Learning Forum

[Baloch]

Dear All, I am doing a time history analysis of an RCC structure for which I am getting some differences between the experimental results and my simulation, which might be due to the materials properties because I am using linear properties for concrete. In Order to better represent the behaviour of concrete I am trying to use the Menetrey- Willam and Drucker Preger model base on the given data (uniaxial compression and young modulus) But when ever I use these two models I face convergence issue which I noticed is at the point where most probably yielding occurs. I derived parameters of both these models based on literature and comparing these with ansys requirements which are fine but still I am having the issue. Additionally, I tried with only Menetrey and Drucker base without any softening behavior still the same problem I am facing. The input of both these models are shown in snap attached. Anyone have idea or experience please share.

[John Doyle]

Your tensile strength (=2.21Mpa) looks reasonable for concrete, but it would not take much to cause convergence trouble if you are exceeding this value. Is any part of your fea model in tension? If this is the case, one possible solution might be to switch to semiiplicit solver. Please refer to the MAPDL Commands Manual for more details on the SEMIIMPLICIT command. Since this is a transient analysis, you could also try adding more system damping to help stabilize the solution. You could try to reproduce the nonconvergence with a simple block model under a force based load and test which of these options best resolves the nonconvergence.