The fibre surface coating (or sizing) is one of the most crucial components involved in the manufacture of glass fibres and their composites. It plays a key role in determining the profitability, processability, and both the short- and long-term performance of a composite product. Given the importance of fibre sizing to the optimisation of the interface, there is a critical need to improve understanding of how this region is affected by fibre sizing composition. The objective of this thesis was to develop methodologies and micromechanical techniques to provide definition of the role of current and developmental glass fibre sizings in the interphase performance of glass fibre reinforced composites for wind turbine blade applications.A round-robin study of the adhesion enhancing capabilities of a wide range of glass fibre sizing components and full sizing packages has been investigated. It was found that glass fibre/epoxy adhesion was increased by application of a glycidoxypropylmethyldiethoxysilane (GPMES) coupling agent compared to amino-, methacryl-, and other epoxysilanes. The adhesion enhancing capabilities of a GPMES coupling agent may also have been responsible for comparable interfacial shear strength (IFSS) values for a series of epoxy-compatible full sizing formulations. However, the use of a matrix-incompatible full sizing resulted in adhesion properties lower than that of unsized fibres despite the presence of an undisclosed coupling agent. An apparent adhesion-inhibiting effect of the thermoplastic film former appeared to counteract any improvement due to the silane. Acetone extraction of fully sized fibres showed that IFSS was improved compared to the as-received fibres in all cases. Increased IFSS compared to the as-received fibres may have been related to removal of residual surface impurities, improved coupling agent homogeneity, or removal of sizing additives and processing agents whose contribution to the properties of the composite are not well defined. Interfacial adhesion values approximating and exceeding those attainable by use of a number of silane coupling agents and full sizing packages were achieved by treating glass fibres with an unreacted epoxy/acetone solution. The coating served as both a model polymeric film former and a post-sizing treatment. Increased IFSS as a result of the epoxy coating application may have been attributable to interdiffusion of the epoxy coating with the epoxy matrix material and enhanced cross-linking at the interface between unreacted DGEBA in the sizing and an amine curing agent.Thermal analysis, spectroscopic, and micromechanical testing methods were used to investigate the thermal degradation of a number of experimental and commercial glass fibre sizings. TGA of three fully sized epoxy-compatible glass fibres (SE1500, SE2020, W2020) indicated that the majority of the mass loss occurred in the 200–400°C range and was attributable to the decomposition of the polymeric film former component of the sizing. An initial 70–80 nm sizing layer was reduced by approximately 50% following 300°C heat treatment and was removed almost entirely following treatment at 500°C. Fourier-transform infrared spectra (FTIR) of thermally degraded fully sized glass fibre bundles indicated that the intensity of spectral bands attributable to an epoxy resin film former decreased linearly with increasing treatment temperature and were removed completely at temperatures of 300– 350°C. Spectra showed excellent correlation with TGA data that indicated that sizing mass loss in in the 200–400°C range was attributable to degradation of an epoxy film former. The accumulation of carbonyl groups may have been due to oxidation of the epoxy film former component in the sizing.Interfacial adhesion measurements indicated that IFSS had an inverse relationship with fibre treatment temperature and was concurrent with decomposition of the glass fibre sizing measured by TGA. Reduced IFSS appeared to onset at 300–350°C and at treatment temperatures of 400°C and above interfacial adhesion was comparable to that of unsized fibres, suggesting that the majority of the sizing components which influence IFSS had been removed. Spectra of fibres treated at 500°C indicated that some residues of degraded silane coupling agent material were present at the glass fibre surface, though the silane coupling agent may have been degraded to the extent that the adhesion-enhancing capabilities were lost. Interfacial adhesion may also have been inhibited after higher treatment temperatures by the accumulation of weakly bound oxidised film former/sizing material on the glass fibre surface.Further investigation of fibre sizing parameters was initially attempted using two epoxy resin matrices used in the production of wind turbine blades. It was discovered that the degree to which these resins formed cured droplets suitable for microbond testing was dependent on a modification to the recommended macroscale cure schedule. Microbond samples showed exceedingly low IFSS values when exposed to immediate heating, failed to cure, and deformed plastically under loading. Poor microdroplet curing and low apparent droplet glass transition temperature was attributable to a stoichiometric imbalance caused by evaporation of components essential to the polymerisation reaction. Off-stoichiometric droplet behaviour was further evidenced by a number of modifications to the curing schedule and sample preparation methodology. The inclusion of a room temperature pre-curing stage for a minimum of 2 h resulted in a profound increase in the apparent IFSS after which further standing times showed no significant improvement.Finally, a novel FTIR method was developed to address a fundamental need for a method to directly characterise the cure state of microbond droplets and investigate poor microscale curing performance in the industrial resin systems. Glass fibres were substituted by 50 μm diameter steel wire filaments in the preparation of microbond samples to improve signal clarity due to the combination of controllable and consistent increased microdroplet size and a favourable reflectance effect. Off-stoichiometric matrix compositions were used as models of the varying extents of curing agent evaporation that might exist in microbond droplets. DSC and FTIR were used to determine the relationship between degree of epoxy conversion and glass transition temperature. Droplet degree of cure was determined by FTIR and droplet glass transition temperature was estimated by comparison to off-stoichiometric data. FTIR spectra of microbond samples cured immediately at elevated temperature showed the presence of unreacted epoxy groups, reduced hydroxyl and secondary amine group accumulation, and a weaker etherification peak, commensurate with spectra indicating a non-stoichiometric epoxyamine network. Quantitative analysis of unreacted epoxy groups and determination of cure state indicated that samples cured immediately at high temperature had degrees of epoxy conversion in the region of 0.55, indicating a loss of up to 60% of the initial curing agent and a sub-ambient glass transition temperature. Conversely, samples that were allowed to pre-cure at room temperature for a minimum of 2 h before heating showed increased degree of cure (0.85–0.93) and droplet Tg some 80°C greater than that of immediately hot-cured samples. Good correlation was shown between micromechanical and spectroscopic methods, in that an increase in IFSS was commensurate with spectra indicating droplets were closer to the stoichiometric ratio and had higher degrees of cure, thus demonstrating a clear relationship between apparent interfacial adhesion and droplet cure state.
Date of Award | 23 Sept 2021 |
---|
Original language | English |
---|
Awarding Institution | - University Of Strathclyde
|
---|
Sponsors | University of Strathclyde |
---|
Supervisor | James Thomason (Supervisor) & Liu Yang (Supervisor) |
---|