Riveted joint's static strength experiments and numerical studies are carried out on four different riveted lap joints having three rivets (single row), five rivets (single row), nine rivets (three rows and three columns) and twenty five rivets (five rows and five columns) along with strain gauges placed at specified locations. From the experiments, it is observed that all the riveted lap joints sustained the estimated joint failure load and failed only after reaching or exceeding the joint failure load. The joint behavior is close to linear, if the joint has large number of rivets. And, if the joint has less number of rivets, the joint behavior is non-linear. Non-linear finite element analysis of riveted lap joints shows that load sharing among the rivets is not uniform as long as rivets do not yield (or in the elastic regime). Once the rivets start yielding, load redistribution takes place until all rivets share the load nearly equal. After rivet yielding and till the final load step or joint failure load, load sharing among the rivets is nearly uniform. The finite element strains, obtained from the consideration of rivet yielding, plate yielding, geometrical non-linear effects and secondary bending effects of the lap joints, are found to be in reasonable agreement with the experimental strains at the failure load noting that our assessment of the rivet behavior is reasonable, even though a simple bi-linear model is used for the rivets. In reality, if plate tearing occurs, the interactions could lead to more complicated behaviors. Further refinement of the bi-linear material model and three dimensional modeling for the rivets could help in refining the numerical model. This was not found necessary for the present work, as we are intent on validating the final load carrying capability from the experiment is the same as in conventional design on one hand and also finite element analysis could make a reasonable prediction of the final load carrying capacity using simple models. The reasonably good agreement between the analysis and experimental strains at the failure load and this in conjunction with the experimental finding that all the riveted lap joints sustained the estimated joint failure load confirms that the rivets share nearly equal load at the joint failure load. This also provides information with regard to the strain picture in the plate around the rivets that are highly loaded. From this observation and in continuation of our discussions, we note that the designs of riveted joints have to be considered from static strength as well as fatigue, where deterioration in residual strength can occur if unequal distribution of rivet loads cause local high stresses and cracking. This needs further investigation and is being carried out. However, in the joints, where more numbers of rivets are used by considering the fitting factor of 1.15, possibilities of initiation of cracks will be lower as local stresses will be lower. Hence, it can be argued that the fitting factors of 1.15 can provide extra rivets that will enable the fail safety in the joint at the design limit load.
Further numerical and experimental failure studies on single and multi-row riveted lap joints
Weitere numerische und experimentelle Schadensuntersuchungen an einfach und mehr-reihig genieteten Überlappstößen
Engineering Failure Analysis ; 20 ; 9-24
2012
16 Seiten, 19 Bilder, 5 Tabellen, 25 Quellen
Article (Journal)
English
An experimental and numerical analysis of riveted single lap joints
Online Contents | 1994
|An Experimental and Numerical Analysis of Riveted Single Lap Joints
SAGE Publications | 1994
|Further experiments on large size riveted joints
Engineering Index Backfile | 1923
|Riveted single lap joints. Part 1: A numerical parametric study
Online Contents | 1997
|Riveted single lap joints. Part 1: A numerical parametric study
SAGE Publications | 1997
|