PSI - Issue 12

A. Castriota et al. / Procedia Structural Integrity 12 (2018) 71–81 Castriota et al. / Structural Integrity Procedia 00 (2018) 000 – 000

75

5

in advance the bending behaviour of the component during the application of fatigue loads. The numerical model was used for the calculation of the deformations in the same points of application of the strain gauges, in order to make comparisons and verify the correct application of the load. A simulation was carried out at the predicted static load failure : at this load level, the spar remained in the elastic field, therefore it was possible to calculate the deformations in the case of other load levels simply by scaling the numerically calculated deformations respect to the load considered. Since this is a large component, the numerical model was built using the Shell 281 element with 8 nodes and the modeling was limited only to the spar in CFRP, thus not modeling the constraint and loads plates in aluminum. A sensitivity analysis was carried out which allowed us to choose an optimal number of elements equal to 10415. The repair near the edge of the hole was not modeled since its presence does not alter the global behaviour of the spar. The static scheme of the spar turns out to be that of a cantilever beam loaded at the tip (Fig. 3a): therefore, all the DOF at the root of the spar were fixed and a transversal load is applied in correspondence of the pin. Since load is applied through a pin on the aluminum end, an auxiliary rigid beam of suitable length has been foreseen. In this way, it was possible to recreate the same load condition that will occur in the laboratory, by inserting the load at the free end node of the additional element. In the numerical model, it was not necessary to insert the anti-warpage guides that are present in the experimental setup: since the numerical model is theoretically perfect and the load is perfectly centered, there is no risk of creating instabilities of torsion due to incorrect positioning of the load. The numerical simulation carried out with ANSYS with a load equal to (Fig. 3b) shows that the greatest deformations occur near the area of the hole on the web, which constitutes the weakest region of the spar and the most probable area of failure. 4.1. Constraint and load

a

b

Fig. 3. Numerical model’s geometry with applied constrains and load (a) and Contour plot of the deformation field of the spar calculated with ANSYS at load equal to (total strain) (b). 5. Discussion of experimental results

5.1. Preliminary static test (Step 0)

A preliminary static test (Step 0) provides the calibration of the measurement setup. In this test, the correctness of the strain gauge readings was evaluated based on what was calculated with the numerical model and the position of the load with respect to the symmetry plane of the spar was evaluated. From the diagram shown in Fig. 4, the absolute difference between numerical and experimental values is very small, although the repair near the hole has not been modeled numerically. The comparison shown in Fig. 5 is relative