PSI - Issue 42

C. Boursier Niutta et al. / Procedia Structural Integrity 42 (2022) 1449–1457 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

1451

3

Two multilevel factorial plans were retained and these are summarized in Tables 1 and 2.

Levels

1

2

3

4

Factor A , beam diameter d Factor B , unit cell size l

0.9 mm (0.8 mm) 6.38 mm

1.2 mm (1.1 mm) 8.35 mm

1.5 mm (1.4 mm)

1.8 mm (1.65 mm)

12.75 mm Table 1. First factorial plan: variation of the beam diameter and size of the unit-cells.

Levels

1

2

3

4

Factor A , beam diameter d Factor B , number of cells n

0.9 mm (0.8 mm)

1.2 mm (1.1 mm)

1.5 mm (1.4 mm)

1.8 mm (1.65 mm)

2x2 4x4 Table 2. Second factorial plan: variation of the beam diameter and number of repeating unit-cells [5]. 3x3

Tables 1 and 2 report the nominal and the actual (in brackets) values of the beam diameter, as measured through an optical microscope. In the first factorial plan, Table 1, the volume of the specimen is constant, i.e., the edge of the cubic specimen is kept equal to 25.5 mm, and the size l of the unit-cell varies. The number n of unit-cells changes in accordance: as the volume of the specimen is constant, the larger the length l , the smaller the number of repeated unit-cells n . In the second factorial plan, Table 2, the volume of the cell is constant, i.e., the size l is fixed to 8.35 mm, and the number of cells is varied: 2x2, 3x3 and 4x4 specimens are tested. The goal of the first factorial plan was thus to investigate the influence of the cell size and diameter on the absorbed energy in a quasi-static crushing test. On the other hand, through the second factorial plan, it was possible to draw conclusions on the effect of the cell repetitions, as the design of lattice components is generally carried out by considering the energy-absorbing properties of the single cell. Fig. 2a shows the carbon nylon specimens printed by FDM, while in Fig. 2b the experimental setup is shown.

(a) (b) Fig. 2. a) Carbon nylon specimens and b) experimental setup of the crushing tests [5].

The specimens were subjected to quasi-static crushing tests through an electro-mechanical testing machine, Zwick Roell Z100. The crosshead speed was imposed to 1 mm/min and the tests were conducted until the material