Please use this identifier to cite or link to this item: https://idr.nitk.ac.in/jspui/handle/123456789/14482
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorBalu, A. S.-
dc.contributor.authorB, Kesava Rao-
dc.date.accessioned2020-08-28T06:29:16Z-
dc.date.available2020-08-28T06:29:16Z-
dc.date.issued2019-
dc.identifier.urihttp://idr.nitk.ac.in/jspui/handle/123456789/14482-
dc.description.abstractDelamination between plies is the most common mode of failure in composite laminates, which occurs due to the presence of matrix cracks, free edges and notches. Prediction of such failure poses a challenging task as it overburdens the computational resources. Therefore, cohesive zone model (CZM) was introduced for representing fracture as a material separation across crack surface. In CZM, cracks and other material discontinuities can be represented using zero thickness fracture process zone in a finite element (FE) framework. Hence, it has become one of the most powerful computational models in predicting the crack initiation and propagation in composites. Further to improve the efficiency of the model, metamodel techniques were introduced to capture the delamination strength of composites. Nevertheless, most of the metamodels are inefficient for highly nonlinear problems and sometimes insensitive to the parameters. Therefore, in this work, a novel CZM is developed based on high dimensional model representation (HDMR) to evaluate the fracture behavior of composites under different mode conditions. The proposed methodology involves the development of CZM using HDMR, implementation of traction-separation laws in FE model using a user-defined subroutine in Abaqus, and minimisation of error using optimisation techniques. An attempt has been made to reduce the computational effort in accurately capturing the delamination strength. The proposed model is employed for capturing the steady state energy release rate (ERR) of a double cantilever beam (DCB) under Mode-I loading. The FE models have been created using HDMR-based response functions. Initially, the CZM is developed for predicting the delamination strength of 51 mm crack size DCB specimens, and the model is then used to predict the ERR variations of 76.2 mm crack size specimens. Subsequently, the numerical results of the developed DCB models are verified with the available experimental data for unidirectional composites (IM7/977- 3). Then, the efficiency of the proposed model is demonstrated by comparing the results with second-order nonlinear regression metamodels. Further, the proposed methodology is extended to assess mixed-mode (MM) failure behavior of the adhesive joints. As a part of experimental study, the composite single leg bending (SLB) specimens are manufactured by using unidirectional carbonfiber reinforced material and epoxy resin, and the tests are conducted in TINUS testing machine as per ASTM D790 under the influence of pure mode dominant conditions in order to obtain the cohesive parameters. Optimization techniques are used to minimize the error between the simulation and experimental values. The MM–CZM is then established and implemented in the SLB joint under various mode mixities for analysing the fracture process. Comparison between the numerical and experimental results shows that the proposed HDMR based approach estimates the failure mechanism efficiently.en_US
dc.language.isoenen_US
dc.publisherNational Institute of Technology Karnataka, Surathkalen_US
dc.subjectDepartment of Civil Engineeringen_US
dc.subjectCohesive zone modelen_US
dc.subjectFinite element analysisen_US
dc.subjectHigh dimensional model representationen_US
dc.subjectResponse surface modelen_US
dc.subjectDouble cantilever beamen_US
dc.subjectSingle leg bendingen_US
dc.titleCohesive Zone Model for Fracture Characterisation in Compositesen_US
dc.typeThesisen_US
Appears in Collections:1. Ph.D Theses

Files in This Item:
File Description SizeFormat 
148056CV14F14.pdf2.28 MBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.