PSI - Issue 12

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

4

395

Both models have the same mechanical properties, shown in Table 1, and the same displacement of 0.5 mm applied along the load direction of the specimen in correspondence of a line of nodes 73.5 mm away from specimen symmetry plane, Figure 3. This distance was chosen to fall within the gripping area of the standard specimen.

Table 1. Specimen material mechanical properties.

R m

E

R s

A s % A m %

Material

[MPa] [MPa] [MPa]

ABS

1250

32

32

3

49

2.1. Numerical analysis of standard ASTM D638 Type I specimen The model, representing a quarter of the standard specimen geometry, was created in Midas NFX 2015 using predominantly parabolic brick elements (CHEXA) for a total of 5435 elements. Symmetry constraints are applied in correspondence of the two ZX and ZY planes; the translation along Z has been constrained at a node outside the area of interest. The model represents a 3 mm thick specimen. This is a linear analysis and simulates a crosshead displacement equal to 1 mm: in this phase the stress concentration phenomena due to the compression induced by the closing of the fixtures on the specimen shoulders are ignored. The outline of the model is shown in Figure 3.

Figure 3. Outline of ASTM D638 specimen FEM model.

2.2. Numerical analysis of equivalent specimen

Starting from the imposed geometric constraints, such as shoulder diameter, the dimensions of the reduced section and the diameter of the holes in the shoulders, different models are created with different connection radii between shoulders and reduced section. The model (see Figure 4) represents a quarter of the geometry and follows that of standard specimen in terms of type of loads, constraints and mesh size. A first model with a radius equal to 76 mm, the same as the standard specimen, is analyzed. Since the stress distribution differs quite significantly from the reference values (see Figure 5), several models have been created varying the transition radius from a minimum of 50 mm to a maximum of 100 mm, with an increase of 5 mm between one model and the other. In addition, since the optimal radius amplitude is between 50 and 55 mm, a further series of models was created with an amplitude ranging from 50 to 55 mm with an increment of 1 mm at a time.