Viscoelastic model for analysing the behaviour of adhesive-bonded FRP-to-steel joints in civil engineering applications

Modelling nonlinear shear creep behaviour of a structural adhesive using deep neural networks (DNN)

“…Several studies have utilised the accelerated testing method, which involves dynamic mechanical analysis (DMA) testing at high temperatures and the application of the timetemperature superposition principle (TTSP), to directly characterise the viscoelasticity of adhesives, which is believed to be the root cause of adhesive creep [10,13,[15][16][17][18]. For example, Houhou et al [13] employed the accelerated test method to investigate the impact of adhesive nonlinear creep on FRP-to-concrete bonded assemblies.…”

“…However, a previous study by the authors [10] has stated that this accelerated testing method may have some limitations, such as high-temperature decomposition effect of the material, errors in the application of TTSP, and further curing of the examined samples. In addition, it may be challenging to obtain a close fit to conventional nonlinear viscoelastic constitutive models (e.g., the modified Burgers model and the parallel rheological framework model, and may further lead to additional errors [10,14]. Therefore, the (non-accelerated) experimental investigation and accurate modelling of nonlinear creep of structural adhesives needs to be further developed.…”

“…Such occurrences can negatively impact the performance of strengthened structures, diminish design efficiency, and limit the widespread adoption of this advanced strengthening technique. Consequently, there is a pressing need for an in-depth understanding of how temperature, stress, and time variables affect the nonlinear creep growth of structural adhesives [6][7][8][9][10].…”