Stress-corrosion cracking of rapidly solidified magnesium-aluminum alloys

Stress corrosion cracking (SCC) of rapidly solidified magnesium-aluminum alloys in aqueous solutions of potassium chromate and sodium chloride was investigated using electrochemical techniques, constant displacement rate tests, and optical and electron microscopy. The electrolyte selection was based on its relevance to the service environment, where passivating chromate ions, applied as a corrosion protection measure, coexist with chloride ions indigenous to seawater and other natural electrolytes. Microcrystalline alloys containing 1, 5, and 9 wt.% aluminum were prepared using a melt-spinning process which yields continuous ribbons 15-25 {dollar}mu{dollar}m thick. The investigation sought to provide an understanding of the effects of aluminum content and rapid solidification on the repassivation rate, relate these changes to SCC behavior, and expand the current understanding of SCC in Mg-Al alloys.; Potential-pulse and scratched electrode experiments demonstrated benefits of both rapid solidification and increased aluminum content. The melt-spun alloys experienced relatively uniform attack, and repassivated faster and more completely than their as-cast counterparts. All the alloys, as well as pure magnesium, failed by transgranular stress corrosion cracking (TGSCC) in aqueous 0.21 M K{dollar}sb2{dollar}CrO{dollar}sb4{dollar} containing 0.6 M NaCl at displacement rates between 5 {dollar}times{dollar} 10{dollar}sp{lcub}-5{rcub}{dollar} and 9 {dollar}times{dollar} 10{dollar}sp{lcub}-3{rcub}{dollar} mm/s. This failure mode was manifested in quasicleavage on the fracture surfaces and in lower maximums in stress intensity and displacement. In chloride solution without chromate, TGSCC occurred only near 3.6 {dollar}times{dollar} 10{dollar}sp{lcub}-3{rcub}{dollar} mm/s, while no stress corrosion was observed in chromate solution without chloride.; Constant displacement rate tests in air after pre-exposure to the electrolyte indicated TGSCC results from a hydrogen embrittlement process. Using realistic estimates of the diffusivity of H in Mg, the results from the constant displacement rate and potential-pulse tests for Mg-9Al are best explained by a model involving the formation of magnesium hydride ahead of the crack tip, and the transition from hydrogen embrittlement to ductile tearing predicted by crack velocity calculations corresponds to that observed experimentally. Crack velocities calculated from potential-pulse results suggest anodic dissolution is not the operative mechanism for TGSCC in these experiments.