A Novel Nanocomposite for Building Façades

Modern building facades provide an innovative solution for designing high-performance building envelopes with increased energy efficiency and outstanding aesthetic appeal. The necessity for high thermal and acoustic insulation has led to the inclusion of significant amounts of combustible materials in many contemporary facade systems. In addition, factors such as cost-effectiveness, availability, and ease of manufacture also influence the choice to incorporate combustible materials. As a result, concerns have been raised about fire safety, specifically the possibility of a fire spreading throughout the building and to nearby structures. The issue of combustible envelope of the building has recently come to light as a significant construction challenge. This PhD thesis provides a comprehensive review of the current state of common facade systems, highlighting both their pros and cons. It also covers cladding materials, fire testing methods, and fire testing simulations used in this PhD. The literature review reveals the lack of a novel cladding material to enhance the fire performance of facade systems, a gap in research on the fire behaviour of three-dimensional (3D) glass fibre-reinforced polymers (GFRP) integrated with flame retardants and nanomaterials for facade applications, and limited utilisation of artificial intelligence (AI) in composite material design. The research presented in this thesis aims to address the aforementioned research gaps. This PhD project aims to develop innovative nano ceramic-based composite cladding material designed for building facades using state-of-the-art 3D GFRPs and phenolic resin. The 3D GFRP cladding offers several advantages by utilising ceramic-based flame retardants, including high thermal resistance, reduced flammability, and thermal stability when exposed to fires. The Taguchi Design of Experiment (DoE) is proposed for optimisation of the cladding material design, incorporating multiple relevant factors and levels in the manufacturing process. The four manufacturing design factors considered in the Taguchi DoE are the flame retardant type (Ceram-FR, ATH-FR, and 3DC-FR), percentage of flame retardant (0.0, 2.5, and 5.0 wt.%), curing regime (60 °C for 2 hours, 90 °C for 2 hours, and 60 °C for 1 hour followed by 90 °C for 1 hour), and dispersion technique (magnetic stirring, three roll milling, and ultrasonication), with each factor having three levels. A scanning electron microscopy (SEM) analysis was used for morphological characterisation of the polymer nanocomposites (without fibreglass fabric) and fibre-reinforced nanocomposite samples to investigate the effectiveness of different dispersion methods and curing regimes. In-plane tension, and in-plane and out-of-plane compression tests were conducted in reference to ASTM D3039, ASTM D3410, and ASTM C365-00, respectively, to evaluate the mechanical properties of the fibre-reinforced composites. Fire-under-tension load tests were conducted at two heat flux levels (19 and 40 kW/m²) within this project. The test results indicate that, under a heat flux of 19 kW/m², the fibre-reinforced composite with 5.0 wt.% Ceram-FR dispersed using magnetic stirring and cured at 60 °C for 2 hours, could maximise the failure time and failure temperature. For samples exposed to a heat flux of 40 kW/m², the analysis revealed the best performing formulation to be 5.0 wt.% ATH-FR dispersed using ultrasonication and cured at 60 °C for 1 hour followed by 90 °C for 1 hour. An investigation into the thermal behaviour and fire performance of the 3D GFRP cladding is undertaken using thermogravimetric analysis (TGA) and cone calorimetry conducted in accordance with the ISO 5660-1 standard. The TGA analysis of the nanopolymers showed that samples containing 5 wt.% Ceram-FR experienced the least weight loss at 850 °C compared to other polymer nanocomposites in both air and nitrogen environments. The optimised formulation (denoted as C10) for minimising critical response parameters of the 3D GFRP cladding during cone calorimetry, such as the ratio of peak heat release rate (PHRR) to the time to reach the peak, the ratio of PHRR to time to ignition, total heat release (THR), and effective heat of combustion, was found to include 5 wt.% Ceram-FR dispersed using magnetic stirring and cured at 60 °C for 1 hour followed by an additional 1 hour at 90 °C. The optimised 3D GFRP cladding demonstrated improved fire resistance, resulting in a lower PHRR and total THR compared to the nine cladding formulations fabricated using Taguchi DoE. This project also involves the development of AI-based simulations using machine learning algorithms to predict the performance of the 3D GFRP cladding when exposed to fires. The AI- predicted results can reproduce the experimental heat release rate data from the cone calorimeter test. Furthermore, the model is a useful tool to assist in calibrating the multiple input parameters required to analyse the 3D GFRP cladding using Fire Dynamics Simulator (FDS), including heat source and pyrolysis reaction properties. An FDS simulation study on a full-scale façade constructed from the optimised 3D GFRP cladding (C10) was conducted in accordance with the BS 8414-2 standard. The fire performance of the 3D GFRP façade system obtained from cone calorimetry was compared against the performance criteria prescribed in AS 5113 for the BS 8414-2 simulation. The results indicate that the façade system using the optimised 3D GFRP cladding could meet most of the prescribed performance criteria. The outcome for the criterion concerning internal temperatures measured at 5 m above the fire source was unclear due to modelling assumptions. It is essential to note that these simulation results should be considered as a preliminary assessment. A comprehensive full-scale façade system test is recommended to validate the FDS findings. The research presented in this PhD thesis contribute to the understanding of the fire performance of flame retardant 3D GFRP cladding materials designed for façade applications. New insights into the mechanical response and survivability of these composite cladding systems during fire were revealed. The research also offers valuable insights into the incorporation of AI in composite material design and the calibration of FDS models, significantly contributing to the development of AI-based modelling in predicting composite material properties during fire exposure.