This paper aims to propose a novel approach to assess the multi-crack behavior of layered fiber-polymer composites. The Compliance and R-curves generated from this novel approach were useful to understand the multiple delamination process, enabling to evaluate separately the strain energy release rate (SERR) related to each crack. A cohesive zone model was developed to simulate the failure process zone of three parallel cracks in web-flange junction (WFJ) specimens extracted from a pultruded bridge deck system subjected to transverse bending. The fracture parameters estimated based on the proposed approach led to a good agreement between the numerical model and the experiments in terms of load vs. displacement curves. Moreover, it was possible to observe that the formation of new cracks may lead to a significant drop on the R-curve, due to the closure of the former cracks.

@en

This paper aims to propose a novel approach to assess the multi-crack behavior of layered fiber-polymer composites. The Compliance and R-curves generated from this novel approach were useful to understand the multiple delamination process, enabling to evaluate separately the strain energy release rate (SERR) related to each crack. A cohesive zone model was developed to simulate the failure process zone of three parallel cracks in web-flange junction (WFJ) specimens extracted from a pultruded bridge deck system subjected to transverse bending. The fracture parameters estimated based on the proposed approach led to a good agreement between the numerical model and the experiments in terms of load vs. displacement curves. Moreover, it was possible to observe that the formation of new cracks may lead to a significant drop on the R-curve, due to the closure of the former cracks.

@en

This paper aims to provide interlaminar fracture mechanics data for pultruded glass fiber-polymer materials, extracted from a composite bridge deck system. Both Modes I and II were investigated, through the adaptation of classical fracture mechanics tests to pultruded specimens. In all, nine data reduction methods were used to determine the Strain Energy Release Rate (SERR) in both Modes and, apart from the Compliance Calibration and Simple Beam Theory methods recommended by current standards, exhibited good agreement between themselves.@en

This paper aims to propose a novel approach to assess the multi-crack behavior of layered fiber-polymer composites. The Compliance and R-curves generated from this novel approach were useful to understand the multiple delamination process, enabling to evaluate separately the strain energy release rate (SERR) related to each crack. A cohesive zone model was developed to simulate the failure process zone of three parallel cracks in web-flange junction (WFJ) specimens extracted from a pultruded bridge deck system subjected to transverse bending. The fracture parameters estimated based on the proposed approach led to a good agreement between the numerical model and the experiments in terms of load vs. displacement curves. Moreover, it was possible to observe that the formation of new cracks may lead to a significant drop on the R-curve, due to the closure of the former cracks.@en

Many recent works have shown that the capacity of pultruded glass fiber-polymer structural members is governed by interlaminar failure. However, there is still a lack of information regarding the fracture properties associated and the best techniques to be adopted. This paper aims to propose a testing methodology and to evaluate the interlaminar fracture of pultruded glass fiber-polymer specimens extracted from a composite bridge deck system. Both Modes I and II were investigated, through Double Cantilever Beam (DCB) and End-Loaded-Split (ELS) tests, respectively. The starter crack (or initial separation) was introduced via a water jet machine and the crack length was measured with the video extensometer technique. In all, nine different methods were applied to obtain the fracture toughness properties. The results are discussed and compared with 2D non-linear finite element models, where the cohesive zone model (CZM) approach was used. The Modified Beam Theory method (MBT_ASTM) presented the best results for crack propagation in Mode I, whereas differences lower than 1% from experiments were obtained when using the Corrected beam theory using effective crack length (CBTE) and the Experimental compliance method (EMC) methods for Mode II.

@en

This paper aims to provide interlaminar fracture mechanics data for pultruded glass fiber-polymer materials, extracted from a composite bridge deck system. Both Modes I and II were investigated, through the adaptation of classical fracture mechanics tests to pultruded specimens. In all, nine data reduction methods were used to determine the Strain Energy Release Rate (SERR) in both Modes and, apart from the Compliance Calibration and Simple Beam Theory methods recommended by current standards, exhibited good agreement between themselves.@en

This paper aims to present an experimental investigation on the behavior of web-flange junctions (WFJs) rotational stiffness of pultruded fiber-reinforced polymer composites (FRP). Channels and I-sections were tested using a simple set-up, which was developed in order to experimentally characterize the junctions in a direct manner. The Digital Image Correlation (DIC) technique was used, allowing overall deflections and relative rotations between web and flange to be monitored. The WFJs' imperfections were analyzed through an optical microscope and correlated with the cracks' formation. Further, damage thresholds are identified using available stress equations for curved composite members and lower bound functions are proposed to simulate the junction stiffness retention. Finally, two Equations are developed in order to analytically predict pultruded junctions' rotational stiffness per unit of width. In general, the theoretical and experimental results agreed fairly well, with a maximum difference of 24% for I-sections and 38% for channels.

