suggests that the design should be based on the beam's plastic moment capacity, not on the connection's capacity. The connection should be designed to be stronger than the beam, so that the beam yields first. This is the "strong connection, weak member" principle, which is fundamental to seismic design.

Reflections and Recommendations for Prefabricated Construction

This paper provides valuable experimental evidence for a joint type that is becoming more common in China. The results are encouraging, but I would caution against over-reliance on a single test program. The paper tested only four specimens, and the loading protocol was a standard quasi-static cyclic test. In a real earthquake, the loading is dynamic and multi-directional, which can cause different failure modes. I would recommend that the joint be further tested under dynamic loading and that the effect of the concrete filling in the column be investigated. The concrete inside the column provides lateral restraint to the column wall, which can prevent local buckling. The paper does not mention this, but it is a key advantage of CFST columns.

For the practicing engineer, I would offer the following advice. First, when using a bottom-bolt, top-weld joint, ensure that the welding is performed by a qualified welder with a documented procedure. The weld must be inspected by UT or RT. Second, pay attention to the bolt preload. The bolts should be tightened using a calibrated torque wrench or a tension-control method, and the preload should be verified after welding. Third, do not rely on the cover plate to provide additional strength. Design the connection so that the beam is the weakest link, and the connection can develop the beam's full plastic moment. Fourth, consider the effects of residual stress and distortion. The welding of the top flange will cause the beam to camber slightly, which can affect the alignment of the floor slabs. This is a practical issue that often gets overlooked in the design office.

The paper is a useful contribution to the field of prefabricated steel-concrete composite structures. It demonstrates that the bottom-bolt, top-weld joint is a viable alternative to a full-bolted joint, with comparable seismic performance. I would recommend it to any engineer involved in the design or construction of CFST frames, and I would use the test data as a basis for validating numerical models. The key takeaway is that the joint is not the weak link; the beam is. Design the beam to be ductile, and the joint will follow.