THE STABILITY OF BUILT-UP AXIAL LOADED COLUMN IN LIGHT OF STR AND EC3

Abstract

Straipsnyje pateikta plieninių spragotinio skerspjūvio kolonų laikomųjų galių, apskaičiuotų vadovaujantis Lietuvoje galiojančių plieninių konstrukcijų projektavimo normų STR 2.05.08:2005 ir Eurokodo 3 nuostatomis, lyginamoji analizė. Skaičiavimai buvo atliekami vienodomis pradinėmis sąlygomis, tik naudoti skirtingi skaičiavimo metodai. Kai kuriais atvejais gautieji rezultatai yra labai prieštaringi ir reikalingi išsamesnės analizės ar eksperimentinių tyrimų. The paper presents the analysis of built-up laced axially loaded steel columns in light of Eurocode 3 and Lithuanian design code STR 2.05.08:2005. The theoretical part analyzes two design methods. Some cases indicate principal differences. According to STR, axial forces are equally divided into two parts for both chords. However, in Eurocode 3, axial force (formula 8) for one chord increases due to the additional bending moment (Formula 6) that depends on the shear stiffness of lacings (Formula 5). For very slender columns, the axial force of one chord, considering Eurocode 3, is 2.7 times bigger than that taking into account the STR method. Another big difference between the methods is that according to Eurocode 3 it is not necessary to check the overall stability of the built-up member round the z-z axis (only checking the stability of one chord round the z1-z1 axis is obligatory). Both methods require checking the stability of one chord round the y-y axis. In two cases, calculations referred to the same initial data (Table 1, 2) applying different design codes. The obtained results are presented in the diagrams. The first case shows that column slenderness in both planes equals λy = λz. The axially loaded column calculated with reference to the STR method has bigger bearing capacity reserve than that calculated considering the Eurocode 3 method. In this case, the stability of one chord round the y-y axis (Fig. 3) is the most dangerous. This example illustrates that the stability condition of the axially loaded column according to Eurocode 3 is not satisfied; thus, a necessity of increasing the column cross-section arises. The main reason for the latter situation is a different method used for calculating the axial force of one chord. This difference is greater for more slender columns. In the second case - column slenderness makes λy = λz/2. When slenderness is λz ≤ 100, the axially loaded column calculated according to the STR method has similar results compared to the Eurocode 3 method (Fig. 10). The most dangerous according to STR is the stability of the entire column round the z-z axis (Fig. 8), whereas in accordance with Eurocode 3 it appears to be the stability of one chord round the y-y axis (Fig. 9). In such a case, the stability condition of the axially loaded column according to Eurocode 3 has more reserve only when slenderness is λz > 100 (Fig. 10). Therefore, calculation according to Eurocode 3 is less safe if compared to the STR method. The main reason is that Eurocode 3 does not require checking the entire column stability round the z-z axis. Hence, for calculating slender columns according to Eurocode 3, some cases (λz > 100) are not very safe, which was also noticed in the numerical investigations provided by other authors Kalochairetis (2011). In some cases, results are controversial, and therefore it is necessary to perform additional analysis or experimental investigation.