Thermal behaviour of a novel cellular beam structural system in fire
Date
2018-03
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
AFRIKAANSE OPSOMMING: Die Suider-Afrikaanse Instituut vir Staalkonstruksie (SAISC) het 'n modulêre staalraamstruktuur ontwikkel wat bestaan uit sellulêre balke. Die selulêre balkstruktuur (SBS) bevat geen beton nie, wat vinniger oprigting moontlik maak en dus die konstruksiekoste aansienlik verminder. Die vervaardigde staal elemente word saamgestel om panele te
vorm wat relatief liggewig is, en wat maklik na die bouperseel vervoer kan word, waar hulle op verskillende maniere opgestel kan word. Die panele kan maklik in die toekoms hergebruik word en byvoegings aan bestaande geboue kan gemaak word. Dus kan meer
ekonomiese staalgeboue in 'n korter tydperk gebou word. Soos vir alle strukture, vereis
die SBS egter 'n brandweerstandsgradering. As gevolg van die unieke en komplekse opset
van die SBS, sal standaard-voorgeskrewe brandontwerpmetodes tot konserwatiewe en
onekonomiese ontwerpe lei, in terme van passiewe brandbeveiligingsmaatreëls. Dus is 'n meer gevorderde prestasie-gebaseerde metode gebruik in hierdie projek, wat van eindige element (EE) ontledings gebruik maak om die gedrag van die SBS in 'n brand te bepaal.
Hierdie tesis, wat deel uitmaak van 'n groter navorsingsprojek, het die termiese gedrag
van die SBS ondersoek. Termiese hitte-oordragontledings is in ABAQUS uitgevoer om die staalbalke en die binne-vloer se temperature te bepaal. Die EE-modelle is bevestig op grond van 'n eksperimentele studie, waarin vier kleinskaalse standaard-vuurtoetse uitgevoer
is. Die eksperimentele en numeriese resultate toon oor die algemeen goeie korrelasie, met variasies in resultate binne die grense van wat verwag word in eksperimentele brandnavorsing.
Voorlopige valideringstudies vanuit die literatuur is ook uitgevoer, waaruit goeie korrelasies behaal is. Dit het die toepaslikheid van die hitte-oordrag modelleringsprosedures
in ABAQUS bevestig en sluit die modellering van gaping-bestraling in.
Die vloerstelsel van die SBS, wat groot balke bevat wat nie by die kleinskaalse eksperimentele toetse ingesluit is nie, is ondersoek deur EE-modelle te ontwikkel vir elk van
die balk elemente in die vloer, wat die vloerborde en staalplate insluit. 'n Addisionele EE-model is ontwikkel vir elk van die balkmodelle, waarvoor die plafonbord weggelaat is, wat veroorsaak dat die staalplaat direk aan die vuur blootgestel word. Dit verteenwoordig
'n mees ongewenste geval waarin die integriteit van die plafonbord heeltemal misluk en af val tydens 'n brand. Dit word voorspel dat die temperatuurverspreiding deur die
dwarssnitte oneweredig is, met die onderste flens- en boonste flenstemperature wat wissel
tussen 220°C en 48°C vir die primêre and sekondêre sellulêre balke, en tussen 176°C en
95°C vir die kanaalbalke, wat die vloerstelsel ondersteun. Sonder die plafonbord word
'n verskil in temperatuur-toename van 300°C tot 500°C verwag, in vergelyking met die
plafonbordmodelle. Dit beklemtoon die belangrikheid van die plafonbord en die invloed
wat dit op die staalbalk-temperature het.
Daar is gevind dat die vloerstelsel, soos tans gespesifiseer is, misluk het met betrekking
tot die isolasiekriteria, waarin die temperatuurstyging in die binne-vloerbord die toelaatbare temperatuur met ongeveer 60°C oorskry. 'n Parametriese ondersoek is dus uitgevoer om alternatiewe oplossings te bepaal wat voldoende isolasie vir die vloerstelsel
sal bied. Verskeie bordmateriale en -diktes is gemodelleer, sowel as verskillende staalplaatstelsels.
