Metodología de optimización para soluciones estructurales en edificación residencial

Laburpena

The main objective of this Thesis is the development of a methodology for decisionmaking on the optimal structural typology for building projects, combining performance criteria, cost, and environmental impact, useful for the different agents involved in the development of residential building projects. This methodology is based on the knowledge generated in the research, through a deep analysis of previous information obtained from real developed projects, and the experience in the sector. The objective was not only to generate the methodology and its underlying knowledge, but also to be useful and practical, so that the research approaches as much as possible to the current reality, in a quantitative way. The research begins with a preliminary analysis of residential building projects, with over 200,000 m2 of constructed area, from which the main results are the significant impact of the so-called "Pillars Index" (number of pilares/m2 of built area) as well as the typologies, geometries, and materials commonly used. In order to model an infinite universe with a finite number of cases, a study of 452 alternatives is settled, thereby obtaining a complete and representative database with exhaustive information of each case study. Implementation cost, rigidity, and environmental impact are selected as the significant variables for the selection of the alternative solutions. The environmental impact assessment is developed by the Life Cycle Analysis (LCA); a scenario that is applied to all the alternatives is established in order to include the lifetime of the structure in the LCA. The results are presented divided according to the typologies and the pillars index, what provides the technician with a range of technically feasible alternatives. Applying the Pareto algorithm to the results simplifies and facilitates the choice as much as possible, reducing the alternatives presented as optimal to 25.2% of the total studied options.