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

Cyclic Load-Sharing Behavior of Steel-Concrete Composite Rock-Socketed Piles

Literature Overview and Study Context

This paper by Liu Mingwei et al. from Chongqing Jiaotong University addresses a problem I have seen repeatedly on inland waterway projects: how does a concrete-filled steel tube (CFST) rock-socketed pile actually share horizontal cyclic loads between the steel shell and the concrete core? The authors built three large-scale models at 1:7.3 scale and applied cyclic horizontal loads of 18.0, 22.5 and 27.0 kN, measuring strains on the outer steel tube, the corresponding inner concrete, and embedded reinforcing bars at multiple heights along the pile. This is not a theoretical exercise; it is a direct attempt to quantify what happens when a ship berthing force or wave action repeatedly loads a pile that is already working in a deep-water, shallow-overburden environment.

From my years on the delivery side, I can tell you that the design of these piles has historically been dominated by the steel tube, with the concrete treated as a passive filler. This paper challenges that assumption in a way that should make every foundation engineer sit up and take notice.

Core Technical Points and Test Observations

Four-Stage Strain Response Under Cyclic Loading

The authors identified four distinct stages in the strain development of both steel and concrete under cyclic loading. These stages are remarkably consistent across all test specimens and load amplitudes.

Stage Description Percentage of Fatigue Life
1 Linear strain growth 8.66%
2 Stable fluctuation 79.66%
3 Severe vibration 6.06%
4 Sharp decline 5.62%

The dominant stage is the stable fluctuation period, which accounts for nearly 80% of the fatigue life. This tells me that the composite system, once it settles into its working state, can tolerate a very long period of cyclic loading without significant degradation. The severe vibration and sharp decline stages are the warning signs that failure is imminent, and they occupy only about 12% of the total life combined. For engineers monitoring these piles in service, the practical implication is clear: if you see strain readings starting to oscillate beyond the stable band, you have very little time to intervene before the pile reaches its limit state.

Steel-Concrete Strain Mismatch at the Same Section

The most striking finding is the large difference between outer steel strain and inner concrete strain at the same pile cross-section. The maximum difference exceeds 80% of the concrete strain. This is not a minor discrepancy; it is a fundamental indication that the steel and concrete are not deforming together at the same section, particularly in regions of lower bending moment.

However, the embedded reinforcing bars tell a different story. The rebar and the surrounding concrete maintain strain coordination throughout the loading history, with the maximum difference not exceeding 20% of the concrete strain. This suggests that the rebar-concrete interface is behaving as a well-bonded composite, while the steel tube-concrete interface is not achieving the same level of strain compatibility.

Bending Moment Distribution Along the Pile

The moment distribution is perhaps the most practically significant finding. At the pile top, where the bending moment is relatively small, the inner concrete carries more than 70% of the section bending moment. As you move down the pile, the steel tube's share of the moment increases progressively. At the pile bottom, the steel and concrete each carry approximately half of the bending moment, and they reach a state of synchronized bending.

This has direct implications for design. If you assume the steel tube carries the entire bending moment, you are being conservative at the pile bottom but significantly over-conservative at the pile top. Conversely, if you assume the concrete carries the majority of the moment throughout, you are underestimating the steel's contribution at depth.

Effect of Load Amplitude on Strain Coordination

The authors also observed that at the same pile cross-section, a larger cyclic load amplitude causes the steel and concrete to reach their coordinated bending state earlier. This is counterintuitive at first glance, but it makes sense from a mechanics perspective: higher loads produce larger deformations, which force the composite action to engage more quickly, overcoming any initial slip or gap at the interface.

Engineering Practice Insights and Field Applications

What This Means for Pile Design

In my experience with deep-water port construction, the design of CFST rock-socketed piles has often relied on simplified assumptions about load sharing. This paper provides quantitative evidence that the actual behavior is more nuanced. For a pile with a given steel tube thickness and concrete core, the moment distribution varies significantly along the pile length, and the design engineer must account for this variation rather than using a single load-sharing factor.

The practical consequence is that the steel tube can be optimized for the lower portion of the pile where it carries a larger share of the moment, while the concrete core can be designed to handle the upper portion where it dominates. This could lead to material savings in the steel tube at the top of the pile and in the concrete at the bottom, provided the transition zone is properly detailed.