'n Parametriese vuur is ook ondersoek om die effek van die verkoelingsfase op die staal elemente te bepaal. Daar is gevind dat die dikte van die plafonbord en die
binne-isolasiebord die belangrikste invloed op die temperatuurstyging binne die vloerstelsel
het. Deur 'n dikker plafonbord te gebruik, voldoen die temperatuurstyging op die bedekte oppervlakke aan die isolasie kriteria vir 'n 20 mm gips tipe X plafonbord, asook
vir 'n 25 mm Promatect-H Kalsiumsilikaat- (CaSi) bord, mits die integriteit van die borde behoue bly. Daar is opgemerk dat die CaSi-bord moontlik beter kan presteer as die gipsbord in terme van integriteit, as gevolg van die hoër voginhoud teenwoordig in die gipsbord, wat tot krimpingskrake kan lei. Die konneksiepunte van die plafonbord moet die verplasings wat deur die SBS ondervind kan word, akkommodeer om die integriteit te
behou.
Die bevindings in hierdie tesis kan gebruik word as grondslag vir verdere navorsing en die beplanning en uitvoering van 'n volskaalse brandtoets op die SBS. Daar is gevind dat die SBS die potensiaal het om voldoende brandweerstand te bied, mits die integriteit behou
word, en sodoende die moontlikheid van strukturele faling verlaag.
ENGLISH ABSTRACT: The Southern African Institute of Steel Construction (SAISC) has developed a modular steel frame structure that consists of cellular beams. The cellular beam structure (CBS) contains no concrete, which allows for a shorter construction duration, thereby significantly reducing costs. The fabricated steel members are assembled to form panels which are relatively lightweight, and which can be easily transported to site where they can be configured in various ways. The panels can easily be re-used in the future and additions to existing buildings can be made. Therefore, more economical steel buildings can be built in a shorter duration. However, as for all structures, the CBS requires a fire resistance rating. Due to the unique and complex configuration of the CBS, standard prescriptive fire design methods may lead to over-conservative and uneconomical designs, in terms of passive fire protective measures. Hence, a more advanced performance-based method has been used in this thesis, which incorporates finite element (FE) analyses and experimental testing to determine the behaviour of the CBS in fire. This thesis, which forms part of a larger research project, has investigated the thermal behaviour of the CBS. Thermal heat transfer analyses have been performed in ABAQUS to determine the steel beam temperatures and the interior floor temperatures. The FE models have been validated based on an experimental study, in which a total of four smallscale standard fire tests were conducted. Experimental and numerical results generally show good correlation, with variations in results within the bounds of that expected in experimental fire research. Preliminary validation studies from the literature have also been performed, from which good correlations were obtained. This validated the suitability of the heat transfer modelling procedures in ABAQUS, and included the modelling of cavity radiation. The sandwich foor system of the CBS, which includes large beams not included in the small-scale experimental tests, has been investigated by developing FE models for each of the beam sections within the floor, which included the floor boards and steel sheeting. An additional FE model was developed for each of the steel floor beams, for which the ceiling board was neglected, thus leaving the steel sheeting directly exposed to the fire. This represented a worst-case scenario, in which the integrity of the ceiling board failed completely and detaches during a fire. The temperature distribution through the beam cross-sections is predicted to be highly non-uniform, with the bottom flange and top flange temperatures ranging between 220°C and 48°C for the primary and secondary cellular beams and between 176°C and 95°C for the channels that support the flooring system. Without the ceiling board, a higher temperature increase of 300°C to 500°C is expected, when compared to the ceiling board models. This emphasises the importance of the ceiling board and the influence it has on the steel beam temperatures. It was found that the foor system failed, as currently specified, with regards to the insulation criteria, in which the temperature increase in the interior floor board exceeded the allowable temperature by approximately 60°C. Therefore, a parametric investigation has been performed to determine alternative solutions that will provide sufficient insulation to the floor system. Various board materials and thicknesses were modelled, as well as different steel decking systems. A parametric fire was also investigated to determine the effect of the cooling phase on the steel members. It was found that the thickness of the ceiling-board and the interior insulation board have the most significant influence on the temperature increase within the floor system. By using a thicker ceiling board, the temperature increase on the unexposed surfaces satisfy the insulation criteria for a 20 mm gypsum type X ceiling board, as well as for a Promatect-H Calcium-Silicate (CaSi) board, while assuming that the integrity of the boards are maintained. It was noted that the CaSi-board may perform better than the gypsum board in terms of integrity, due to the higher moisture content present in the gypsum board, which can lead to shrinkage cracks. The fixities of the ceiling board should be able to accommodate the de ections which may be experienced by the CBS, in order to maintain integrity. The findings in this thesis can be used as a foundation for further research, and the planning and execution of a full-scale fire test on the CBS. Ultimately, it was found that the CBS has the potential to provide sufficient fire resistance, based on the assumption that the integrity is maintained, thus reducing the possibility of structural failure.
