Cohesive modeling is one of the unique methods which has been used to model adhesive bonding in computer aided engineering (CAE) industry. There exist numerous conventional methodologies which involve the usage of hexa and penta elements by assigning cohesive material properties. These methods inherently are error-prone in terms of modeling errors and result in increased modeling and computation times. A conventional method of cohesive zone modeling (CZM) has a drawback of higher computation and modeling time. Due to this problem, sometimes engineers tend to avoid simulations and rely only on some sort of approximation of crack from previous designs. This approximation can lead to either product failure or overdesign of the product.
A new modeling technique is discussed in this paper to simulate crack initiation and propagation with cohesive zone modeling, The new approach models the cohesive zone as contact between two bodies, thus eliminating the need to use cohesive elements which eliminates the modeling errors and drastically reduces the modeling and computation times.
The approach uses the damage model method which allows modeling of interaction between two surfaces based on available experimental data. This will yield output in terms of damage initiation, damage evolution, cohesive opening, and its status. The validation of model is done with the double cantilever beam (DCB) test. The specimen is made with two panels and flaw is introduced at their interaction. The strain life approach is used to calculate the fatigue damage under testing cycles. The result is compared with the experimental data for accuracy. This paper is focused on the aspects of a new modeling technique to analyze adhesive bonds using finite element methods by Altair OptiStruct and its comparison with a conventional method of cohesive zone modeling. Also, Altair's Design Explorer is used for exploratory studies with automated machine learning (AutoML) to predict the results for nonsimulated runs.