PSI - Issue 12

Lorenzo Bergonzi et al. / Procedia Structural Integrity 12 (2018) 392–403 Lorenzo Bergonzi / Structural Integrity Procedia 00 (2018) 000 – 000

10

401

crosshead displacement divided by L0 for the two geometries, it can be seen that the curve relative to the equivalent specimen shows a deviation due to settling of the specimen in the fixture seat while load on the shoulders increases (ref. Figure 12 b). The ASTM specimens also exhibit this deviation, although very small as seen in Figure 12 a: the word "TM" indicates the curves with deformation determined considering the free-to-deform portion of specimen.

Table 7. ABS mechanical properties obtained with ASTM and equivalent specimens considering 0° orientation

F m

R m

E

A s %

s M

ASTM D638

[N]

[MPa]

[MPa]

[mm/mm]

[mm]

Average value 1185.10 29.97 1803.29

2.04 0.05

2.09 0.04

Standard dev.

19.66

0.50

22.23

F m

R m

E

A s %

s M

Equivalent

[N]

[MPa]

[MPa]

[mm/mm]

[mm]

Average value 1195.32 29.91 1859.81

2.00 0.04

3.67 0.10

Standard dev.

12.50

0.31

17.51

% Variation

0.9%

0.2%

3.1%

1.6%

75.5%

(a)

(b)

Figure 12. Stress - extensometer deformation (a) and stress – L0 deformation (b) for ABS 0° orientation: as can be seen (b) there is an initial toe region due to specimen settling in fixture seats, whilst good accordance exists for the curves obtained using an extensometer (a)

5. Conclusions Starting from the specifications stated in the MaCh 3D project, a specimen geometry has been identified that is equivalent to what is defined by ASTM D638 - Type I. The individuation of the initial specimen and fixture conformation has been carried out by FEM analysis, obtaining a good match between the results from the equivalent and standard geometry. Further numerical study on fixture seating, which allows the self - alignment of the specimen during loading and considerably facilitates the operations test set-up, has allowed to avoid the presence of stress concentration on the sides of the specimen which could cause its failure outside the measurement section. The numerical results were confirmed by experimental tests: these were performed on a traditional UTM, in order to compare the two geometries using the same machine. Results of the individual series of tests have concordant values and a similar data dispersion.