No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

In conventional finite element analysis of reinforced concrete the steel bars are normally assumed to lie along the concrete element edges and very often the bond gripping the steel to the concrete is assumed to be infinitely stiff. The first assumption makes it difficult to model all steel bars leading to the inclusion of only a few representative bars. Shear reinforcement is usually ignored. Thin concrete cover also creates difficulty by causing long thin finite elements in that region. The second assumption does not reflect the true behaviour of the system. In this research a new method for the modelling of steel in reinforced concrete by finite element analysis has been developed which allows all steel reinforcement to be included in the analysis. The method is based on modelling the steel and concrete separately, the two materials being interconnected by the bond forces between them. Thus, bond stiffness is naturally included in the analysis. Such interconnection of steel and concrete is achieved by an interface bond matrix which is derived from the relative displacements between the steel and the concrete at the steel nodes. A linear bond slip relation is assumed for the bond, and a linear stress strain relation is assumed for the concrete and the steel. The work has extended also to nonlinear bond stress-slip relation. Concrete is represented by 8-noded isoparametric quadrilateral elements, and the steel is represented by two noded bar elements. The bond is represented by springs joining each steel node to all 8-concrete nodes. The solution of the resulting system of equations is achieved in an iterative manner which converges quite rapidly, and which requires less computation than the direct solution needs. Three types of problems are analysed in two dimension to demonstrate the application of this new method. These are beam, cantilever and pullout problems. The first two, being real problems, demonstrate the ability of the method to handle complex steel arrangements, thin concrete covers and anchorage of steel, while the third problem shows the application of load to the steel rather than to the concrete. Concrete and steel deformations and stresses are calculated at their nodes. Bond stresses are given at all steel nodes. In the nonlinear bond analysis, deterioration of bond will be demonstrated in pullout and pushout tests at high loads.

The mechanical properties of alumina trihydrate (ATH)-filled epoxy resin at loadings of up to 100 parts by weight ATH per hundred of resin (epoxy and hardener) (pphr) have been investigated. A low peak exotherm, increased Young's modulus and increased critical strain energy release rate (G[sub]IC) and critical stress intensity factor (K[sub]IC) can be achieved by incorporating a dispersion of ATH into an epoxy resin. However, the high filler loadings required for effective fire resistance reduce tensile strength and elongation. Tensile modulus increases with filler loading in line with previous studies and theoretical equations. However, the tensile strength is higher and the ultimate elongation lower than current theories predict. The tensile and fracture process in ATH-filled epoxy follows linear elastic fracture mechanics, but can be considered in two parts. The initiation of a crack occurs from a large critical flaw, either as a large particle or agglomerations of particles. A flaw can also be formed on the application of a tensile load, when large stress concentrations cause localised microcracking of the matrix. The propagation of a flaw requires more energy and is dependent on several possible mechanisms. Shear yielding and associated crack blunting are shown to be the most important mechanisms, whilst minor contributions from matrix microcracking and debonding of ATH particles are possible. The absence of crack pinning in this study is believed to be due to the inherently weak nature of ATH particles. The presence of a 10pphr rubber dispersion in ATH-filled epoxy only increases the values of G[sub]IC and K[sub]IC at low filler loadings. Amine-terminated butadiene acrylonitrile rubber (ATBN)-modified epoxy matrix exhibits little adhesion to ATH and therefore the efficiency of stress transfer between particle and matrix is reduced, diminishing shear yielding.

No description available.

The development of synthetic adhesives based on epoxy resins has opened new possibilities for bonding structural materials together. The present work was concerned with the use of epoxy resins to strengthen reinforced concrete beams by externally bonded steel plates. It was found in the first part that the assessment of the properties of the epoxy adhesive is of paramount importance as they varied considerably with the thickness of the test specimen and the rate of loading. The adhesive proved to offer a bond stronger than concrete in shear and resulted in a composite action between the beams and steel plates. Preloading the beams prior to strengthening them did not have any adverse effect on their behaviour. The added strength from the plates was fully exploited even in beams which were held under a preload of 70% of their ultimate strength while being strengthened. Stopping the plate in the shear span, short of the support, created a critical section where premature bond failure occurred beyond a certain plate thickness. Failure was caused by the combination of high peeling and bond stresses present in the region where the plate was stopped. These stresses were due to the transfer of tensile forces from the plate to the bars in that region and were higher with thicker plates. Bonding steel plates on the tension face of the beams increased their shear capacity by 9 to 15%. This may have been due to dowelling action from the plates which had a greater contact area with concrete than an equivalent amount of internal steel bars. The use of externally bonded steel as shear reinforcement was effective but requires further investigation. The external web strips failed prematurely as compared to equivalent stirrups. The long term deformations in plated beams were highly affected by the conditions of their environment but despite 47 month exposure no visual deterioration of the concrete-epoxy-steel joint was observed.