PSI - Issue 81

Nazar Loboda et al. / Procedia Structural Integrity 81 (2026) 221–227

226

diagram was obtained (Fig. 6). Therefore, analysis of this diagram leads to the conclusion that the cohesive contact model deviates from the experimental results by approximately 5%, which is within acceptable limits. This confirms the validity of the developed cohesive contact model and the correctness of its numerical implementation. After numerical modeling of adhesive joints using two different approaches - cohesive elements and cohesive contact, a comparative analysis of the obtained results was carried out to evaluate their accuracy and practical applicability in engineering calculations. For both methods, load-displacement diagrams were constructed (Fig. 6), illustrating the behavior of the joint up to complete failure. The discrepancy between the curves does not exceed 6%.

Fig. 6. Comparison chart of calculation results.

Both methods use identical or very similar parameters of the adhesive layer material, but their implementation in the software environment has fundamental differences. The approach using cohesive elements involves explicit modeling of the adhesive layer using special elements of small thickness. In contrast, cohesive contact is implemented as a contact interaction between the surfaces of the bodies without the need to construct a separate geometry for the adhesive layer.

3.4. Influence of mesh size

During numerical modeling, an important step is the selection of the optimal finite-element mesh size, since the accuracy and stability of the calculation results largely depend on this parameter. The mesh quality directly affects the representation of the stress-strain state in the stress concentration zones, particularly in the adhesive layer region. An excessively coarse mesh can lead to the loss of local effects, distorting the results and reducing simulation accuracy. At the same time, an overly fine mesh significantly increases the number of elements and, consequently, computational costs, which is not always advisable if the increase in accuracy is not significant.

Fig. 7. Influence of mesh size to: (a) cohesive elements model; (b) cohesive contact model.

In order to determine the optimal finite-element mesh size, a series of calculations with varying degrees of mesh refinement were carried out for both types of numerical models. Figure 7 shows the influence of mesh size on the calculation results for the model with cohesive elements (Fig. 7.a) and the model with cohesive contact (Fig. 7.b). Analysis of the results showed that reducing the elements size below 0.5 mm does not significantly improve the agreement with the experimental load-displacement curve, but leads to a significant increase in computational costs. Based on this analysis, it was determined that the optimal element size is 0.5 mm for both the model with cohesive elements and the model with cohesive contact. This choice provides a balanced relationship between the accuracy of strength prediction, the reproduction of the adhesive joint failure behavior, and computational efficiency, allowing the developed numerical models to be considered sufficiently accurate and suitable for further engineering applications.