High cycle fatigue of a die cast AZ91E-T4 magnesium alloy
This study reveals the micro-mechanisms of fatigue crack nucleation and growth in a commercial high-pressure die cast automotive AZ91E-T4 Mg component. Mechanical fatigue tests were conducted under R=-1 conditions on specimens machined at different locations in the casting at total strain amplitudes ranging from 0.02% to 0.5%. Fracture surfaces of specimens that failed in the high cycle fatigue regime with lives spanning two orders of magnitude were examined using a scanning electron microscope. The difference in lives for the Mg specimens was primarily attributed to a drastic difference in nucleation site sizes, which ranged from several hundred micrometers to several mm's. A secondary effect may include the influence of average secondary dendrite arm spacing and average grain size. At low crack tip driving forces (Kmax < 3.5 MPa (m)^1/2) intact particles and boundaries act as barriers to fatigue crack propagation, and consequently the cracks tended to avoid the interdendritic regions and progress through the cells, leaving a .ne striated pattern in this single-phase region. At high driving forces (Kmax > 3.5 MPa (m)^1/2) fractured particles and boundary decohesion created weak paths for fatigue crack propagation, and consequently the cracks followed the interdendritic regions, leaving serrated markings as the crack progressed through this heterogeneous region. The ramifications of the results on future modeling efforts are discussed in detail.
A recent push by the automotive industry to lower the fuel consumption and cost of production automobiles is providing enhanced motivation for the study of lightweight cast materials for structural components. Cast materials have clear economic advantage compared to wrought materials for mass production of components due to their lower long-term processing and assembly costs. On the other hand, processing related variability of monotonic and fatigue properties curtail the potential advantage of cast materials. Although cast and wrought materials can often have comparable maximum properties, cast materials typically show considerably more scatter in fatigue and monotonic properties. Consequently, castings have typically been designed by a worst-case-scenario paradigm, whereby the component is assumed to have the weakest material in the location of the highest stresses. Such an approach leads to over-designed components with material placed in unnecessary regions of the casting. The variability in the properties of as-cast materials is a direct consequence of the extremely strong dependence of the resulting microstructure on local solidification mechanisms. For example, the dimensions of the casting dictate the local cooling rate, which in turn produces a geometry dependent dendrite cells size and porosity level. A more robust design methodology for cast components would entail predicting the distribution of critical microstructural parameters as a function of the geometry of the casting, followed by life estimation based on the predicted microstructural features. To facilitate the interactive microstructure-based design of cast components for long life behavior, the mechanisms of fatigue and the sources of variability must be linked to critical microstructural features in the as-cast material. The present study examines the links between fatigue mechanisms and microstructure in cast magnesium (Mg) alloys, which have found a myriad of applications in industry [1–3] .
Total strain amplitude versus cycles to failure data from 34 AZ91E-T4 Mg specimens are presented in Fig. 1. The variability in the fatigue life of the small specimens exceeds two orders of magnitude for a given applied strain in the HCF regime. The variability in the fatigue life data in the low cycle fatigue regime is less severe, although still one order of magnitude on life.
The samples were etched to reveal microstructural features caused by the solidification of the casting. Fig. 3 shows high contrast SEM images of samples S1, S3, and S4 after etching. The etching reveals a dendritic structure within the grains. Based on the Mg–Al phase diagram, AZ91 Mg should contain a-Mg with trace solid solution Al and b-Al12Mg17. Fatigue crack formation The overall fracture surfaces of samples S1–S4 are presented in Fig. 4. In Fig. 4, the number of cycles to failure is indicated beneath the corresponding sample number. An arrow indicates the location of the fatigue crack nucleation sites on individual samples. Sample S4 has multiple arrows highlighting a large region on the fracture surface covered by subsurface trapped oxides. The extremely low number of cycles to failure of sample S4 is undoubtedly linked to these trapped oxides, which are found over nearly a third of the fracture surface.
- ↑  Jambor A, Beyer M. New cars – new materials. Mat Des 1997;18:203–9.  Decker RF. The renaissance in magnesium. Adv Mater Proc 1998;154:31-3.  Froes FH, Eliezar D, Aghion E. The science, technology, and applications of magnesium. JOM 1998;September:30–4.
Citation: High cycle fatigue of a die cast AZ91E-T4 magnesium alloy, M.F. Horstemeyer, N. Yang, Ken Gall, D.L. McDowell, J. Fan, P.M. Gullett, Acta Materialia