ENGLISH ABSTRACT: The Southern African Institute of Steel Construction (SAISC) has developed a modular steel frame structure that consists of cellular beams. The cellular beam structure (CBS) contains no concrete, which allows for a shorter construction duration, thereby significantly reducing costs. The fabricated steel members are assembled to form panels which are relatively lightweight, and which can be easily transported to site where they can be configured in various ways. The panels can easily be re-used in the future and additions to existing buildings can be made. Therefore, more economical steel buildings can be built in a shorter duration. However, as for all structures, the CBS requires a fire resistance rating. Due to the unique and complex configuration of the CBS, standard prescriptive fire design methods may lead to over-conservative and uneconomical designs, in terms of passive fire protective measures. Hence, a more advanced performance-based method has been used in this thesis, which incorporates finite element (FE) analyses and experimental testing to determine the behaviour of the CBS in fire. This thesis, which forms part of a larger research project, has investigated the thermal behaviour of the CBS. Thermal heat transfer analyses have been performed in ABAQUS to determine the steel beam temperatures and the interior floor temperatures. The FE models have been validated based on an experimental study, in which a total of four smallscale standard fire tests were conducted. Experimental and numerical results generally show good correlation, with variations in results within the bounds of that expected in experimental fire research. Preliminary validation studies from the literature have also been performed, from which good correlations were obtained. This validated the suitability of the heat transfer modelling procedures in ABAQUS, and included the modelling of cavity radiation. The sandwich foor system of the CBS, which includes large beams not included in the small-scale experimental tests, has been investigated by developing FE models for each of the beam sections within the floor, which included the floor boards and steel sheeting. An additional FE model was developed for each of the steel floor beams, for which the ceiling board was neglected, thus leaving the steel sheeting directly exposed to the fire. This represented a worst-case scenario, in which the integrity of the ceiling board failed completely and detaches during a fire. The temperature distribution through the beam cross-sections is predicted to be highly non-uniform, with the bottom flange and top flange temperatures ranging between 220°C and 48°C for the primary and secondary cellular beams and between 176°C and 95°C for the channels that support the flooring system. Without the ceiling board, a higher temperature increase of 300°C to 500°C is expected, when compared to the ceiling board models. This emphasises the importance of the ceiling board and the influence it has on the steel beam temperatures. It was found that the foor system failed, as currently specified, with regards to the insulation criteria, in which the temperature increase in the interior floor board exceeded the allowable temperature by approximately 60°C. Therefore, a parametric investigation has been performed to determine alternative solutions that will provide sufficient insulation to the floor system. Various board materials and thicknesses were modelled, as well as different steel decking systems. A parametric fire was also investigated to determine the effect of the cooling phase on the steel members. It was found that the thickness of the ceiling-board and the interior insulation board have the most significant influence on the temperature increase within the floor system. By using a thicker ceiling board, the temperature increase on the unexposed surfaces satisfy the insulation criteria for a 20 mm gypsum type X ceiling board, as well as for a Promatect-H Calcium-Silicate (CaSi) board, while assuming that the integrity of the boards are maintained. It was noted that the CaSi-board may perform better than the gypsum board in terms of integrity, due to the higher moisture content present in the gypsum board, which can lead to shrinkage cracks. The fixities of the ceiling board should be able to accommodate the de ections which may be experienced by the CBS, in order to maintain integrity. The findings in this thesis can be used as a foundation for further research, and the planning and execution of a full-scale fire test on the CBS. Ultimately, it was found that the CBS has the potential to provide sufficient fire resistance, based on the assumption that the integrity is maintained, thus reducing the possibility of structural failure.
Description
Thesis (MEng)--Stellenbosch University, 2018.
Keywords
Finite element analysis, Heat transfer, Sandwich floor system, Thermal analysis, Building materials -- Testing, Condition monitoring (Structural engineering), UCTD