PSI - Issue 17

Ludvík Kunz et al. / Procedia Structural Integrity 17 (2019) 222–229

223

Ludvík Kunz et al./ Structural Integrity Procedia 00 (2019) 000 – 000

2

1. Introduction

Growing production of Ti6Al4V alloy components produced by additive manufacturing (AM) technologies in aerospace, chemical and biomedical industry requires more knowledge on mechanical properties of AM materials. For reliable service of components which are exposed to cyclic loads deeper understanding of their fatigue behavior is required. For prediction of fatigue life the knowledge of data characterizing the fatigue strength and growth of fatigue cracks is a necessary prerequisite. Direct metal laser sintering (DMLS) is the most often used AM technology for production of end-use components of complex shape. Broad variety of processing parameters of commercial DMLS systems like scan speed and scanning strategy, laser power, layer thickness, component size, temperature, non-equilibrium phenomena during manufacturing process and input powder material influence the microstructure of the final product. Because the microstructure produced by DMLS is substantially different from the microstructure of conventionally produced alloy the mechanical properties can differ substantially. For this reason, the known data characterizing this alloy manufactured conventionally i.e. by forging, casting and rolling cannot be used for design purposes of additively produced parts. The mechanical behaviour of AM Ti6Al4V alloy in relation to the microstructure and AM processing parameters were investigated intensively in the last years. The available knowledge concerning the as-built material was reported in many papers, e.g. in a recent review paper Liu and Shin (2019). Engineering components manufactured from the Ti6Al4V alloy are often operated in load bearing structures, which have requirement on low weight. Typical examples are aerospace structures. Tensile strength and fatigue properties are thus the most important material parameters for prediction of component service life. Particularly the knowledge on the fatigue behaviour is crucial, because the resistance to the crack initiation and propagation can vary strongly in dependence on material microstructure, which is dependent on a wide range of production DMLS parameters and subsequent heat treatment. The fatigue life is determined by the initiation and propagation of fatigue cracks. DMLS can produce fully dense material in theory, however, in practice non-optimal processing parameters result in formation of porosity in the form of gas pores or lack-of-fusion pores. The consequence is a reduced fatigue limit. This issue was investigated intensively recently, e.g. Beretta and Romano (2017) with the result that the defect tolerant design concept with extreme value defect ratings can be adopted for AM parts. For prediction of fatigue life with existing cracks the knowledge on crack growth and thresholds for crack propagation is inevitable. Despite the fact that data on fatigue crack growth are available in literature, more data in relation to the manufacturing and post-processing is still needed in order to get deeper understanding of the material behaviour. The data on the growth of long fatigue cracks and threshold values indicate large scatter. Moreover, the available data are exclusively related to the cyclic loading at the stress ratio R = 0.1. The growth of long fatigue cracks and thresholds for crack propagation in Ti6Al4V manufactured by DMLS was investigated and described in numerous papers. The crack growth resistance differs substantially according to the parameters of the DMLS manufacturing process and according to the post-processing treatment. Especially the threshold values of the stress intensity factor  K th for crack growth lie in a wide range. The aim of this paper is to experimentally determine the crack growth curves and threshold values  K th for Ti6Al4V manufactured by DMLS and heat treated at three temperatures, namely 380, 740 and 900 °C and to investigate the influence of building direction, laser power and the stress ratio R on the long crack growth. Compact tension (CT) specimens according to the ASTM E 647-08 standard were manufactured on two EOSINT systems. M270 machine was equipped with 200 W laser, the laser beam speed was 0.8 ms -1 and bed hatch distance 100  m. The layer thickness was 30  m. The M290 machine with laser power of 400 W enabled larger layer thickness of 60  m. The laser scanning strategy was optimized by the AM Company BEAM-IT Fornovo Taro, Italy. The Ti6Al4V powder optimized for the used machines had the following chemical composition (in wt. %): Al 6.06, V 3.90, O 0.085, N 0.006, H 0.002, Fe 0.250, C 0.007, Ti balance. An example of the powder particles is presented 2. Material and Experiments