I would like to have an extra opinion about the way of handling the procedure of modeling-designing foundation of big span industrial one storey steel buildings.
In such buildings with (usually typical and repetitive) frames with big span (>15m.) there is no use of assigning connective ground beams between opposite spread footings so the later ones have to carry the large moments of frames “by themselves”.
In order not to have huge footings (the large dimensions of which are mainly caused by the small values of vertical loads hence the verification for rotation and even worst the verification of eccentricity in the location of the resultant vertical force, becomes unfavorable) we have to model realistically the foundation by assigning proper strings to the supporting nodes. [As concerns a proper and practical way of modeling-designing spread foundation, I have already posted my proposal as a comment to the wish list in the forum (as well as I have already addressed it directly to the support team sometime in the past)].
Besides, we may assign continuous footing simultaneously with spread foundation in the cross direction (not in the “effective” direction into the plane of the big span frames but in the across direction, that is anyway required either for practical reasons or for essential reasons) in order to obtain some kind of “relief” at the in-plane moments. For this reason we must model elastic ground along the ground beam (-continuous footing) in both Kz and Hx directions and Hx is the direction that would contribute to the transfer of moments (in-plane to the big span frame). This kind of contribution is relatively small (that is somehow justified) but I am wondering if it is properly handled in Robot.
If someone assigns just only perimetrical continuous footing around the building without assigning spread footings, everything would be “fine” and all kind of verifications (of continuous footings) would be positive. But these results would be “fake” because in the design calculations of continuous footing, Robot doesn’t take in to account the reactions caused by Hx.
Continuous Footing Design module in RSA ignores columns/ base moments about longitudinal axis of the footing and ignores torsional/rotation interactions between footing and soil (resulting in non-uniform soil stress distribution in the direction perpendicular to the axis of continuous footing). It is considering only vertical loads and bending about axis perpendicular to the axis of continuous footing (and torsion but without interaction with soil). Furthermore, it is not possible to directly display reactions resulting from Hx rotation stiffness (such as Ky, Kz Reaction on the “diagrams for bars” option). These two issues have been registered as a SR sometime in the past.
So, in such cases this kind of modeling (without spread footings) isn’t realistic in Robot.
And something more: How really can the option of “moments transferred to the ground beams” in the “optimization” window be properly used !? (This transfer of moment can only be calculated internally in the analysis and not assigned manually in the stage of design..).
Regards.
![Autodesk Design & Make Report](https://forums.autodesk.com/t5/image/serverpage/image-id/1350519iCB55845EBFC43016/image-size/large?v=v2&px=999" "Autodesk Design & Make Report")