Thesis (MEng)--Stellenbosch University, 2015. / ENGLISH ABSTRACT: FRC (Fibre-reinforced concrete) has become a common form of secondary and even primary reinforcing in some applications throughout the world. In South Africa, the structural applications are limited primarily to steel fibres while cheaper, lighter and more durable synthetic fibres have been side-lined due to low stiffnesses. The purpose of this research project is to investigate the behaviour of synthetic fibre-reinforced concrete (SynFRC) using fibres which are locally available in South Africa, and to propose a performance-based specification and test method for the use of these fibres. In order to achieve this, single fibre pull-out tests were performed on four locally available polypropylene fibres. It was found that the average bond stresses of the fibres are influenced primarily by the fibre cross sectional shape, longitudinal geometry and surface treatment, and secondarily by the aspect ratio. The W/C ratio had little effect on the single fibre performance of non-treated fibres, but appeared to have a slight effect on the single fibre performance of the surface treated fibre. From the experimental results, the highest fibre bond stress will be generated by using a fibre with an X-shaped cross section, longitudinal crimping and applying a surface treatment to this fibre. It also appears that the bond stress distribution for flat fibres is close to uniform, while the bond stress distribution for non-flat crimped fibres has a high mechanical interlock component at the surface end. Macro-mechanical performance tests were performed by means of the BS EN 14651 (2007) three point beam bending test and the ASTM C1550 (2012) Round Determinate Panel Test (RDPT). These tests were selected following a thorough literature review. The RDPT was found to be more consistent and able to identify trends which the three point beam bending test could not. In addition, the three point beam bending test’s most popular output, the Re,3 value tended to be misleading with varying W/C ratios, and it is recommended that the equivalent flexural tensile strength be used instead if the three point beam bending test is used. The macro-mechanical testing showed that increasing the fibre dosage did increase post-cracking performance. The flat fibres’ performance was significantly better than that of the non-flat fibres, and also increased at a faster rate with increasing fibre dosage. The post-cracking performance decreased with increasing W/C ratios and increasing aggregate sizes. The macro-mechanical performance was inversely proportionate to the single fibre performance. The macro-mechanical performance decreased with increasing fibre bond stress, and increased with increasing equivalent diameter, which equates to fewer fibres in a set volume of fibres. Finally, basic principles were developed from the data. These principles were used to predict the RDPT and three point beam bending test performance parameters based on fibre dosage, single fibre properties (bond stress and equivalent diameter), W/C ratio and aggregate size from the available data. The principles can be further refined with more experimental data. / AFRIKAANSE OPSOMMING: Vesel-gewapende beton word regoor die wêreld as ’n algemene vorm van sekondêre en selfs primêre versterking gebruik. In Suid-Afrika is die strukturele toepassings hoofsaaklik tot staal vesels beperk, terwyl goedkoper, ligter en meer duursame sintetiese vesels vermy word as gevolg van lae styfhede. Die doel van hierdie navorsingsprojek is om die gedrag van sintetiese-veselversterktebeton (SynFRC) te ondersoek deur gebruik te maak van vesels wat in Suid-Afrika beskikbaar is, en 'n prestasie-gebaseerdespesifikasie en toetsmetode vir die gebruik van sintetiese vesels voor te stel. Enkelveseluittrektoetse is op vier plaaslik beskikbare polipropileen vesels uitgevoer. Daar is gevind dat die gemiddelde verbandspanning van die vesel hoofsaaklik deur die vesel deursnee vorm, lengte meetkunde en oppervlak behandeling beïnvloed word, en tweedens deur die aspek verhouding beïnvloed. Die W/C-verhouding het min effek op die enkelveselprestasie van nie-behandelde vesels, maar het 'n effek op die enkeleveselprestasie van die oppervlak-behandelde vesel gehad. Die eksperimentele resultate wys dat die hoogste vesel verbandspanning deur 'n vesel met 'n X-vormige deursnit, lengte krimping en toepassing van 'n oppervlak behandeling gegenereer sal word. Dit blyk ook dat die verbandspanningverspreiding vir ’n plat vesel naby aan uniform is, terwyl die verbandspanningverspreiding vir ’n nie-plat gekrimpde vesel 'n hoë meganiese grendeling komponent op die oppervlak ente het. Makro-meganiese prestasietoetse is uitgevoer deur middel van die BS EN 14651 (2007) driepuntbalkbuigtoets en die ASTM C1550 (2012) RDPT. Hierdie toetse is ná ’n deeglike literatuuroorsig gekies. Die RDPT is meer konsekwent en is in staat om neigings te identifiseer wat die driepuntbalkbuigingtoets nie kan nie. Daarbenewens, met wisselende W/C verhoudings, is die driepuntbalkbuigtoets se gewildste resultaat, die Re,3-waarde geneig om misleidend te wees. Dit word aanbeveel dat die ekwivalentebuigtreksterkte in plaas van die Re,3-waarde as die drie punt balk buig toets resultaat gebruik word. Die makro-meganiesetoets het getoon dat die verhoging van die veseldosis ’n toename in na-krakingprestasie veroorsaak. Die plat vesels se prestasie was aansienlik beter as die van nie-plat vesels, en het met 'n toenemende veseldosis teen 'n vinniger koers verhoog. Die na-krakingprestasie het met toenemende W/C en die verhoging van die klip grootte afgeneem. Die makro-meganieseprestasie was omgekeerd eweredig aan die enkelveselprestasie. Die makro-meganieseprestasie het met toenemende vesel band stres verminder, en het met 'n toenemende gelykstaande deursnee (wat gelykstaande is aan minder vesel in 'n stel volume van vesel) vergroot. Ten slotte is basiese beginsels uit die data ontwikkel. Hierdie beginsels is gebruik om die RDPT en driepuntbalkbuigtoets prestasieparameters gebaseer op veseldosis, enkelveseleienskappe (verbandspanning en ekwivalentediameter), W/C-verhouding en klip grootte van die beskikbare data te voorspel. Die beginsels kan met meer eksperimentele data verder verfyn word.