Acceptable life safety risks associated with the effects of gas explosions on reinforced concrete structures
- Autores: Ramón Hingorani
- Directores de la Tesis: Peter Tanner (dir. tes.), Carlos Zanuy Sánchez (codir. tes.)
- Lectura: En la Universidad Politécnica de Madrid ( España ) en 2017
- Idioma: español
- Tribunal Calificador de la Tesis: Luis Albajar Molera (presid.), David Izquierdo López (secret.), Joan Ramón Casas Rius (voc.), Dimitris Diamantidis (voc.), Angel Arteaga Iriarte (voc.)
- Materias:
- Enlaces
- Tesis en acceso abierto en: Archivo Digital UPM
- Resumen
Accidental actions on structures may be characterized as low probability - high consequence events. On one hand, their occurrence during the envisaged design working life of the structure is unlikely. On the other, if not appropriately accounted for, the associated effects on structures might entail significant damage. Since moreover such effects are subjected to high uncertainties, decision-making related to structural safety accounting for accidental actions is generally difficult and prone to be based on irrational grounds. Among such actions, gas explosions, are a good example. Despite the continuous modernization of gas installations and appliances, available statistics from different countries show that the occurrence rate of such explosions in buildings does not seem to decrease in a significant way. While the hazard potential is known and recognized, and although dealt with in many design codes, gas explosions are seldom accounted for in the design and evaluation of ordinary building structures. The low occurrence probability evokes reluctance to allocate resources to mitigate the associated risks, which, as a consequence, are often ignored and sometimes consciously accepted. The question if “doing nothing” is a justified practice cannot be easily answered however, since under the implicit approach adopted in everyday practice for verification of structural safety the risks are not quantified nor are acceptable risk levels established.
On this background, the study aims at exploring methods and tools for the practical application of explicit risk analysis in connection with gas explosions in buildings. A procedure is established for quantification of implicitly acceptable structure-related risks to persons, based on the probability of structural collapse and the consequences of such a failure in terms of loss of human life. The procedure adopted is applied to a representative set of building structures with RC members (beams and columns), which is obtained by varying the parameters with the greatest effect on design within reasonable limits. Following their identification, the most relevant hazard scenarios to these members are represented in terms of limit state functions (LSF). Based on the established LSF’s, a strict design (Ed = Rd) according to a consistent set of codes is carried out, so that structural member performance complies exactly with the safety requirements that reflect current best practice. The basic variables involved in the LSF are stochastically characterized, where special attention is paid to the dynamic effects associated with the explosion-induced high loading rates on the members, such as the contribution of inertia forces, energy dissipation and strain rate-sensitive material behaviour.
Quantification of these effects is addressed in a deterministic dynamic analysis where the explosion load is represented as an idealised pressure-time function, compatible with simplified models. Under consideration of dynamic material properties, member flexural response is obtained assuming a single degree of freedom system, whereas the reaction forces, representative for the shear forces, are determined from the dynamic equilibrium formulation applied to the members themselves. For the beams, a comparative study is conducted, where the deployed simplified models are validated by means of non-linear finite element analysis. The analysis of the columns under dynamic bending moment-axial force interaction requires a specific solution algorithm that accounts for the axial force dependent formulation of structural resistance under consideration of both the material- and geometrical non-linearities involved.
In the subsequent reliability analysis of the structural members, the mentioned algorithm is coupled to a purpose-developed FOSM-based iterative procedure in order to obtain the most likely failure point for the established LSF. Taking account of the occurrence probability of a gas explosion event, implicitly acceptable structural failure probabilities for both columns and beams are derived and analyzed in the light of target ceilings demanded by structural codes. The findings suggest significant scope for a more rational formulation of design rules for accidental situations related to gas explosions.
For the estimation of the structural failure consequences, a regression model is developed from previously compiled and statistically evaluated data on explosion-induced structural collapse scenarios in buildings. The model delivers estimations for the number of fatalities as a function of the area affected by structural collapse and the occupancy rate of this area. Reasonable hypothesis are adopted in order to account for the possibility of system collapse given a local member failure.
Subsequently, implicitly acceptable risk profiles are obtained for each of the representative building structures where account is taken of the fact that, in addition to the considered accidental load scenarios, certain member failure modes might be triggered by persistent load arrangements associated with normal building use conditions. Acceptance criteria for structure-related life safety risks are deduced from the findings. Such criteria facilitate the adoption of rational decisions on both, the need and the appropriate choice of risk-reduction measures to counteract the effects of gas explosions in buildings. The design of key elements, upon which depends the stability of the structure, or a large part of it, may be one of these strategies. For this purpose, acceptable risks are translated into target failure probabilities for individual structural members, defined as a function of the potential failure consequences. In spite of their notional character, such target values provide a rational basis for the calibration of the implicit rules in structural codes and standards for verification of structural safety in relation to gas explosions.
