ABAQUS粘接单元应用.ppt

Over the next several sides we will compare surface-based and element-based cohesive behavior. We will begin by discussing the differences in preprocessing, including mesh refinement considerations. Surface-based cohesive behavior is easier to model because cohesive elements do not need to be defined. For surface-based cohesive behavior the cohesive constraint is enforced at each slave node; hence, refining the slave surface relative to the master surface will likely lead to improved constraint satisfaction and more accurate results. For element-based cohesive behavior the cohesive constraints are calculated at the material integration points which are located on the edges between nodes in the element thickness direction. In this case, refining the cohesive element mesh relative to that of the connected structures (while still maintaining one element in the thickness direction) will likely lead to improved constraint satisfaction and more accurate results. “Sticky” contact must be modeled using cohesive contact (i.e., surfaces or parts of surfaces that are not initially in contact may bond on coming into contact; subsequently the bond may damage and fail). Debonded surfaces can either stick or not stick if contact occurs again. For example, you can use cohesive contact (but not cohesive elements) to model Velcro, which will re-stick each time contact is established. Both cohesive contact and cohesive elements can model situations where the surfaces will not re-bond once the cohesive interface has failed (e.g., dried glue). Surface-based cohesive connections only allow for one failure mechanism, while element-based cohesive connections allow for multiple failure mechanisms. Interface thickness effects are never considered for cohesive surfaces; in cohesive elements with traction-separation response, thickness effects can be incorporated by either specifying a nonzero thickness for the interface or by requiring the initial constitutive thickness to be determined