Calculation methods of strengthening of existing structures

Abstract

This thesis is a theoretical evaluation on whether the sequence of installation can be neglected when determining the capacity of structural systems, as postulated by the elasto-plastic lower bound theorem. The thesis further aims to identify the conditions under which this approach is considered acceptable. To address this approach, two theoretical studies were conducted:

-Sequential buttweld capacity -Capacity of sequential installation of columns

For the theoretical study of sequential buttweld capacity, a hypothetical situation was examined, where a sequence of buttwelds is added to a structure. Each buttweld is then loaded to its design capacity prior to the introducing the next buttweld. The aim of this study was to capture the influence of the installation sequence of sequential buttwelds. Installation of two and ten sequential buttwelds were examined. Finite Element Analysis (FEA) was conducted in Abaqus and simple hand calculations were conducted in Mathcad. The results indicates that the initial weld experienced a higher level of strain for the ten-step sequence compared to the two-step sequence. This further indicates that the strain increases with number of sequences and will endure some strain within the non-elastic domain. However, the results also indicated that the anticipated strain level in the initial weld is well below the threshold of rapture failure. For a ductile material like steel, the system demonstrated sufficient capacity even after the sequence of installation.

The second theoretical study was a study into buckling of columns and evaluation of capacity of sequential installation of columns. For this study the capacity, ductility and slenderness of undamaged tubular members with circular hollow section (CHS) were investigated. Further, these were used in an evaluation of capacity of two sequentially installed columns. No experimental test was executed in this thesis, hence, the study relied on previous experimental test reports executed by Chen and Ross (Chen & Ross, 1978), Vo and Hestholm (Vo & Hestholm, 2019) and Riise (Riise, 2021). From the experimental studies, a total of 10 undamaged columns with varying geometric and material properties were utilized. Capacity of each column was calculated according to NORSOK N-004 and compared to experimental results. A proposal for calculation of the level of ductility is suggested, through a ductility index. In addition, an analysis of how this ductility index varies with slenderness and Diameter-to-thickness (D/t) ratio for experimentally tested columns is presented. Three columns were selected from the experimental test set (specimen 6, Specimen 8 and Riise1) where nonlinear buckling analysis was performed in Abaqus. The FE-model were then compared to experimental test results. FE-model and experimental test showed reasonable comparison, but full agreement in the post-buckling phase was and is a challenge. A simple evaluation of the ultimate system capacity of two sequentially installed columns were simulated in Mathcad. From the calculations, an indication of the required ductility for the individual component to allow for ignoring the installation sequence was found, as a function of slenderness and diameter to thickness ratios.