"Step" command in FLAC

FLAC
1

Hi to all,

I’m performing a slope stability analysis in FLAC 8.0 in which I’d like to replicate the dynamic action related to a water wave impact on the slope. The idea is to use the function apply sxx along the (i,j)-cells interested by the wave impact, by applying different values of stress (i.e., sxx) per each time-step of analysis.
I realized that it doesn’t make sense to reach the equilibrium using the solve command. In fact, my idea is to replicate the dynamic action of water waves impact on a slope by the use of a pushover analysis. This latter consists in reproduce the multiple and subsequent applied pressures on the slope, that, in “real-life” are related to a dynamic phenomena (which lasts approximately 30 seconds), but in the pushover analysis (i.e., FLAC model) I replicate these actions referring to separate (but subsequent) static calculation steps. However, it doesn’t make sense to reach the equilibrium, because it would mean considering the applied pressures not as instantaneous but as applied pressures acting over time.
So, I’m trying to use the step command instead that the solve one, but now the crucial question is: how many steps should I consider for my purposes?
I’m not performing a creep analysis, so the steps are not directly referred to the real time.

2

why not use dynamic option and create a history of applied sxx (sxx vs time) and then apply it along the boundary and solve dynamic time to 30 seconds?

3

Because I want to apply the pressure only along a specif boundary and not to the whole domain. The boundary would be represented by a portion of the slope on which the waves impact.
Can the dynamic analysis perform also in this kind of setting?

4

And also because the pressure distribution is not uniform along the slope surface that I want to stress.

5

Yes, you can do that! please note: 1) the sxx should be applied at least on two gridpoints, and 2) since the pressure distribution is not uniform along the slope, you need to apply the sxx history in different segments:
apply sxx xx history xx i=i1, i2 j=j1,j2
apply sxx xx history xx i=i3, i4 j=j3,j4
xxxx

6

Thank you so much jwang! I will take into account your precious suggestion.
Do you think the the idea of the pushover analysis and the timestep solution does not make any sense at all? I mean, if I would try to perform and compare both the methods (dynamic and pseudo-static), in your opinion how could I manage the timesteps for the pushover analysis?