@en

This paper aims to present an experimental investigation on the behavior of web-flange junctions (WFJs) rotational stiffness of pultruded fiber-reinforced polymer composites (FRP). Channels and I-sections were tested using a simple set-up, which was developed in order to experimentally characterize the junctions in a direct manner. The Digital Image Correlation (DIC) technique was used, allowing overall deflections and relative rotations between web and flange to be monitored. The WFJs’ imperfections were analyzed through an optical microscope and correlated with the cracks’ formation. Further, damage thresholds are identified using available stress equations for curved composite members and lower bound functions are proposed to simulate the junction stiffness retention. Finally, two Equations are developed in order to analytically predict pultruded junctions’ rotational stiffness per unit of width. In general, the theoretical and experimental results agreed fairly well, with a maximum difference of 24% for I-sections and 38% for channels.@en

This paper aims to investigate the performance of pultruded glass fiber reinforced polymer (pGFRP) I-section columns subject to short-term concentric compression, bringing up a discussion on the relevant parameters affecting their local buckling behavior and the interpretation of tests results. An experimental program was carried out, including a detailed material characterization and twenty-nine compression tests on short I-columns made of either polyester or vinyl ester matrices, with variable flange width-to-section depth ratios (bf/d), column lengths and local end conditions. The theoretical critical loads predicted using generalized beam theory (GBT) and finite element method (FEM) were compared to experimental results obtained by Southwell and Koiter techniques, with better agreements obtained for the latter. It is shown that, besides the length and local end conditions for loaded edges, post-buckling with associated non-linear elastic strains distribution throughout the cross-section and damage prior to buckling have relevant influence on the measured critical loads. On the other hand, the influence of the rotational stiffness of web-flange junctions were considered small for the material and cross-sections studied. Finally, results have shown that the usual boundary condition adopted in literature approaches a clamped condition instead of simply-supported one.@en

This paper aims to investigate the performance of pultruded glass fiber reinforced polymer (pGFRP)I-section columns subject to short-term concentric compression, bringing up a discussion on the relevant parameters affecting their local buckling behavior and the interpretation of tests results. An experimental program was carried out, including a detailed material characterization and twenty-nine compression tests on short I-columns made of either polyester or vinyl ester matrices, with variable flange width-to-section depth ratios (b_f/d), column lengths and local end conditions. The theoretical critical loads predicted using generalized beam theory (GBT)and finite element method (FEM)were compared to experimental results obtained by Southwell and Koiter techniques, with better agreements obtained for the latter. It is shown that, besides the length and local end conditions for loaded edges, post-buckling with associated non-linear elastic strains distribution throughout the cross-section and damage prior to buckling have relevant influence on the measured critical loads. On the other hand, the influence of the rotational stiffness of web-flange junctions were considered small for the material and cross-sections studied. Finally, results have shown that the usual boundary condition adopted in literature approaches a clamped condition instead of simply-supported one.

@en

In this paper, a discussion on the local buckling (LB) behaviour of pultruded glass-fibre reinforced polymer (pGFRP) columns and on the appropriate techniques for testing and analysing data is presented. Comments on the influence of relevant parameters such as length, end-conditions, material properties, flange-to-web width ratio and post-buckling behaviour in the experimentally determined critical loads and theoretical predictions are made. Signature curves (critical stress x length plots) and global x local buckling slenderness ‘maps’ are presented as important ‘tools’ for predicting column behaviour. To support the discussion, the results of an ongoing experimental program involving eighteen stub columns with clamped-ends and different flange-to-web width ratios are reported. Comparison with theory is made and differences between tests and predictions are explained.@en

This paper aims to present the results of an experimental program on the local buckling behavior of GFRP I-section column. Characterization tests were made to determine the material properties and twenty-eight stub columns having three different flange-to-web ratios, overall lengths and end-condition were tested under compression. With the aid of the software GBTUL, based on the Generalized Beam Theory, the so-called signature curves (critical stress versus length) were obtained and lengths were determined to ensure ‘pure’ local buckling. Experimental data are reported and a discussion on the observed behavior is made, including strength, pre-buckling behavior, post-buckling reserve of strength and the influence of length and end-conditions in the critical load. It is also shown that the use of traditional Southwell plot technique leads to greater critical loads.@en