Design, built and test of a composite beam

The objective of this project is to design, build and test a representative aircraft component made from composite materials. The component in question is an undercarriage stay beam, and it will be subjected to a combined load case with axial, bending and torsional forces. The beam was manufactured via hand lay-up with prepreg material. The fibres were cut either by hand or using a ply cutter. The plies were laid up by hand on a male tooling. The male tooling was a 3D printed sand tooling, and was removed via wash out with water. The part was vacuum bagged and cured in an autoclave. Estimating the amount of material used in the design was conducted using CATIA as a design tool. The CAD software allows delimiting to detail the geometric form of the plies, and thus to efficiently and correctly distribute the pieces necessary for the manufacture. Efficient use of the material significantly depends on the distribution of the plies and their arrangement inside the cutting patterns. The mass of the composite beam structure was estimated by modeling the design in the CATIA software which provides a precise estimate of the area and mass contribution of each individual ply. The amount of uni-directional and woven material used in a part was estimated using the CATIA software. The beam was modeled in ABAQUS FEA software using 2D shell elements. The geometry was created by lofting through sections, after which it was partitioned to represent the various areas of pad-up and ply-drops that were defined above. Coupling constraints were applied around the holes to represent the presence of the pin, and the clamping constraint caused by the bushings. The pins were also modeled with correct strength and stiffness Finite Element Analysis (FEA) was performed in ABAQUS to assess the static stiffness and strength and buckling modes. It was determined that the FEA model was accurate and the results were taken to be the most accurate and current set of results, as they matched the hand calculations reasonably well but also took into account several design changes made since the hand calculations were performed. Buckling analysis via FEA was conducted for 10 eigenvalues (buckling modes) at ultimate load, using a linear perturbation, buckling analysis. The subspace eigensolver was selected, with 10 requested eigenvalues, 30 vectors per iteration and 100 maximum iterations.