Welcome to this issue of Structures and Buildings. The papers this month cover a wide range of topics, from detailed consideration of joints in both steel and precast concrete, to the seismic response of braced buildings and the dynamic response of reinforced-concrete beams under impact, to steel plate girders with tubular compression flanges and the shear behaviour of fibre-reinforced-concrete beams. All the papers relate to practical problems faced by practitioners and offer new methods of analysis, design or implementation.A neural network model has been developed by Doran et al. (2017) to study the seismic response of braced steel-framed buildings. The use of steel concentrically braced frames is restricted to low-to-moderate seismic regions and low-rise buildings. Recent studies of the response of such frames under moderate earthquakes has shown them to perform better than expected and this paper sets out to investigate further using two example buildings, one of three stories and a second much taller one of nine stories. As step-by-step time-history analyses are time consuming to conduct, the paper explores how the use of a neural network model might capture the behaviour of the frames under seismic action. The development of the neural network is described in detail before conducting sensitivity analyses. Valuable insights into the dynamic response of concentrically braced frames to seismic loads are provided alongside evidence to suggest that neural network models could form the basis of practical design tools.When using precast concrete it is clearly important to tie together the individual components to achieve a robust structure. For precast wall elements, this is commonly done using U bars, through which a lock bar is inserted and grouted in place. As installation of the wall panels is impeded by the protruding U bars, alternative solutions using high-strength wire ropes have been developed. Joergensen et al. (2017) have conducted a series of tests on precast concrete shear joints incorporating high-strength wire ropes and developed a design model to predict the failure load. An accurate design model is important because wire ropes, although very strong, do not yield and therefore may result in undesirable brittle failure of the joint. It is therefore essential that the wire rope is the strongest part of the joint and ensure that failure occurs in a ductile manner by yielding of the lock bar and crushing of the joint mortar. The development of the design model is described in detail and compared with eleven tests recently conducted by the authors and twenty-nine reported in the literature.Conventional steel plate girders are, as the name implies, fabricated from flat plate. Replacing the compression flange with a tubular section may improve material efficiency by improving local capacity and resistance to lateral torsional buckling. Kharoob (2017) presents the results of a detailed finite-element investigation of the effect of tubular section shape (i.e. radius of gyration and section modulus) on the...