School of Mechanical Engineering Seminar
Wednesday, December 8, 2020 at 14:00     
Wolfson Building of Mechanical Engineering, Room 206

Ido Simon

Ph.D. student of Prof. Leslie Banks-Sills

Failure Prediction of Bonded Composites with Application to Structures with External Ply-Drops

The motivation of the transportation and aerospace industries to reduce the weight of their manufactured products has greatly encouraged the development and use of advanced lightweight materials such as carbon fiber reinforced polymers. The adoption of such materials along with the desire to further reduce weight has also increased the use of adhesive bonding in load bearing structural elements. Employment of adhesive bonding and carbon fiber reinforced polymers in such structures requires reliable computational tools and experimental methods to assess the load bearing capabilities of the bonded joints.

The aim of this investigation is to develop a methodology and a numerical tool for predicting debond failure in a bonded composite structure. The structure investigated here is a simplified one, which may be thought of as representing a composite laminate with external ply-drops. The laminate is composed of uni-directional carbon fiber reinforced polymer plies which were manufactured by means of a pultrusion process. Such plies are cured separately as part of the pultrusion process. The cured plies are then bonded using a two-part epoxy adhesive which is cured at room temperature. The resulting laminate is denoted as a secondary bonded laminate. In order to achieve the aim of this investigation, both experimental and numerical work is carried out.

The experimental work may be divided into two parts. In the first part, the modes I, II and mixed modes I/II fracture toughnesses of the bonded composite are determined and a mixed mode I/II initiation fracture toughness criterion is obtained. In the second part of the experimental work, a secondary bonded pultrusion structural element specimen containing a ply-drop is tested under different types of quasi-static loadings. The ply-drop specimen tests are carried out to examine and assess the different failure modes of this type of structural element. In addition, the results of the ply-drop specimen tests are used in the calibration and validation of the numerical model.

The numerical model is based on two fracture mechanics concepts: the cohesive zone model and the virtual crack closure technique. It is implemented in a hybrid fracture discrete finite element. A fully coupled mixed mode traction separation relation is developed. The hybrid fracture discrete element is coded using the Abaqus® finite element analysis user element (UEL) subroutine. The UEL is equipped with run and arrest fracture criteria in order to model unstable debonding failure. Following the development of the numerical tool, it is calibrated using the results of the fracture toughness and some of the ply-drop specimen tests. Then, a validation stage is performed. In the first part of the validation, the ply-drop specimen tests which were not used in the calibration are considered. In the second part of the validation, a blind test, which consists of a quasi-static test on a specimen with a similar geometry as the ply-drop specimen but under a different type of loading, is carried out. The numerical tool is used to predict damage initiation and delamination propagation of the blind test and the numerical results are compared to the experimental ones. The results are well correlated.

Join Zoom Meetin https://us02web.zoom.us/j/82108132163?pwd=Z2h4UzNzUS9mbXplT0lMU1pZenFEQT09