Entrained air bubbles in concrete improve the freeze–thaw resistance by alleviating stresses caused by freezing water within the microstructure. It is necessary to achieve a uniform distribution of entrained air bubbles in the paste to achieve optimal freeze–thaw performance. The technique of measuring the air-void system parameters in hardened concrete is specified by ASTM C457/C457M-12. In this standard testing procedure, the freeze–thaw parameters of interest are determined by optically observing and measuring the traversal lengths of air voids, paste, and aggregate particles along straight lines across the polished surface of the concrete specimen. ASTM C457/C457M-12 has been criticized as time-consuming and expensive, and so there has been a concerted effort by the research community to develop improved methods. Automated methods that use digital images to measure freeze–thaw parameters exist, but they require operator-selected threshold levels and significant sample preparation time and labor. To address these shortcomings, this paper presents an automated method to measure freeze–thaw parameters using a high-resolution flatbed scanner, automated image segmentation, and short sample preparation time and labor. Polished concrete surfaces are treated with phenolphthalein and a florescent chalk powder to distinguish phases of interest. The surfaces are scanned using a flatbed scanner, and spectral–spatial analysis is used to identify the individual phases of aggregates, paste, and air voids. The ternary images are used to measure parameters of the air-void system. To validate the rigor of this automated method, six groups of concrete specimens provided by a commercial laboratory were tested using the proposed automated method and manual ASTM C457/C457M-12. This comparison was conducted in a “blind” fashion, so as to avoid potential biases. A fair agreement is found for normal concrete specimens, but not for concrete containing lightweight aggregates.

No abstract

No abstract

Concrete creep results in the strain increment of concrete-filled steel tube. In this article, a new method was established for creep prediction of concrete-filled steel tube considering the creep-recovery effect of the concrete core with creep model and separate creep-recovery model, named two-function method. Five creep models including the recent proposed B4 model and two separate creep-recovery models were used in the method. First, the creep models were compared with 197 test curves of creep of sealed concrete. The results show that the B4 model offers the highest prediction accuracy for load duration time longer than 200 days. Then, the two-function method was used for creep prediction of normal strength concrete-filled steel tube and high-strength concrete-filled steel tube. For comparison, the creep prediction method for concrete-filled steel tube based on the principle of superposition was also adopted. The principle of superposition did not predict the creep of concrete-filled steel tube accurately. The calculation results based on the two-function method are closer to the test data, especially for normal strength concrete-filled steel tube. Finally, the uncertainty of creep in concrete-filled steel tube was analyzed considering the variability of the concrete strength. The results indicate that the variances calculated by the methods with the MC 2010 creep model are smaller than that of the B4 model. The creep strains of concrete-filled steel tube calculated by the two-function method are subject to normal distribution.