PSI - Issue 34

Zoltan Major et al. / Procedia Structural Integrity 34 (2021) 191–198 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

193

3

sequences of the loading - number of load levels, cycle numbers - were varied. The test frequency was in the range of 1-10 Hz and the tests were frequently performed at room temperature (RT). The local crack tip loading was characterized by tearing energy (T). The length of the typically blunted crack was optically measured and crack propagation rate values, da/dN were calculated and subsequently combined with the calculated tearing energy values. The tearing energy was calculated using the deformation energy values corrected with the crack length. The basic assumption was that for appropriate planar tensile specimens no geometry factor is needed for calculating the tearing energy for sharply notched specimens [Mars, 2001]. While the stable part of the crack growth curves can be used for predicting the fatigue life time, the threshold values of these curves were used to predict the endurance strain limit,  eth of model components. The flexibility of the additive manufacturing makes the production and usage of unique non-standardized specimen configurations easily possible. The cylindrical dumbbell specimen reveals several advantages: o Due to the higher stiffness both displacement and force-controlled tests can be performed reliably o Due to the higher stiffness both uniaxial tensile and compressive (without buckling) testing are possible (variation of the R ratio) o The cylindrical specimen can be loaded in torsion and proper shear stress-strain curves can be determined o Multiaxial (axial/torsional) loading is also possible o Local stress concentrations can be introduced (notches with various radius) The degree of inherent anisotropy of additively manufactured materials depends on a number of factors and is typically underestimated by practical engineers. Based on some similarities with composite materials, we can consider some basic corner points of a proper fatigue methodology. The engineer’s perception of the phenomenon of fatigue is so closely associated with the behavior of macroscopically homogeneous, isotropic materials that there is often a tendency to treat other materials (i.e., fiber reinforced composites or additively manufactured metals and polymers) as they were conventionally manufactured materials. Since anisotropy is a characteristic of additively manufactured materials that we should accept we have to take it into account in the design. Besides the need to understand the mechanisms by which fatigue damage occurs in additively manufactured polymers, we also need methods that can reliably predict the development and accumulation of this damage and thus the likely lifetime of the material (or component) [Harris, 2003 and Aidy, 2018]. In recent study, we preferred the stress-based technique in order to properly compare PA12 and TPU. In our research project we are using these two commercially available materials for manufacturing various lattice structures [MOAMMM, Horizon 2020, grant agreement No 862015 . ]. The commercial availability is an important issue regarding the quality assurance both in the process and for the component. The results generated on laboratory test specimen level should be transferred to components containing lattice structures. While in this paper we are focusing on the load-controlled fatigue test of SLS TPU materials, displacement controlled fatigue tests will also be performed and local strain-based Wöhler-curves will be determined in the next phase, for additively manufactured (SLS) TPU. These results can then be compared with existing results of injection molded thermoplastic polyurethane materials using similar cylindrical specimen configurations. In addition, there is also a need for shear strength data of the interface and the multiaxial combination of axial/torsional data, which are currently under investigation.