Afternoon All,
We am working on a Robot model for a temporary, cantilever protection deck for a job we are working on involving an old, concrete-encased steel frame with non-load bearing, cast in-situ infill walls. This involves installing penetrations through the non-load bearing RC walls, fixing temporary steel needle beams to the sides which cantilever out of the building with the backspan supported by temporary steel framing (see attached screenshots). The catch deck has to be designed for a 10 kPa UDL live load in addition to scaffold leg loads.
I have used linear releases along all vertical panel edges and along all bottom horizontal panel edges to prevent load sharing between the concrete-encased steel columns and edge beams and modelled the penetrations we need to made to accommodate the cantilever needle beams using the opening tool.
The results of the analysis show that the cantilever needle beams seem to be taking up substantially more load than expected. After further investigation, we noticed that the axial force diagrams for the steel columns which are embedded within the concrete wall show an anomaly at the cantilever needle beam level whereby the column appears to go into tension above the needle beam and back into compression below needle beam level. The axial load diagram for the columns above needle beam level increment with each floor as one would expect and at the floor below assume a more typical axial load diagram. Our suspicion is that Robot thinks that because the needle beam runs through the wall (albeit through an opening) it assumes that the needle beam and RC wall are connected and therefore we are getting load from the column getting taken up by the needle beam, transferred back into the wall, taken down to the edge beam below and then transferred back to the column at the level below hence why it appears to return to 'normal'.
Does this sound like a plausible explanation and if so, does anyone have any good suggestion as to how to stop this from happening?
Anthony
