Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Materials Science and Engineering

Major Professor

Raymond A. Buchanan, Peter K. Liaw

Committee Members

Chain T. Liu, Charlie R. Brooks, Charles S. Feigerle, Easo P. George


Although the mechanical behavior of Zr-based bulk metallic glasses (BMGs) has been studied, fatigue studies in controlled environments and testing conditions have not been adequately performed or understood. Early fatigue examinations of rectangular bar specimens tested in bending had resulted in fatigue strengths substantially lower than anticipated [1,2,3,4,5]. Preliminary work performed by Peter et al. and Wang et al. on uniaxial button-head specimens have yielded in very different fatigue behavior with fatigue-endurance limits comparable to conventional high-strength, crystalline alloys [6,7,8,9,10]. Between all S-N results studied, the fatigue-endurance limits for Zr-based BMGs have been observed to range from 150 MPa to 1 GPa [1,2,3,4,5,6,7,8,9,10]. Testing conditions, sample preparation, and the quality of the amorphous alloy may provide the understanding for this variability in fatigue behavior.

In the following thesis, several investigations were engaged to better understand changeability in the fatigue behavior of a Zr-based BMG, Zr52.5Al10Ti5Cu17.9Ni14.6 (at. %). The studies were primarily conducted to explain how the loading conditions, the sample preparation, the quality of the glass materials, the test environment, or the chemical composition affect the degradation behavior of BMGs. Fabrication, corrosion, metallography, and cyclic loading were types of investigations performed.

Fatigue testing was conducted in air and vacuum environments. By comparing the results, the environmental effects due to water vapor (in air) on the fatigue lifetime of BMG-11 were evaluated. It was concluded that water vapor does not have a detrimental effect on the fatigue lifetime of BMG-11. Indeed, the observed lifetimes in vacuum were shorter than those in air. Early fatigue tests in vacuum with and without the use of an ionization gauge seemed to indicate that the dissociation of the residual water vapor to atomic hydrogen in vacuum via a hot tungsten-filament ionization gauge could be a factor in the shorter fatigue lifetimes in vacuum than in air. Further testing has disproved this hypothesis. Closer examinations of the surface of multiple samples has led to the discovery of mechanical wear and fused copper (from copper grips) near the crack initiation site.

Because of the possible impact the ionization gauge may have had on the fatigue behavior of BMG-11, hydrogen-charged samples were tested in air and compared to uncharged samples in order to understand any detrimental effects hydrogen may have on the fatigue lifetime around the fatigue-endurance limit. Though the ionization gauge did not seem to play a detrimental role in the fatigue lifetime of amorphous samples tested in vacuum, charged hydrogen embrittles the material with increases in hardness values and lower fatigue lifetimes for cathodically-charged samples. These results could impact Zr-based BMGs’ usefulness in hydrogen-rich environments.

Fatigue studies were performed on button head, uniaxial specimens with different surface finishes in order to better understand the influence the average surface roughness and/or critical surface defects may have on the fatigue behavior. It was hypothesized that geometric, surface flaws could lower the observed life of a BMG sample by shortening the crack initiation phase and providing local stress concentrators. Careful studies of surface conditions indicate that fatigue-endurance limits are greatly impacted by the average surface roughness with possible reductions over fifty percent. Lastly, a rectangular bend-bar sample was finished with a coarse grit paper on the tension side of the sample, and observed for the location of crack initiation.

Four-point and three-point bending fatigue studies were conducted to observe the effect of variability in loading conditions versus uniaxial tension studies, and to observe any impact from testing volume on the fatigue life of BMG-11. Both, three point and four point bend results seemed to exhibit slightly better fatigue behavior compared to the uniaxial tests. However, little difference was observed between three-and four-point bending. These results are similar to those found with a preliminary study of uniaxial specimens with varying testing gauge lengths.

Lastly, a study was performed to better understand the effect crystallinity has on the fatigue behavior of Zr-based bulk metallic glasses. The crystalline phases of BMG-11 have an extremely low ultimate bending strength, around 100 MPa, and are very brittle. Any interfaces between crystalline impurities and the glassy matrix are prime locations for crack-initiation sites. A large volume fraction of crystallinity has been shown to dramatically lower the fatigue lifetime of a Zr-based BMG. This careful study of fatigue behavior leads to the conclusion that the detrimental effect crystallinity and geometrical surface defects have on the fatigue-endurance limit and the fatigue lifetime explain the variability in previously reported results.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."