Monitoring and Inspection Implications

The four-stage strain response provides a valuable diagnostic tool for field monitoring. If strain gauges are installed on these piles, the transition from stable fluctuation to severe vibration is the key indicator that the pile is approaching its fatigue limit. In a real port environment, this would trigger a more detailed inspection, possibly including ultrasonic testing of the steel tube and core sampling of the concrete to assess any internal damage.

Interface Bond Quality

The large steel-concrete strain mismatch at the same section raises questions about the bond quality at the steel-concrete interface. In my experience with composite piles, the bond can be compromised by several factors: inadequate surface preparation of the steel tube interior, shrinkage of the concrete during curing, and the presence of voids or gaps at the interface. The authors' finding that the rebar maintains good strain coordination with the concrete suggests that the concrete itself is sound, but the steel-concrete interface is the weak link.

For manufacturing and fabrication, this means that the internal surface of the steel tube must be properly cleaned and prepared before concrete placement. Any rust, mill scale, or oil contamination will reduce the bond and lead to the kind of strain mismatch observed in this study. In my plant experience, we have found that sandblasting the interior of the steel tube to a near-white metal finish, followed by the application of a bonding agent, significantly improves the composite action.

The Role of the Rock Socket

The paper focuses on the steel-concrete interaction, but the rock socket itself is equally important. The pile is socketed into rock, and the load transfer from the pile to the rock mass is a critical part of the overall behavior. The authors' test setup appears to model this reasonably well, but in practice, the rock quality, socket length, and socket roughness all influence the pile's performance. I would want to see additional testing that varies the rock socket parameters to understand how they interact with the steel-concrete composite behavior.

Key Questions and Reflections

Is the 1:7.3 Scale Adequate?

The test models are at 1:7.3 scale, which is a reasonably large scale for laboratory testing. However, I have seen in practice that scale effects can be significant in composite structures. The bond between steel and concrete, the shrinkage behavior of concrete, and the strain distribution all depend on the absolute dimensions. A full-scale test or at least a larger-scale test would be valuable to confirm the findings.

What About the Fatigue Life?

The paper discusses the four stages of strain development but does not provide a direct fatigue life prediction. The cyclic load amplitudes of 18.0, 22.5, and 27.0 kN represent a range of loading conditions, but the actual fatigue life in terms of number of cycles to failure is not explicitly stated. For a port structure that may experience millions of load cycles over its service life, this is a critical missing piece of information.

How Does This Compare to Existing Design Codes?

Most design codes for composite piles use a simplified approach to load sharing, often based on the relative stiffness of the steel and concrete components. This paper suggests that the actual behavior is more complex, with the load sharing varying along the pile length and depending on the load amplitude. The design codes may need to be updated to reflect this more nuanced behavior.

Study Insights and Implications for Practice

This paper has given me a deeper appreciation for the complexity of CFST rock-socketed piles. The key takeaway for me is that the steel and concrete do not always work in perfect coordination, and the design engineer must account for the variation in load sharing along the pile length. The four-stage strain response is a valuable diagnostic tool for field monitoring, and the finding that the steel-concrete interface is the weak link in the composite system has direct implications for fabrication quality.

For the manufacturing side, the most important lesson is that the internal surface preparation of the steel tube is critical to achieving good composite action. This is a quality control issue that must be addressed during fabrication, not after installation. In my experience, the cost of proper surface preparation is minimal compared to the cost of a pile that fails to achieve its design capacity.

For the design side, the paper suggests that the steel tube can be optimized for the lower portion of the pile while the concrete core can be optimized for the upper portion. This is a departure from the traditional approach of designing both components for the maximum load, and it could lead to significant material savings.

For the field monitoring side, the four-stage strain response provides a clear warning system. If strain gauges are installed on these piles, the transition from stable fluctuation to severe vibration is the signal that the pile is approaching its fatigue limit. This would allow for timely intervention, such as reducing the berthing loads or strengthening the pile.

In conclusion, this paper is a valuable contribution to our understanding of CFST rock-socketed piles. It provides quantitative evidence of the complex load-sharing behavior and offers practical insights for design, fabrication, and monitoring. I would recommend that every engineer involved in the design or construction of these piles study this paper carefully and consider how its findings might apply to their specific projects. The four-stage strain response and the variation in load sharing along the pile length are particularly important findings that should inform future design practice.