Contributions to experiment design in civil engineering research projects

Abstract

The use of statistically designed experiments in research projects that involve the analysis of results from experimental programmes is not common in Civil Engineering. The statistical design of experiments allows for the refined analysis of data, obtained from the modelling of response variables. Moreover, fractional designs can be used to work with several control variables, at different levels, without making the study expensive and reducing the necessary time for the realization of experiments. This article discusses the advantages of using experiment designs, in order to make a contribution for improving the quality of research methods that involve experimental programmes. Some examples of experiment design in the field of construction materials technology are described, indicating the main benefits achieved. Also, data analysis obtained from complete and fractional experiments are presented, indicating that the same information obtained from complete experiments can be obtained from fractional experiments following statistical criteria.