Summary:
This report presents the suitability evaluation of various theoretical approaches for the structural design of the flexural strengthening of bridge deck slabs with textile reinforced concrete (TRC) at ultimate limit state (ULS). TRC is a composite material which combines the high tensile strengths of the fiber textiles with the bond capacity and the mechanical and thermal protection provided by fine-grained concrete (i.e. cement-based mortars). Besides being used in manufacturing of new TRC elements, these materials can also be used as strengthening for existing reinforced concrete (RC) elements. The most frequently used textiles in this type of applications are made from Carbon, Basalt, Glass and PBO (poly-phenylene-2,6-benzobisoxazole a.k.a. Zylon) fibers. Such fibers are bundled together into highly resistant yarns with the help of resin. These yarns are then woven into textiles, usually applied in the form of meshes. In flexural strengthening of existing RC elements, the preparation of the support surface plays an important role. It has to be roughened by hydro-demolition before applying the strengthening system in order to ensure a monolithic behavior at the interface. Once this process is completed, the first mortar layer can be applied, usually by spraying, followed by successive textile-mortar layers until the desired strengthening level is reached. Besides offering a protection layer to the steel reinforcement of the existing RC element, the mortar plays an important structural role as anchorage and bond agent for the textile yarns. The anchorage by bond was studied in an experimental campaign, complementing existing experimental data from the literature and allowing to calibrate a series of parameters of the theoretical model proposed in the literature. The study concluded with an analytical proposal for determining the bond anchorage capacity of the textile yarns embedded in mortar. More details on the proposal and the experimental campaign are given in this report. To identify a suitable theoretical approach for dimensioning the flexural strengthening of existing one-way slab elements, a database with experimental results was created by gathering data from the literature. This database contains data from approximatively 150 experiments on strengthened elements and unstrengthened reference elements. Each entry describes 58 different parameters covering input data such as test setup, cross-sectional geometry, material properties and experimental failure loads. In a first step, the collected data was evaluated empirically to statistically determine the type of textile materials used, the types of observed failure and increases in flexural capacity. The latter increases, on average and independently of the textile material, by 20-25% per textile layer. In a second step, different theoretical approaches for ULS design of the flexural strengthening of slab elements with TRC were applied to evaluate their suitability by comparing the theoretical flexural capacity with the experimental results. A first approach considers rigid bond between textile and surrounding concrete matrix and results in a large overestimation of the flexural capacity and a high coefficient of variation (COV). This legitimates the introduction of efficiency coefficients on the axial stiffness of the textiles, as also proposed in literature, which are due to the particularities of internal bond of the textiles but also to the bond between textile yarns and the surrounding mortar. Thus, the second approach applied goes into the other extreme, considering no bond between textile and matrix but end-anchorages of the textile being provided by the mortar. This approach returns conservative estimations of the flexural capacity, yet it is not satisfactory due to a high COV. The third and fourth approaches applied evaluate the consideration of friction and bond coefficients, respectively, between textile and surrounding matrix. However, due to their inapplicability to 3-point bending configurations and high COVs, they were also considered unsatisfactory. Finally, an analytical method is retained that introduces strain limits for the flexural tension forces in the textiles while the existing steel reinforcement is yielding, thereby relating to approaches known from fiber-reinforced polymer strips. Thanks to thoughtful calibration of the strain limits, this approach results in an average of 1 for the ratio between experimental results and theoretical predictions and the lowest COV of all evaluated approaches. This approach has the advantage of being aptly practice-oriented. However, such a dimensioning approach does not explicitly address the governing rupture mode. In load tests, anchorage or delamination failures of the textiles are normally observed which are, both, rather related to loading from shear forces than bending moments. As such, it is recommended to complement the flexural dimensioning by strain limits with a verification of the anchorage capacity of the textiles in the theoretically uncracked zone (in analogy to SIA 166). This report also presents a dimensioning model for this anchorage capacity that shows good agreement with experimental results and acceptable COV. Furthermore, the model also proofs that the anchorage capacity is limited and that the tensile strength of the textiles can usually not be exploited. For a practical dimensioning, design values of the stain limits are required, being based on characteristic values. Practically applicable strain limits for the latter can be recommended for carbon and PBO fiber textiles only, as too few experimental data is available for textiles made of glass and basalt fibers. Textile reinforced concrete represents a viable method for strengthening existing reinforced concrete elements in bending. The main advantages of this strengthening method are the ease of application which involves tools that are already in use on the construction site, and better fire resistance compared to adhesively bonded solutions with fiber-reinforced polymers. The potential impact of flexural strengthening with TRC on shear or fatigue capacity as well as the serviceability behavior of deck slabs could not be assessed due to missing literature data. Further investigations are required on the influence of bond between textile and matrix on the strengthened length, and to determine the effects of flexural strengthening with TRC on the shear strength of bridge deck slabs. A potential strengthening capacity loss over time due to fatigue loading is a further important subject to be studied. Such investigations should target real-scale tests with basalt and PBO strengthening and, in addition, more general structural conditions (higher steel reinforcement ratios, more variable slab slenderness etc.). Particular attention should be given to the identification of the efficiency of three and more textile layers, to be identified in anchorage tests with variable anchorage length and real-scale flexural tests. In combination with additional theoretical evaluations on the influence of the textile bond behavior over the whole strengthened length, such results would allow to refine the proposed strain limits and to identify governing control sections in more detail. The textile bond model presented here can serve as a basis for all these investigations. Last but not least, the results and conclusions of such evaluations should be reflected in normative guidelines (e.g. through a revision of SIA 166). This will further require the derivation and identification of characteristic values, conversion coefficients, and partial safety factors for the strengthening materials.