Feature Article Figure 10. Finite element analysis of the test fixture. Figure 11. Load application fixture. separately. Since the aim of the test was to obtain load-strain data and validate the developed numerical models, the part was only loaded to a pre-defined level and then unloaded to avoid any failures. Several strain gauges were attached to different parts of the fuselage to monitor and recorded the deformation caused by the load. The recorded strain values were as expected; with the fuselage experiencing tension on the top, compression on the bottom and almost no longitudinal strains on the sides. From the collected data it was evident that nearly all the accumulated strain in the loading phase was removed in the unloading step and loading-unloading strain curves were overlapping. This verified that no local damage occurred in the test around the measurement areas. Digital Image Correlation (DIC) was also used to record the deformation of a bigger area compared to the strain gauges. The area covered was a square 300mm x 300mm on the top of the structure. For further validation of the simulation approach the SAMPE Journal, Volume 52, No. 2, March/April 2016 full scale test of the prototype was modelled in Abaqus standard. The experimental set-up was substituted by MPCs that transferred the load to the structure (Figure 12). When the strain gauge measurements were compared to local strains predicted by the analysis, there was a noticeable difference between the top and the bottom of the cylindrical section as shown in Figure 13. The strains predicted by the model in tension agreed very well with the experiment; however, the compressive strains at the bottom were found to be lower in the simulations. This applied for both the skin and the ribs. This difference was caused by the modelling idealisation and application of the load by MPC constraints. In reality the loads are transferred differently in the tension and compression sections of the structure. The compression at the bottom of the cylinder is applied by all the surfaces coming in contact with the test fixture plate, while the tensile loads at the top are transferred mainly through the ribs bolted to the fixture plate by integrated pins. 31