The disproportionate collapse of building structures has become a growing concern amid the increasing frequency and intensity of extreme events. Conventional design strategies address this issue by ensuring extensive structural continuity, allowing loads to be redistributed after local failures. However, when major initial damage occurs, such continuity can inadvertently promote failure propagation rather than preventing it. This limitation can be overcome by designing buildings to segment in a controlled manner once a collapse is triggered in order to isolate the initial damage and prevent its propagation.

This thesis introduces a novel performance-based framework for enhancing the robustness of buildings through controlled structural segmentation. The approach aims to limit collapse propagation by intentionally isolating damaged portions of the structure once a failure occurs.

An extensive forensic analysis of forty real building collapses was first carried out to identify the most common hazards, initial failures, and propagation mechanisms leading to disproportionate collapse. This analysis enabled the definition of critical scenarios to be considered in the assessment of segmentation strategies and supported the calibration of consequence models for estimating human and economic losses.

The proposed framework is structured in two main phases. The first involves the identification of feasible segmentation configurations based on geometric and structural criteria. The second evaluates their cost-effectiveness through a risk-based cost-benefit analysis that compares the expected risk reduction achieved by segmentation with its implementation cost. Designed to be threat-independent, the methodology provides a quantitative basis for determining when segmentation is a viable robustness-enhancing measure for framed buildings. Its application to a reinforced concrete case study demonstrated that segmentation can be a cost-effective strategy for mitigating collapse risks associated with a wide range of natural and anthropogenic hazards.