Application of PLS Technique to Optimization of the Formulation of a Geo-Eco-Material

Abstract

Earthen construction is one of the most familiar construction method practiced since the ancient times. It has been successfully used around the world. The raw earth material as a geo-eco-material is one of the most abundant, basic building materials. It is low technology, straightforwardly worked with simple tools. It requires very low energy to manufacture through low environmental impact, especially when the material is sourced on construction site. Nowadays, building construction with raw earth materials needs remarkable mechanical performance. For this, a raw earth treatment with binders is one of the techniques used to make better its strength and durability. This paper presents the use of Design of Experiments through D-optimal mixture design as a tool to optimize a raw earth concrete formulation to achieve a desirable compressive strength. A multivariate statistical regression technique of PLS, Partial Least Square projections to latent structures, was chosen to evaluate the design. This PLS technique was selected because of the complicated experimental design data along with different constraints on model. The obtained results illustrate that PLS technique can be a useful tool to improve and optimize a raw earth concrete formulation.