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

Numerical Simulation of Axial Compressive Behavior in Concrete-Filled Double-Skin Steel Tube Short Columns Using OpenSEES

Introduction and Literature Overview

The paper by Zhang Yuan, Zhang Long, Ren Hongwei, and Chen Jianwei, published in World Earthquake Engineering (Vol. 32, No. 1, 2016), addresses a structural form that has gained considerable traction in high-rise buildings and bridge piers: the concrete-filled double-skin steel tube (CFDST) column. The specific configuration studied is an outer square steel tube with an inner circular steel tube, filled with high-strength concrete in the annular space. This geometry offers advantages in both architectural aesthetics and structural efficiency, but its axial compressive behavior is complex due to the interaction between the outer steel tube, the concrete core, and the inner tube. The authors employ the OpenSEES finite element platform to simulate this behavior, using a fiber-based modeling approach with the Mander concrete constitutive model to account for the confinement effect provided by the steel tubes.

From my perspective as someone who has spent decades in steel pipe manufacturing and structural steel fabrication, this paper is a welcome contribution because it bridges the gap between theoretical structural engineering and the practical realities of steel tube production. The CFDST column is not merely an academic curiosity; it is a real construction solution used in seismic zones and high-rise construction, and its performance depends heavily on the quality and dimensional accuracy of the steel tubes supplied. The paper's focus on simulation is valuable, but the practical engineer must understand that the simulation is only as good as the material properties and geometric tolerances that go into the model.

Core Technical Points and Interpretation

The Mander Concrete Model and Confinement

The paper selects the Mander concrete constitutive model to represent the high-strength concrete. This is a critical choice. The Mander model was originally developed for confined concrete, and it accounts for the increase in compressive strength and ductility that occurs when concrete is laterally confined. In a CFDST column, the outer steel tube provides confinement to the concrete, but the degree of confinement depends on the tube's wall thickness, yield strength, and the shape of the cross-section. A square outer tube provides less uniform confinement than a circular tube because the flat sides tend to bulge outward under internal pressure from the concrete. The paper's choice of the Mander model implicitly acknowledges this non-uniform confinement, but the practical engineer should note that the model parameters must be calibrated against actual material test data. In my experience, the difference between a well-calibrated Mander model and one using default parameters can be 10-15% in predicted ultimate load, which is significant for design.

Fiber Model Discretization

The fiber model approach used in OpenSEES discretizes the cross-section into small fibers, each with its own material properties. This allows the model to capture the nonlinear behavior of both the steel and concrete, as well as the interaction between them. The paper uses an outer square tube and an inner circular tube, which means the concrete annulus has a complex geometry that must be carefully meshed. The authors report good agreement between their simulation and experimental results, which suggests that their meshing strategy and material parameter selection are sound. However, from a practical standpoint, the fiber model's accuracy is highly sensitive to the number of fibers used. Too few fibers and the model becomes overly stiff; too many and computational cost becomes prohibitive. The paper's approach appears to strike a reasonable balance, but the engineer applying this method should perform a mesh convergence study to ensure their own results are reliable.

Comparison with Experimental Results

The paper's key contribution is the validation of the OpenSEES modeling approach against experimental data. The authors report that the simulation results "agree well" with the test results, which is encouraging. However, I would caution that "agree well" in academic terms often means a 5-10% deviation in ultimate load, which is acceptable for research but may not be sufficient for design verification. In real projects, we typically require a safety factor of at least 1.5 on ultimate load, so a 10% modeling error is acceptable if the design uses conservative material properties. The practical engineer should also note that the experimental results used for validation were likely obtained from small-scale specimens, and the behavior of full-scale CFDST columns may differ due to size effects, welding residual stresses, and geometric imperfections that are difficult to replicate in small-scale tests.

Integration with Manufacturing and Delivery Practice

Steel Tube Dimensional Tolerances

The performance of a CFDST column depends critically on the dimensional accuracy of the steel tubes. The outer square tube must have straight sides and sharp corners, while the inner circular tube must be round and concentric with the outer tube. In my experience, square tubes are typically manufactured by cold-forming from steel plate, and the corner radius and wall thickness uniformity can vary significantly between manufacturers. The paper's model assumes perfect geometry, but in reality, the outer tube may have a corner radius that deviates from the nominal value, and the wall thickness may vary by as much as 10% from the specified value. These variations can affect the confinement of the concrete and the local buckling behavior of the tube. The practical engineer should specify tight tolerances on the tube dimensions and verify them upon delivery, because a poorly manufactured tube will not perform as the simulation predicts.

