Robustness of building floors under column removal. Experimental and theoretical study

Within research project ITSAFE, partially funded by the Spanish Government, two full-scale structures were built, one consisting of a single story two-span 7.00.14.00 m2 RC frame with a solid slab and another one consisting of a two-story 7.00x7.00 m2 frame also of RC with solid slabs. In the two-span frame, one of the central supports was first demolished using a pneumatic hammer resulting in rather limited damage (a 14-15 cm deflection at the removed support location).

However torsional cracks appeared at the interface between column and slab at one of the outer supports. When the second central support was removed the structure collapsed due to failure of the support-slab connection. The same type of cracking was observed in the two-story structure, where the column removal was dynamic, and a 22 cm deflection was measured.

From these full-scale experiments, several models of finite elements were developed for the study and analysis of the robustness of the structures. The results were published in three articles and two conferences. This thesis addresses a problem that is not yet resolved in the existing literature, and explain how the loads acting on a slab are resisted after the collapse of a column, in particular how much of the load is resisted by bending and what part is resisted by membrane effect. Existing studies looking at the problem in depth are limited to unidirectional elements, in which they materialize in an artificial longitudinal coercion which allows the axil to be measured and the membrane effect to be evaluated. The problem with real slabs is that there are no external elements that resist an axil, but rather a self-balancing system of axils is formed, which allows the generation of membrane force and the tensile points deform vertically more than the compressed points so the compressions generated at these points allow them to resist greater bending forces. This thesis addresses these issues by relying on the experimentation of the structures on 1:1 scale mentioned above and developed within the scope of the Group of Structural Engineering of the Polytechnic University of Madrid (UPM) and in theoretical analyses using finite elements calculations using LS-DYNA and Sofistik computing programs.

RESUMEN

Esta tesis aborda un problema que no está aún resuelto en la bibliografía existente que es explicar cómo se resiste las cargas actuantes en una losa en el caso en que se produzca el fallo de un pilar, en particular qué parte de la carga se resiste mediante flexión y qué parte se resisten mediante efecto membrana. Los estudios existentes que analizan el problema en profundidad se limitan a elementos unidireccionales, en los que se materializa de forma artificial una coacción longitudinal lo cual permite medir el axil y evaluar el efecto membrana. El problema de las losas reales es que no hay un elemento externo que resista un axil, sino que se forma un sistema autoequilibrado de axiles, que permiten generar fuerza de membran en la medida en que los puntos traccionados se deforman en vertical más que los puntos comprimidos y en la medida en que las compresiones que se generan en estos puntos permiten resistir esfuerzos de flexión mayores. Esta tesis aborda estos temas apoyándose en la experimentación de estructuras a escala 1:1 mencionadas anteriormente y desarrolladas en el ámbito de Grupo de Ingeniería Estructural de la Universidad Politécnica de Madrid (UPM) y en análisis teóricos mediante elementos finitos como modelos sofisticados utilizando LS-DYNA y Sofistik.