Diseño de experimentos factorial completo aplicado al proyecto de muros de contención

Abstract

This paper applies a complete factorial design to a five-meter wall to evaluate which variables most influence the response. This method is used for two target functions, CO2 emissions and the cost of the structure. To do this, 32 evaluations of the structure are performed using a computer program and a statistical analysis is carried out. The results of this analysis show that the most statistically representative factor is the thickness of the wall and the length of the toe is of little importance for both target functions. The result of the models considering only the variables without the interaction results in an R2 greater than 95%, so the interaction between variables, although it is proven to exist, is not relevant to the case study. This methodology allows to reduce the complexity of structural problems, reducing the number of variables.