10 - Statistical analysis techniques in ACT

Abstract

Effective management of concrete includes a quantitative knowledge of the key attributes, monitoring techniques, decision methods, their limitations and an ability to interpret the measured values. Statistical techniques are used extensively to understand and compare variations between concrete batches, to modify and control the production of concrete and to form the basis of Quality Control and Quality Assurance. This chapter focuses on the statistical techniques that are exploited and applied explicitly to current practical circumstances in ACT, such as trends and errors, estimation of parameters, checking test results, mix design, compliance, and quality control. For analysis two statistical distributions, the t-distribution, and the F-distribution, are introduced and their use to provide estimates of key population parameters are explained. Objectives are to: understand the concept of sampling to provide an estimate of a key (population) values and calculate of the precision of estimates for large sample sizes using the normal distribution. The concept of decision making based on sample data is covered and applied to the comparison of mean and variances of key parameters. Regression models are also presented for an examination of possible relationships between linked parameters and the derivation of useable functional relationships.