Mode I cohesive zone model parameters identification and comparison of measurement techniques based on uncertainty estimation

Abstract

Adhesive bondline mechanical behaviour is frequently described with cohesive zone models (CZM). For mode I loading condition these phenomenological laws simply represent the evolution of the peel stress as a function of the two adherends relative displacement normal to the joint. Generally, these laws are identified rather than really measured using experimental data obtained from crack initiation and propagation experiments such as the Double Cantilever Beam Test (DCB). The uncertainty on parameter estimation are generally not indicated, as for a DCB test it is only the critical energy release rate that has the most influence on the results. However, the uncertainties on the other parameters prevent the use of the identified TSL for other mechanical tests where mode I solicitations are predominant. In this article, the purpose is to evaluate the methodologies reliability for the assessment of mode I CZM. To do so, several methods used to evaluate CZM parameters are compared in terms parameter estimation reliability. Synthetic noisy data are considered for a χ² function minimisation. Then, sensitivity calculations are performed to determine the estimated parameters standard deviation. By applying this procedure on different type of synthetic measurements (respectively P(), J(,), backface strain and DIC) the ability of these different techniques to capture the best parameters for a chosen CZM shape can be rigorously evaluated.

Estimating crack tip position in adhesively bonded joints subjected to mode II quasi‐static loading

“…For the simulation of the crack propagation within the adhesive layer, the cohesive zone model (CZM) 25 was used considering both a bi-linear and a trapezoidal TSL, as depicted in Figure 3. While a bi-linear TSL is commonly associated to brittle adhesives and the a trapezoidal one to ductile ones, in some researches, 25,34,35 the bi-linear TSL was successfully employed to model ductile adhesives. To explore this further, both triangular and trapezoidal models were used for the ductile adhesive considered in this study.…”

Bonded flush repairs for aerospace composite structures: A review on modelling strategies and application to repairs optimization, reliability and durability

“…However, there is no standard procedure for parameters identification. It can be directly based on measurements [99][100][101][102][103], or indirectly identified by calibrating a FE model to match to experimental tests. The same is true for the maximum stress parameter 𝜎 0 , which is often extracted from other tests like thick-adherend shear test (TAST) or butt joint but could be also extracted from crack propagation tests.…”