PSI - Issue 75

Kalle Lipiäinen et al. / Procedia Structural Integrity 75 (2025) 19–28 Lipiainen et al./ Structural Integrity Procedia (2025)

21 3

Therefore, the input load data for re-designing the component was readily available, and the component was considered as a good case study for the L-PBF research. Achieving airworthiness for the component was not the objective of this research study but instead to identify the readiness of L-PBF techniques to produce components for aerospace engineering. The components are installed inside the center box of the tailplane between lower and upper skins. One tailplane consists of two fittings, one of each side, LH & RH. In Hawk Mk.66, the tailplane acts as a horizontal stabilizer and elevator. The fittings carry the aerodynamical forces between tailplane and airframe resulting from the pitch control. 2.2. Material and component design for additive manufacturing Before component design for AM, material selection was done. The objective was to fulfil the fatigue life requirement (~35 000 cycles at the given loads of 20 kN and -13.2 kN) with decreased mass (the original mass was 202g). High-strength aluminum alloys, such as AlSi10Mg, would have had good technology readiness levels, but not optimal material properties regarding fatigue strength for this application. Some new aluminum alloys for powder bed fusion had better material properties but low technology readiness levels or limited amount of fatigue data. In the meantime, studies on the fatigue properties of Ti64 material were on-going. In general, many potential applications seemed to end up to Ti6Al4V alloy and it was also chosen for this application to gain more knowledge. The material fatigue data for the L-PBF Ti6Al4V was not available before the component design process started and, therefore, the maximum cyclic stress value was determined to be around 200 MPa based on the earlier studies by Kahlin (2017). Considering the structural details in the component, the AM tail fitting was seen as a representative notched bar with as-built surface quality. The material fatigue test results of the studied titanium material are illustrated in Fig. 2. The effect of surface conditions is highlighted by the results with the R = -0.3 fatigue tests (Fig. 2a). With the machined surfaces, very high fatigue performance was obtained.

2000

2000

2000

Machined R = -0.66 As-built R = -0.66 As-built with shot peened surfaces R = -0.66

Nominal stress range [MPa]

Machined R = -1 As-built R = -1 As-built with shot peened surfaces R = -1

Machined R = -0.3 As-built R =-0.3 As-built with SP surfaces R = -0.33

Component EOS R = -0.66 Component SLM R = -0.66

200

200

200

10000

100000

1000000

10000

100000

1000000

10000

100000

1000000

Fatigue life [cycles]

Fig. 2. (a) Small-scale fatigue test data with stress ratios of R = -0.33, R = -0.66 and R = -1. The results from the database of University of Oulu.

In the original structure, the load is directly applied to the web of the part. It could be seen that achieving a more optimal geometry for the new part would be challenging if the weight target was to be maintained. Therefore, the design process started with topology optimization by Altair Inspire 2021 software. The optimization considered the requirements for minimized weight and maximized stiffness. The optimization was carried out with few types of shape controls and boundary conditions while the load cases were similar to ones of the original structure. The design process was iterative, and FEA work was executed with Inspire also to enable swift iterations in the design. Several different geometries containing polyNURBS shapes were created, but all of them included stress concentrations that were difficult to eliminate. Also, these geometries would have required a significant amount of support structures during manufacturing. Finally, the topology optimization was used only to support engineering conclusions. The final design was acquired with CATIA V5 while the FEAs during the iterations was conducted with Altair Inspire. The final design included a double web configuration. While the surface area per volume increased with double web design it considered to be more damage tolerant design. This geometry also allowed to produce