Welding and Residual Stresses

The steel tubes used in CFDST columns are typically welded from plate or formed from strip. The welding process introduces residual stresses in the tube wall, which can affect its buckling behavior under axial compression. The paper does not explicitly account for residual stresses, but the practical engineer must be aware of them. In my experience, the residual stresses from a submerged arc weld on a square tube can be significant, especially near the weld seam. These stresses can cause premature local buckling or reduce the effective yield strength of the tube. The paper's model assumes a stress-free tube, which may overestimate the ultimate load capacity. To address this, the engineer should either specify stress-relieving heat treatment for the tubes or incorporate a residual stress distribution into the model. The latter is difficult and requires detailed knowledge of the welding process, so stress-relieving is often the more practical option.

Concrete Placement and Quality Control

The performance of the CFDST column also depends on the quality of the concrete fill. The paper uses high-strength concrete, which typically has a low water-to-cement ratio and may require admixtures to achieve adequate workability. In practice, placing concrete into a narrow annular space between two steel tubes is challenging. The concrete must flow evenly around the inner tube without leaving voids, and it must be compacted thoroughly to avoid honeycombing. The paper does not address this aspect, but it is a critical practical concern. In my experience, the most common cause of failure in CFDST columns is not structural overloading but poor concrete placement, which leads to voids and reduced load transfer between the concrete and the steel tubes. The engineer should specify a concrete mix with high flowability and use a placement method that ensures complete filling, such as pumping from the bottom with a tremie pipe.

Key Questions and Reflections

One of the most important questions raised by this paper is whether the fiber model approach can reliably predict the behavior of full-scale CFDST columns. The paper validates the model against small-scale tests, but the transition to full scale introduces uncertainties related to size effects, concrete shrinkage, and creep. In my experience, the confinement effect in a large-diameter tube is less pronounced than in a small-diameter tube, because the concrete core is larger relative to the tube wall thickness. This means that the Mander model parameters may need to be adjusted for full-scale columns. The practical engineer should not blindly trust the simulation results but should verify them against full-scale test data or established design codes.

Another question is the applicability of the model to seismic loading. The paper focuses on axial compression, but CFDST columns are often used in seismic zones where they are subjected to combined axial and lateral loads. The fiber model approach can be extended to cyclic loading, but the Mander model may not accurately capture the degradation of concrete under cyclic loading. The practical engineer should consult the relevant seismic design codes and ensure that the simulation accounts for the full range of loading conditions.

Study Insights and Implications

This paper provides a valuable framework for simulating the axial compressive behavior of CFDST columns using OpenSEES. The key insight is that the fiber model approach, combined with the Mander concrete model, can accurately predict the ultimate load capacity and the load-deformation behavior of these columns. However, the practical engineer must be aware of the limitations of the model, particularly regarding residual stresses, geometric imperfections, and concrete placement quality. The paper's contribution is not just the simulation methodology but also the validation against experimental data, which gives confidence that the approach can be used for design purposes.

From a manufacturing perspective, this paper reinforces the importance of dimensional accuracy and welding quality in steel tubes used for CFDST columns. The engineer should specify tight tolerances on tube dimensions, ensure that the welding process is controlled to minimize residual stresses, and implement a rigorous quality control program to verify the tube properties upon delivery. The paper also highlights the need for careful concrete placement, which is often the weak link in CFDST construction.

Reference Value and Outlook

The paper is a useful reference for structural engineers who are designing CFDST columns and need a reliable simulation tool. The OpenSEES model described in the paper can be adapted to different cross-section geometries and material properties, making it a versatile tool for parametric studies. However, the practical engineer should not rely solely on simulation; they should always validate the model against experimental data and use conservative material properties in design.

Looking ahead, I see a need for more research on the behavior of CFDST columns under combined loading, particularly seismic loading. The paper's focus on axial compression is a good starting point, but the real challenge is to understand how these columns behave under cyclic loading and how the steel tube and concrete interact during an earthquake. The practical engineer should follow the literature closely and be prepared to adapt their design approach as new research emerges.

In conclusion, this paper is a solid contribution to the field of CFDST column analysis. It provides a validated simulation methodology that can be used for design, but it also highlights the need for careful attention to manufacturing quality and construction practice. The veteran engineer's advice is simple: use the simulation as a guide, but never forget that the real performance of a CFDST column depends on the quality of the steel tubes and the concrete placement, which are the responsibility of the fabricator and the contractor, not the finite element model.