Cohesive element vs Traction - separation behavior (interaction /contact proprety)

[AleksaBgd94]

Hello everyone,

I am trying to model the masonry infill by simplified micro modelling approach. I defined one model in which the interaction between the bricks and between bricks and frame is defined by assigning additional contact properties (cohesive behavior and damage) to the interaction. However, I would like to run calculations in which the mortar is defined by cohesive elements. I read that the two approaches yield similar results, but I would like to check it.

I have some questions regarding the input parameters. If the one uses the interaction property to define the elastic behavior of the mortar, the values Knn, Kss and Ktt are entered as input data. In the case of cohesive element, the one should define Enn, Ess, Ett (figures attached). Are these values actually the same since the actual length is L=1 and it is ensured that ε = δ ?

Furthermore, during the definition of the model, should this cohesive element be modelled as a part with actual thickness? In the papers it is usually called zero-thickness element, but should the one model it as a part with very small thickness(for example 1 mm) and then assign the cohesive element section to the part etc.?

Thanks in advance.

 

[FEA way]

Element-based cohesive behavior is more accurate and allows you to account for the thickness of adhesive layer but it’s also more difficult to use and solves longer.

Cohesive elements must be embedded in a model. You can do it using tie constraints. Create a finite thickness cohesive layer and position it properly with respect to the rest of the model, define surfaces for tie constraints, mesh the parts and create tie constraint using surfaces created before.

Check Abaqus documentation and articles, for example the one titled "Determination of Failure Modes in Brick Walls Using Cohesive Elements Approach" by L. Kowalewski and M. Gajewski.