PSI - Issue 73

Přemysl Pařenica et al. / Procedia Structural Integrity 73 (2025) 125 – 129 Přemysl Pařenica and Petr Lehner / Structural Integrity Procedia 00 (2025) 000–000

127

3

designed between the purlin and the supporting structure so that all load transfer is provided by the clip only - this prevents local buckling at the point of insertion and ensures uniform load transfer from the purlin stand. The last option (c) combines a reinforcing clip with transverse reinforcement with a single (straight) truss placement. In contrast to conventional fits, the bottom flange of the Z-profile is also bolted to the supporting structure. 2. Numerical models To determine the effect of the flange helical connection on the bearing capacity of the girder, a numerical model of all connections (see Figure 2) was created in Ansys (ANSYS, 2020), which is based on finite element methods (FEM) (Bathe, 2008). The type of tie beam connection by bolting only at the standoff location was not physically tested in the research program. Comparison of the details of the tie beam fits is made by numerical simulations. The aim of the presented simulation was to compare three different geometries in terms of performance. All models created were performed as complex spatial models and included geometric, structural and material non-linearities. A very extensive and detailed description of the models can be found in previous studies (Pařenica et al., 2024, 2021) .The settings and boundary conditions correspond to validated numerical models (Pařenica et al., 2024) . SOLID185, SHELL181, BEAM188, TARGET170, and CONTA175 finite elements were used for the numerical model. The number of nodes was approximately 65,000, and the number of elements was approximately 50,000. The mesh size of the finite elements was approximately 15 mm. The supports were set using remote displacement (see Figure 1, red bases), and the load was applied via the upper crossbar in three separate steps by displacement. The first step was preloading in the bolts, the second step was linear loading, and the third step was plastic behaviour. Due to the tuned boundary conditions and settings, this procedure can be considered as a good estimation of the behaviour of another type of connection of thin-walled purlins and does not serve to obtain exact values of the bearing capacity of the modelled connection. The main motivation for developing this model is to present a comparison of different connection types by means of numerical simulations, whose parameters and calculation settings are partially validated by experimentally obtained data. For comparison, the connection detail for thin-walled Z300 section purlins with a thickness of 1.89 mm and a bearing width of 200 mm with purlin spans of 3.0 m and 5.1 m was chosen. Figure 3 shows an example of a 3.1 m long truss model with a reinforcing clip (Pařenica et al., 2024) . 3. Results As mentioned above, three types of models were developed in terms of different connections and in combination of two spans. In total, six variants were numerically analysed and can be compared in two sets. Figures 3 and 4 show force-displacement diagrams of simulations of three types of connections.

Fig. 3. Graphs of force-displacement diagram obtained from numerical models - span 3.1 m.