Issue 13

With limit state design, sidesway of buildings may become more critical than collapse; there is considerable difficulty in specifying ‘safe’ sidesway, since wind gusts are dynamic and opposed by inertia as well as by frame stiffness. Dynamic analysis necessitates using large computers but many designers prefer simplified ‘desk’ methods of analysis. Two such methods for static behaviour have already been developed at the Building Research Station, employing substitute frames, and have been accepted in recent British and European Codes of Practice. This paper describes an analogous method for dynamic behaviour of multistorey frames, in particular the natural frequencies and modal shapes that could be developed to study the effects of forced vibration. The largest matrices involved are now only of size 2 x 2 irrespective of the number of members in the multistorey frame. It is shown how composite stiffening effects of walls and floors can be introduced into the design, which highlights the need for more tests in real buildings. E.H. Roberts and R.H. Wood

Results are given of bending tests in the elastic range on seven composite beams with the concrete slab in tension. They show that, when cracks in the concrete first occur, they are wider than would be predicted by the existing theory for cracking in reinforced concrete members. The reason is that cracking releases the locked-in stresses due to the restraint of shrinkage of the slab by the steel beam. When account is taken of this, the accuracy of the methods of BS 5400: Part 5 for the prediction of hogging curvature and mean tensile strain at a given bending moment, is much improved. It is shown in another paper that these results also enable the widths and spacings of cracks in the concrete slab to be predicted more accurately. R.P. Johnson and R.W. Allison