PSI - Issue 24

Filippo Cianetti et al. / Procedia Structural Integrity 24 (2019) 526–540 Author name / Structural Integrity Procedia 00 (2019) 000–000

531

6

(a)

Fig. 5: Discretization for Experimental Modal Analysis

2.3. Experimental Validation: from the Quadrature Picking to the Modal Assurance Criterion

To strengthen the predictive capacity of the dynamic model, an experimental test campaign was carried out from which the mode shapes and the natural frequencies were derived. The object, suspended by means of wires, was discretized in 8 nodes (Fig. 5) and excited with an impact hammer, recording the response through a mono-axial accelerometer in the Z direction with reference to the indicated triad in Fig. 4 (b) (out of plane in Fig. 5). From the input and output signals the experimental FRFs have been calculated according to the estimator Døssing (1988a):

R i , o R i , i

H 1 =

(12)

where the input and output R i , o cross-spectrum is divided by the auto-spectrum of the input R i , i , so as to minimize the noise on the response. The estimator H 1 is shown to converge to the real FRF at the increase of the number of averages on measure Døssing (1988a), which in this case are 10 for each node.

10 2

10 1

Experimental Numerical

Experimental Numerical

10 1

10 0

ms -2 /N

ms -2 /N

10 0

10 -1

10 -1

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Hz 10 -2

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Hz 10 -2

(a) FRF Node 1 to node 1

(b) FRF Node 2 to node 1

Fig. 6: Numerical vs Experimental FRF.