Model
I modeled two blocks (high stiffness) with the dimensions 5 x 5 x 5 and connected them with the interaction property "cohesive behavior". The blocks are lined up in x-direction. In my first step, I want to separate these two blocks in normal direction (x) followed by a tangential sliding in shear 1 direction (y) in the second step. The separation in x direction depicts a pre-damaging process.
Parameters
Material properties of two blocks:
Elastic
- Young's Modulus of blocks = 1000000
- Poisson's ratio = 0
Interaction properties (cohesive behavior + damage):
Elastic
- Knn = 0
- Kss = 10
- Ktt = 0
Damage
- Initiation: Quadratic separation (delta_n0 = 0.1 / delta_s0 = 1 / delta_t0 = 1000000)
- linear Damage evolution with total/plastic displacement delta_f = 9
Loading:
First step
- u1 = 0.7
Second step
- u3 = 10
Question
I comprehend the calculation of the damage variable in the first step:
D = (delta_f * (delta_n - delta_n0))/(delta_n * (delta_f - delta_n0))
How is the overall damage variable calculated in the second step due to the pre-damaging?
Niklas Müller, the overall damage variable in the second step of your ABAQUS two step model is calculated based on the damage variable from the first step and the additional damage that occurs in the second step. In the first step, the damage variable is calculated using the quadratic separation initiation criterion and the linear damage evolution law that you specified. The quadratic separation initiation criterion states that damage will initiate when the normal separation, shear separation, or tangential separation exceeds a certain threshold value (That you have defined). The linear damage evolution law states that the damage variable will increase linearly with the total or plastic displacement. In the second step, the damage variable is updated based on the new loading conditions. If the damage variable is less than 1.0, then it will continue to increase according to the linear damage evolution law. However, if the damage variable is equal to 1.0, then the material is completely damaged and will not support any further load.
Alternatively, the simulation could be done in one step using amplitude ((0,0),(T/2,1),(T,1)), value 0.7 for U1 and amplitude ((0,0),(T/2,0),(T,1)), value 10 for U3, where T is the step time. Of course, this is not equivalent to the original two-step model, but it would be interesting to compare the results.
Damage in Abaqus is purely empirical. Unless you have damage test data from the system you are analyzing guessing will give you better results.
In the second step of your analysis, you're essentially dealing with a situation where the material has already experienced some damage due to the pre-damaging process in the first step. The damage variable at the end of the first step will influence the behavior of the material in the second step.
To calculate the overall damage variable in the second step, you need to consider the damage accumulated from the first step as well as any additional damage accrued during the second step.
The damage variable at the end of the first step, denoted as D_1, would be calculated using the formula you provided:
- Badges
- Science topic