Analysis of a lap splice in a lattice girder system

Report, 2003

To enable load-carrying in two directions in lattice girder systems, transverse reinforcement is needed. In the present study, the possibility to put transverse reinforcement in the precast concrete panels and complement with lapped reinforcement across the joints at the construction site was studied. The behaviour of such a joint, when subjected to bending, was investigated in two-dimensional finite element analyses. The analyses show that the cast joint between the precast concrete and the in-situ cast concrete is the weak link in this detailing, as could be expected. In the analyses where a rather large amount of transverse reinforcement was used, ø8 s150 NPs 700, the joint could be loaded close to yielding of the reinforcement; then opening of the cast joint occurred in all of the analyses. When no bent reinforcement, crossing the cast joint, is present, the failure mode will most likely become brittle. It is therefore recommended to have bent reinforcement crossing the cast joint. Two analyses were carried out with two different placement of bent reinforcement, ø8 s150 B500B in both cases. When the bent reinforcement was placed close to the joint, it obtained large stresses rather early in the analysis, and the analysis became unstable for rather low rotations. When the bent reinforcement was placed further away from the joint, the transverse reinforcement reached yielding, and the deformation capacity was approximately doubled compared to the other analyses. For small amounts of transverse reinforcement, or reinforcement with a lower yield strength, it might be possible to use the studied detailing even without bent reinforcement crossing the joint. In one analysis where the transverse reinforcement had a reduced yield strength, 500 MPa instead of 700 MPa, the reinforcement reached yielding, and it was possible to keep the yielding moment some additional rotation before the joint opened up. The modelling of the cast joint is of very large importance for the results of the analyses. The modelling of that was checked through analyses of joints between precast and in-situ concrete tested by Nissen et al. (1986), who made a large experimental investigation on the interaction between precast and in-situ concrete. Still some uncertainty about the input parameters remain. Furthermore, long term effects such as shrinkage and creep were not included in these analyses. It is recommended to do further studies, including full-scale testing of lap splices before this detailing is used in practice.