Recycled Self-Compacting Concrete Filled Steel Tube Axially Loaded Short Columns
Overview and What This Paper Really Says
Zhou Jinghai, Su Ming, Meng Xianhong and Yang Yongsheng from Shenyang Jianzhu University have done something that the industry has been talking about for years but rarely tests: they have taken recycled aggregate concrete, made it self-compacting, and put it inside a steel tube to see if the composite column can carry load. The motivation is clear — recycled concrete is cheaper and more sustainable, but it has lower strength and higher shrinkage than natural aggregate concrete. The steel tube provides confinement and prevents the concrete from spalling, which may compensate for the lower material properties. The authors used an orthogonal test to optimise the mix design, then fabricated short columns and loaded them axially to failure. The key finding is that the optimum mix is 50% recycled aggregate replacement, 10% superplasticiser, and 8% fly ash. The ultimate load of the recycled self-compacting concrete filled steel tube (RSCFST) is lower than that of a conventional concrete filled steel tube (CFST), but the displacement at maximum load is essentially the same — about 7 mm. The authors conclude that recycled concrete is entirely feasible for use in CFST structures.
Technical Interpretation and Key Parameters
The orthogonal test is the right tool for this problem. With three factors and three levels, a full factorial would require 27 mixes; an orthogonal array reduces that to 9 or 12, which is practical for a laboratory. The factors and levels are summarised below.
| Factor | Level 1 | Level 2 | Level 3 | Optimum |
|---|---|---|---|---|
| Recycled aggregate replacement (%) | 30 | 50 | 70 | 50 |
| Superplasticiser dosage (%) | 6 | 8 | 10 | 10 |
| Fly ash content (%) | 5 | 8 | 12 | 8 |
The optimum mix gives a workable self-compacting concrete with a 50% recycled aggregate replacement. The authors do not report the compressive strength of the concrete itself, which is a gap — the strength of the core concrete is the single most important parameter in a CFST. However, the fact that the ultimate load of the RSCFST is lower than that of the conventional CFST suggests that the recycled concrete has a lower cube strength, even with the same mix proportions.
The key mechanical result is that the ultimate load of the RSCFST is lower, but the displacement at maximum load is about 7 mm for both types. This is significant because it means the stiffness and the ductility of the composite column are not greatly affected by the recycled aggregate. The steel tube is doing most of the work in the elastic range, and the concrete only contributes significantly after the steel yields. The confinement provided by the steel tube is what allows the concrete to reach its full strength without spalling, and this mechanism is the same whether the aggregate is natural or recycled.
Real Practice, Real Failures, and What I Would Tell a Site Engineer
I have shipped thousands of tonnes of steel tube for CFST columns, and I have seen the problems that arise when the concrete does not fill the tube properly. Self-compacting concrete is the answer to that problem — it flows into the tube without vibration, which is essential for tall columns where vibration is impossible. The authors are right to focus on self-compacting behaviour. But recycled aggregate concrete is notoriously variable. The quality of the recycled aggregate depends entirely on the source — if it comes from a demolition site with contaminated concrete, the chloride content and the sulphate content can be high, leading to corrosion of the steel tube from the inside. I would insist on a full chemical analysis of the recycled aggregate before it is used in a CFST.
The other practical issue is shrinkage. Recycled concrete has higher shrinkage than natural concrete, and in a CFST the shrinkage creates a gap between the concrete and the steel tube. This gap reduces the bond and, more importantly, reduces the confinement effect. The authors do not measure shrinkage, but the 7 mm displacement at maximum load suggests that the bond is not severely compromised. Still, on site I would specify a shrinkage-compensating admixture or a post-filling grouting method to ensure intimate contact between the concrete and the steel tube.
The orthogonal test gives a mix design, but it does not give a full stress-strain curve. For design purposes, the engineer needs the complete constitutive relationship of the core concrete, not just the optimum mix. I would ask the laboratory to produce the full stress-strain curve for the optimum mix and compare it with the natural aggregate concrete. The design formulas in the existing codes assume a certain concrete behaviour; if the recycled concrete behaves differently, the design formulas may not be valid.
Key Questions and Reflections
The most important question is whether the 50% replacement level is economically and practically viable. Recycled aggregate is cheaper than natural aggregate, but the additional superplasticiser and fly ash add cost. The net saving may be small. The real benefit is environmental — using recycled aggregate reduces landfill waste and reduces the demand for natural aggregate. In a country like China, where construction waste is a major problem, this is a significant advantage. But the engineer must weigh the environmental benefit against the lower ultimate load. If the column is designed for a fixed load, a larger column will be needed, which may negate the cost saving.
Another question is the long-term behaviour. The authors only tested short-term axial loading. In practice, a CFST column is subjected to sustained loads, creep and shrinkage of the concrete, and temperature variations. Recycled concrete has higher creep than natural concrete, which means the load will transfer from the concrete to the steel tube over time. This is not necessarily bad — the steel tube can carry the additional load — but it changes the stress distribution and may affect the serviceability limit state. I would like to see a long-term creep test before recommending recycled concrete for permanent structures.
Study Insights and Practical Implications
This paper is a useful first step, but it is not the last word. The authors have demonstrated that recycled self-compacting concrete can be used in a CFST column without catastrophic loss of capacity or ductility. That is a valuable finding. But the practical implications are limited until the long-term behaviour and the full stress-strain relationship are established. For a veteran like me, the lesson is that material innovation must be tested under the conditions that will occur in service, not just under short-term laboratory loading. The steel tube is forgiving — it will carry the load even if the concrete is weak — but it cannot compensate for a concrete that shrinks away from the tube or creeps excessively.
I would recommend that the next step be a full-scale column test with the optimum mix, instrumented with strain gauges on the steel tube and the concrete, and loaded for at least 90 days to measure creep and shrinkage. The results should be compared with the existing design formulas for CFST columns. If the formulas are conservative for recycled concrete, then the material can be used with confidence. If not, a new design formula is needed. The paper has opened the door; the industry must now walk through it carefully.
Zhuojin Pipe Fitting Co., Ltd