This thesis deals with composite joints. Designing such joints is more difficult than metallic joints due to the mechanical properties of composite materials. Composites are anisotropic and have a limited ability of yielding. The low degree of yielding means that stress concentrations are not relieved by plastic deformation, which is important in multi-fastener single-lap joints. The distribution of load between the fasteners may be more uneven than in metallic joints due to that the stress concentrations around the holes are not relieved. Single-lap joints have an eccentric load path which generates a nonuniform bolt-hole contact pressure through the plate thickness. This generates out-of-plane deflection of the joint, termed secondary bending. Such nonuniform contact stress severely limits the strength of the joint. The nonuniform contact stress distribution is affected by several factors, e.g. bolthole clearance and secondary bending. The first part of the work is devoted to investigating secondary bending, and its effect on stresses in the joint. A novel technique to study secondary bending has been developed and used in a parametric study. It is based on the calculation of specimen curvature from out-of-plane deflections measured with an optical technique. It is shown that the specimen curvature is correlated to the conventional definition of secondary bending, which involves strain measurements on both sides of the plate. The two most important parameters affecting specimen curvature was found to be the overlap length and the thickness of the plates. The finite element method was used to study the influence of secondary bending on joint strength. Secondary bending was changed in magnitude by altering the length of the overlap region in a two-fastener specimen. It was found that secondary bending affects the local stress field around the fasteners and that it may change the strength and the mode of failure. The second part is concerned with the load distribution and prediction of joint strength. A detailed finite element model was developed to calculate the load distribution while accounting for bolt-hole clearances, bolt clamp-up, secondary bending and friction. An experimental programme was conducted in order to validate the finite element model by means of instrumented fasteners. Good agreement between simulations and experiments was achieved and it was found that bolt-hole clearance is the most important factor in terms of load distribution between the fasteners. Sensitivity to this parameter was found to be large, implying that temperature changes could affect the load distribution if member plates with different thermal expansion properties are used. Calculating the load distribution in structures with a large number of fasteners is in general not feasible with detailed finite element models based on continuum elements. A simplified, computationally effective model of a multi-fastener, singlelap joint has been developed by means of structural finite elements. The model accounts for bolt-hole clearances, bolt clamp-up, secondary bending and friction. Comparisons with the detailed finite element model and experiments validated the accuracy of the simplified model. A parametric study was conducted where it was found that an increased stiffness mismatch between the plates generates a more uneven load distribution, while reducing the length of the overlap region has the opposite effect. Increasing the stiffness of a fastener shifts some of the load from the nearest fasteners to that particular fastener. An idealized optimization study was conducted in order to minimize bearing stresses in the joint with restrictions on the increase of joint weight and net-section stresses. Maximum bearing stress was reduced from 220 MPa to 120 MPa while both weight and net-section stresses decreased. A procedure to predict bearing strength based on the results from the simplified model was developed. It was established by an experimental programme that fiber micro-buckling is the initial failure mode. The stress state in the laminate was determined through force and moment equilibrium, based on output from the finite element model. An existing criterion was used to predict the fiber microbuckling, and thus the initial failure. Predictions were compared with experiments which validated the method. The small computational cost required by the procedure suggests that the method is applicable on large scale structures and suitable to use in conjunction with iterative schemes such as optimization and statistical investigations. / QC 20110121
| Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:kth-4006 |
| Date | January 2006 |
| Creators | Ekh, Johan |
| Publisher | KTH, Farkost och flyg, Stockholm : KTH |
| Source Sets | DiVA Archive at Upsalla University |
| Language | English |
| Detected Language | English |
| Type | Doctoral thesis, comprehensive summary, info:eu-repo/semantics/doctoralThesis, text |
| Format | application/pdf |
| Rights | info:eu-repo/semantics/openAccess |
| Relation | Trita-AVE, 1651-7660 ; 2006:22 |
Page generated in 0.0024 seconds