City Research Online

Thermo-mechanical damage modelling for collapse assessment of steel buildings under blast and fire loads

Lu, W. (2019). Thermo-mechanical damage modelling for collapse assessment of steel buildings under blast and fire loads. (Unpublished Doctoral thesis, City, University of London)

Abstract

The aim of this research is to develop a coupled thermo-mechanical damage model for implementation in finite element software in support of fire-induced collapse assessment of steel structures. The need for properly modelling steel deterioration behaviour remains a challenging task in structural fire engineering because of the complexity inherent in the damage states of steel at large strains and high temperatures. A fully three-dimensional damage-coupled constitutive model is developed based on the hypothesis of effective space elastoplasticity and isotropic damage theory. The coupled damage is simulated by a coupling formulation between a mechanical damage component and a thermal damage component in attempt to capture the coupled damage growth under combined actions of mechanical loading and fire loading. The proposed damage model comprises a limited number of parameters that could be identified at unloading slopes of stress-strain relationships through tensile coupon tests. Alternatively, an inverse analysis type of calibration procedure could be adopted when coupon test data is unavailable. The proposed damage model is successfully implemented in the finite element software ABAQUS and calibrated with a comprehensive range of experimental results and established numerical results. The damage-affected structural response is accurately reproduced under various loading conditions and a wide temperature range, demonstrating that the proposed damage model is a useful tool in giving a realistic representation of steel deterioration behaviour under combined actions of fire and mechanical loads.

Three-dimensional FE models of a five-storey and a ten-storey steel-framed office building are developed in ABAQUS and the proposed damage model is adopted in assessing their susceptibility to progressive collapse. Three types of accidental scenarios are investigated : (i) fire only scenario, (ii) post-blast fire scenario, and (iii) fire-triggered explosion scenario. The location of the compartment where triggering loads occur is varied and the most vulnerable location is at the mid-height of both building systems. Estimation of ultimate failure time by incorporating damage model with the suggested damage parameter set has the potential to be utilized as a useful tool in helping designers to determine how much time is realistically available for evacuation before progressive collapse occurs in this type of buildings. Results show that the proposed damage model significantly affects the limit state of steel buildings under fire, and especially under combined actions of blast and fire. Compared to conventional numerical approaches, the consideration of coupled thermo-mechanical damage accumulation results in an 8.25% ∼ 23.47% decrease of collapse resisting time. A key finding from this study is that the alternative load path, which is a crucial factor in deciding the survival of buildings upon local column failure, may be severely compromised due to the coupled thermo-mechanical damage propagation in surrounding columns. Based on the identified collapse mechanisms, effective strategies are suggested to improve the survivability of buildings under blast and fire.

Publication Type: Thesis (Doctoral)
Subjects: T Technology > TH Building construction
Departments: Doctoral Theses
Doctoral Theses > School of Mathematics, Computer Science and Engineering Doctoral Theses
School of Mathematics, Computer Science & Engineering > Engineering > Civil Engineering
URI: http://openaccess.city.ac.uk/id/eprint/21816
[img] Text - Accepted Version
This document is not freely accessible until 1 March 2022 due to copyright restrictions.

To request a copy, please use the button below.

Request a copy

Export

Downloads

Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login