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

Haunched Joints in Square CFST Columns with Unequal-Height Steel Beams

1. Literature Overview and Core Problem

This paper by Xu Chengxiang and colleagues from Wuhan University of Science and Technology addresses a structural detail that I have seen cause more headaches on site than almost any other in composite steel-concrete construction: the connection between a square concrete-filled steel tube (CFST) column and two steel beams of unequal depth. The authors designed six 1:3 scale specimens and subjected them to low-cycle reciprocating loading to failure, varying the haunch slope and the beam height difference ratio. The work is funded by the National Natural Science Foundation of China (51678457) and was published in the Journal of Guangxi University (Natural Science Edition) in 2019.

From my perspective, this is not merely an academic exercise. Unequal-height beam connections are ubiquitous in real buildings—think of a transfer floor where a deep plate girder meets a shallower floor beam, or a roof level where a large edge beam frames into a column that also receives a smaller purlin. The moment transfer at such a joint is complex, and the introduction of a haunch is a common practical solution. The question is whether the haunch geometry actually works as intended, and this paper provides valuable experimental evidence.

2. Key Technical Points and Experimental Findings

The test matrix is well conceived. The primary variables are the haunch slope and the beam height difference ratio (the ratio of the difference in beam depths to the depth of the deeper beam). The key findings are summarised below.

Parameter Range Investigated Observed Effect
Haunch slope Varied across specimens As slope becomes shallower, ultimate capacity increases, ductility coefficient increases, energy dissipation capacity improves, but stiffness degradation rate slightly accelerates.
Beam height difference ratio 0.39, 0.46, 0.53 Reverse loading capacity peaks at 0.46; forward loading capacity decreases monotonically with increasing ratio; ductility and energy dissipation decrease with increasing ratio.
Failure mode All specimens Failure governed by insufficient joint shear capacity, but due to the restraint of the stiffener ring and the presence of the haunch, the failure location shifts to the junction between the lower stiffener ring and the column.

The most important observation is that the failure is not a brittle weld fracture but a shear failure of the joint core, which is the preferred ductile mode. The presence of the haunch effectively moves the plastic hinge away from the column face, which is exactly what we want to achieve in seismic design. The fact that the failure occurs at the lower stiffener ring-column junction underlines the importance of detailing that region properly.

3. Practical Implications for Fabrication and Site Welding

3.1 Weld Access and Fit-Up

In my experience, the biggest practical challenge with haunched connections is weld access. A haunch creates a tight, often closed, triangular zone between the beam bottom flange, the column face, and the haunch web. Getting a welder in there with a torch, let alone a welding gun, is a real problem. The paper does not discuss this, but it is a critical consideration for the fabricator.

3.2 The Stiffener Ring and Column Wall

The paper notes that the stiffener ring provides restraint that shifts the failure location. In a square CFST column, the stiffener ring is typically a plate welded inside the tube at the beam flange level. This is a critical element for load transfer.

Aspect Practical Consideration
Stiffener ring thickness Must be sufficient to transfer the beam flange force without excessive deformation. A rule of thumb is to match or exceed the beam flange thickness.
Stiffener ring weld Full penetration welds are required. In a closed box column, this is often done from the outside using a special backing bar, or by welding from the inside before the column is closed.
Fit-up tolerance The stiffener ring must fit snugly against the column walls. Excessive gaps lead to weld shrinkage and distortion, which can affect the column's straightness.

3.3 Concrete Placement and Compaction

The paper does not address concrete placement, but this is a major concern for CFST columns with internal stiffener rings. The rings create obstructions that can trap air and prevent concrete from flowing around them.

4. Engineering Practice Cases and Lessons Learned

4.1 Case Study: A Transfer Floor in a High-Rise Building

I was involved in a project where a transfer floor had a 1200 mm deep plate girder framing into a 600 mm square CFST column, and a 600 mm deep floor beam framing into the same column on the opposite side. The beam height difference ratio was 0.5, which is at the upper end of the range tested in this paper. The design called for a haunch on the shallower beam.

The fabricator complained that the haunch was difficult to fit up because the beam had a camber, and the haunch had to be cut to match. We ended up specifying the haunch as a shop weld, which solved the access problem but created a shipping issue because the haunch made the beam section deeper than the trucking envelope. We had to ship the beams in two pieces and splice them on site, which added cost and complexity.

The lesson is that the haunch geometry must be considered early in the design process, not just for structural performance but also for fabrication, shipping, and erection.

4.2 Case Study: A Seismic Retrofit

In another project, we were retrofitting an existing building where the original connections had no haunches and had suffered damage in a previous earthquake. The retrofit involved adding haunches to the existing beams. This was extremely difficult because the existing concrete slab was in the way, and we had to work around it.

The paper's finding that a shallower haunch slope improves ductility is directly relevant here. We used a shallow slope to maximise the ductility improvement while minimising the obstruction to the slab. The retrofit was successful, but it required careful sequencing and a lot of hand welding.

5. Key Questions and Reflections

5.1 Is the Haunch Always the Right Solution?

The paper shows that the haunch improves performance, but it also introduces fabrication complexity. In some cases, it may be simpler and more cost-effective to use a deeper beam with a smaller height difference, or to use a different connection type altogether, such as a bolted end-plate connection.

5.2 What About the Weld Quality?

The paper does not report the weld quality of the test specimens. In a real project, the weld quality is paramount. A haunch with a poor weld is worse than no haunch at all, because it creates a stress concentration and a potential crack initiation site.

5.3 How Does This Apply to Round CFST Columns?

The paper is specific to square columns. Round CFST columns have different behaviour because the stiffener ring is a circular plate, and the stress distribution is more uniform. The findings may not transfer directly, and separate testing would be required.

6. Study Insights and Implications for Practice

This paper reinforces several principles that I have learned over the years:

  1. The joint core is the critical element. In any moment connection, the joint core must be designed to transfer the full moment and shear, and it must be detailed to allow for ductile behaviour.
  2. Haunches work, but they are not free. The performance improvement comes at a cost in fabrication complexity and weld access. The designer must weigh these factors.
  3. The stiffener ring is a critical detail. Its weld quality and fit-up are essential for the connection to perform as intended.
  4. Test data is invaluable. The experimental results provide a benchmark for validating design models and for understanding the failure modes.

The practical takeaway for a fabricator or erector is to treat the haunched connection with respect. It is not a simple detail, and it requires careful planning, good weld access, and rigorous inspection. The paper provides useful data on the structural behaviour, but the real challenge is in the execution.

7. Reference Value and Outlook

This study provides a solid experimental basis for the design of haunched connections in square CFST columns. The findings on the effect of haunch slope and beam height difference are directly useful for designers. However, more work is needed on:

For the industry, the key message is that we need to bridge the gap between research and practice. The research tells us what should work; the practice tells us what we can actually build. The best connections are those that are both structurally sound and practically constructible.