PSI - Issue 28

12

Giovanni Pio Pucillo et al. / Procedia Structural Integrity 28 (2020) 1998–2012 GP Pucillo et al. – Part I / Structural Integrity Procedia 00 (2019) 000 – 000

2009

μ m/m

μ m/m

μ m/m

hole #

15900 4350 900 -2550 -6000 -9450 -12900 -16350 -26000

28000 21187.5 14375 7562.5 750 -6062.5 -12875 -19687.5 -26500

14700 4925 1250 -2425 -6100 -9775

ε xy

ε yy

ε xx

#2

11200 4925 1250 -2425 -6100 -9775 -13450 -17125 -20800 -10100 -13400 -16700 -20000 11209 3100 -200 -3500 -6800 -13450 -17125 -22500

μ m/m

μ m/m

μ m/m

8390 1300 -1700 -4700 -7700

23200 17100 11000 4900 -1200 -7300 -13400 -19500 -25600

ε xy

ε yy

ε xx

#5

-10700 -13700 -16700 -19700

μ m/m

μ m/m

μ m/m

31000 24062.5 17125 10187.5 3250 -3687.5 -10625 -17562.5 -24500

10602 2200 -1400 -5000 -8600 -12200 -15800 -19400 -27400

ε yy

ε xx

ε xy

#6

y

x

Fig. 12. Residual strain field measured by 2D-DIC after 2.0% of CE.

4. Discussion - 2D-DIC vs. strain gauge data In this Section, the experimental results obtained by strain gauges and by 2D-DIC are compared. Being the 2D DIC technique limited to measurement of in-plane displacements and strains of a planar object surface, experimentally measured residual strains can be considered accurate only for the central (flat) part of the rail web surface. For this reason, DIC results are compared with those obtained by strain gauges only along the rail longitudinal axis, i.e. along the direction at 0°. The comparison between results concerning hole #6, hole #2, and hole #5 is shown in Fig. 13. As it is shown in Fig. 13-a, DIC results referred to hole #6 are in very good agreement with strains measured by strain gauges, being the percentage difference between the strains measured by DIC and those measured by strain gauges of about 5%. Concerning hole #2 (Fig. 13-b), a good agreement between the results obtained with the DIC and the strain gauges is observed in terms of trend of the residual strains as a function of the hole distance, even if a higher percentage difference ( ≃ 25 %) was found with respect to what observed previously for hole # 6. This can be due to some imperfection in the speckle pattern or, more probably, to out of plane displacements of the rail during the cold expansion process. In fact, as explained in Section 2.2, the rail was fixed on the worktable only by mean of brackets and clamps, which may not remove any undesirable movement of the rail during the process. Concerning hole #5 (Fig. 13-c), the strain gauges and the DIC results appear almost not in good agreement. This disagreement could be attributed, as for hole #2, to the speckle and to out of plane displacements of the rail; moreover, during this experiment a non-parallelism between the CCD sensor and the rail web surface was observed, which strongly affects the results.