Use this URL to cite or link to this record in EThOS:
Title: The influence of matrix stoichiometry on interfacial adhesion in composites for wind turbine applications
Author: Minty, Ross Forbes
ISNI:       0000 0004 7657 078X
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
It is well known that the fibre-matrix interface plays a key role in defining the mechanical properties of fibre composite materials. The ability to efficiently transfer stress between the matrix and the fibres is critical in ensuring the required performance level needed for advanced composite materials. Stress transfer across the fibre-matrix interface is often reduced to a discussion of 'adhesion'. Past discussions of thermosetting matrices have typically focussed on the chemistry of thematrix system, specifically the task of maximising the level of chemical bonding between the fibre and the matrix to produce the strongest interface. However, many authors have also commented on the potential for residual radial compressive stresses formed at the interface to be a significant contributor to the strength of the interface. There is still a significant weight of opinion that holds that even if these residual stresses at the interface can contribute to the stress transfer capability, then chemistry and chemical reactions must play an active role in defining their magnitude. As such it was the objective of this thesis to develop an understanding of how chemistry and residual stresses formed at the interface could be interrelated to influence the stress transfer capability of the interface. First an understanding was established of how the amine-to-epoxy group ratio (R) of an amine cured epoxy influences the thermomechanical properties of the matrix. Overall it was shown that the R value had a major influence over all of the thermomechanical properties studied. The glass transition temperature (Tg) was shown to change significantly depending on the R value, with a maximum of 87.3 °C observed at a value (R [approx. equal to] 1.25) slightly above the stoichiometric point. The matrix Tg decreased as the R value deviated from this value, approaching room temperature for the extreme ratios. Above Tg, the linear coefficient of thermal expansion (LCTE) was shown to reach a minimum at the stoichiometric ratio due to this ratio inducing the highest crosslink density. Below Tg, the R value appeared to have a less clear influence. The storage modulus (E') of the matrix was also shown to be affected by the R value, with the stoichiometric ratio possessing the largest magnitude of E' for temperatures 20°C above Tg. However, for temperatures 20 °C below Tg the storage modulus decreased in magnitude the closer the R value was to R [approx. equal to] 1.25. This was the ratio measured to possess the largest Tg value and thus the highest temperature at which E' was plotted relative to the other ratios. The effect of changing the R value on the interfacial shear strength (IFSS) was investigated using the microbond test. This was done in combination with changing the surface chemistry of the glass fibre and purity of the hardener. Results showed the magnitude of IFSS to be significantly affected by the R value, independent of the fibre sizing applied. The chemistry of the fibre sizing was shown to influence the maximum IFSS achievable and the R value at which it would occur, however the magnitude differences were not as significant. From these results it was concluded that the R value of the matrix has a greater influence than the chemistry of the fibre sizing in defining the level of adhesion at the fibre-matrix interface. The changes in IFSS were shown to correlate with Tg and the decrease in the contribution of residual thermal stresses at the interface. However, this contribution only represented a portion of the total IFSS value measured. It was concluded that other mechanisms, such as cure shrinkage stresses, must provide the remaining portion of IFSS shown. To expand upon this, the influence of temperature in combination with the other variables discussed was studied using the microbond test within a thermomechanical analyser. At lower temperatures the maximum IFSS value was shown to occur at R [approx. equal to] 1.0. The magnitude of IFSS was then shown to decrease as the R value deviated further from this value, again independent of the chemistry of the fibre sizing. Above Tg it was observed that a small value of IFSS remained which appeared to possess a linear relationship with the level of amine present within the Rratio. It was hypothesized that the magnitude of IFSS being greater for excess amineratios (R > 1) resulted from a combination of an increase in the level of hydrogen bonding and a variation in the shear failure behaviour of the matrix due to the differing crosslink densities. Using Nairn's model, a correlation was shown to exist between the IFSS values measured and the potential total contribution of residual stresses for R [approx. equal to] 1.0. However, as the ratios deviated further from R [approx. equal to] 1.0 the degree of correlation was shown to decrease. It was concluded that assumptions made by the model regarding the contribution of cure shrinkage stresses appeared to be oversimplifications once the R value deviated significantly from R [approx. equal to] 1.0. To address this a novel technique using hot-stage microscopy was used to measure the cure shrinkage of a minute epoxy droplet upon a single fibre during the curing process. The results showed that as the R value was increased, the level of cure shrinkage increased. Rheometry was then used to study the influence of the R value on the gel time of the matrix and applied to the data collected using the hot-stage method. This cure shrinkage data was then reapplied to the model where again good correlation was shown for R [approx. equal to] 1.0, yet the discrepancies regarding off-stoichiometric ratios remained. It was concluded that this may be due to a lack of understanding regarding the Tg of a microdroplet and the adhesion mechanisms of a rubbery statepolymer. Overall it was concluded that Nairn's model supports the hypothesis that residual radial compressive stresses at the interface can contribute significantly to the stress transfer capability of the interface. Since these stresses were shown to be affected by the R value it would also satisfy the need for chemistry to be involved significantly in some role.
Supervisor: Thomason, James ; Yang, Liu Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral