Multi-parameter Linear Regression Analysis About the Experimental Results of Pervious Concrete Strength

摘要

In this paper, based on 28-day cubic compressive strength f_(cu) and splitting strength f_(ts) of pervious concrete with different porosity, aggregate size and of cubic size, correlation analysis was conducted among these three factors and strength using the software of SPSS (Statistical Package for the Social Sciences), in which a logarithmic function was used to do linear regression fitting about the relationship among the three parameters and the strength and then a formula was proposed. The analysis result shows that there is a certain higher-order function relationship between the three influencing factors and the strength, and it is contribute to analysis the impact of three parameters on strength by using natural logarithm to linearize the relationship of parameters. Studies have shown that porosity is the main factor that affects both f_(cu) and f_(ts), and side length of the cube has a significant influence on f_(ts) but has no significant linear correlation with f_(cu). And f_(ts) may be more sensitive to cube size on linear correlation than f_(cu).