Good morning,
I am seeing some quite different results in 2 otherwise similar models - where the only difference is in the propped capping beam is modelled as a plate in one model, and as beams in another.
In context: The model is a bored piled and propped cofferdam into a slope. At one end of the retaining structure (out of the slope) the retained height reduces, and so the bored pile wall can work in cantilever.
In the 'Beam model' we are seeing some significant moments / torsion on the capping beam at this interface between propped and cantilever wall, particularly where it is connected to a vertical section of capping beam,to broadly follow the change in retained levels.
At this interface, Bending moments in the vertical beam, (due to the relative back rotation profile at the top of the propped retaining wall, against the relative forward rotation at the cantilever section) are high - this is moment connected to the horizontal capping beam and the vertical moments are transferred to the horizontal beam as torsion.
In the 'equivalent plate' model we see a 75% reduction in this torsion, and I wonder if there is something lost in the transfer of moment to torsion when using these as equivalent beam elements. Intuition tells me that the beam model is the most correct way in which to model the capping beams, and has always been my preferred modelling option.
Is there a 'best practice' when modelling a capping beam that may explain the differences in the magnitudes of these bending moments and torsion?
Regards
Scot