The compressive strength of concrete is an important parameter in construction practice. At present, there are few reports on the prediction model of the compressive strength of concrete at a super early age. For some engineering vibration analyses, it is very necessary to study the growth law of compressive strength of concrete at a super early age. To this end, a new prediction model is proposed in this work to analyze the variation of compressive strength for the concrete at a super early age. The innovations of this work mainly lie in two aspects. The first innovation is to propose a new compressive strength-age mathematical model to predict the variation of compressive strength more accurately. The second innovation is to develop a new robust regression analysis method to obtain the fitting parameters in the mathematical model more effectively. Using the experimental data of the super early age concrete, the proposed prediction model is compared with the existing power function model and the hyperbolic function model. The results of the comparative study show that the prediction model proposed in this work is more reasonable and reliable. Taking C40 under natural curing as an example, it has been shown from the comparative study that: (1) The total fitting error of the proposed model is approximately 60% of that of the power function model, and approximately 17% of that of the hyperbolic model; (2) The fitting standard deviation of the proposed model is approximately 49% of that of the power function model, and approximately 15% of that of the hyperbolic model; (3) The 28 day strength of concrete predicted by the proposed model is more in line with the actual strength growth law of concrete.
Keywords: compressive strength; concrete; predictive model; regression analysis; super early age.