Monitoring Gallium-Induced Damage in Aluminum Alloys Using Nonlinear Resonant Ultrasound Spectroscopy

Abstract

Nonlinear Resonant Ultrasound Spectroscopy is a nonlinear ultrasonic technique which allows monitoring small variations in the microstructure of a medium and thus allows materials characterization and monitoring of damage evolution. Application of the technique to monitor Liquid Metal Embrittlement induced by gallium penetration in aluminum is presented here. To define indicators of material degradation, data treatment using the Singular Value Decomposition approach is introduced and discussed. Experimental results show that nonlinear properties are correlated with the state of the liquid metal in the solid matrix, allowing to identify different phases in the process of gallium diffusion along grain boundaries and within the bulk of individual grains. Furthermore, the evolution of gallium damage allows to study correlations between nonlinear, fast and slow dynamic properties.

Figures (12)

• Figure 1: Definition of NRUS parameters. Schematic power spectrum of the resonance peak. The $Q$ factor is determined from the resonance frequency and full width at half maximum ($W$) and the asymmetry factor $\Phi$ as a ratio of half widths $B$ to $A$ at 1/10 of the peak height. Experimental results shown correspond to data taken after one hour from the beginning of gallium embrittlement and at low excitation amplitude.
• Figure 1: The phenomenological functions as the most significant singular vectors of the followed indicators for the second sample tested.
• Figure 1: Resampling of the data matrix. a) Data of relative velocity change shown in time of measurement and output amplitude coordinates. b) Histogram of output amplitude values from the whole monitoring. Red dashed lines indicate the $A_\text{out}$ values used for resampling. c) Resampled values superimposed on original data for selected measurements.
• Figure 1: a) Modified NRUS amplitude sequence including a downward ramp. The sequence timing agrees with previous probing. b) Relative velocity change vs. strain: computed without baseline reference correction (red), with baseline reference correction (black) and with amplified baseline reference correction (green). c) Baseline reference vs. strain: measured (red) and amplified (green) baseline velocity variation. d) Slow dynamic response of gallium damaged sample. Measurement are taken on the first sample after 40 hours of gallium damaging. Note that the relaxation curve is amplified by a factor of 2 to allow appreciating the temporal evolution.
• Figure 1: Linear and nonlinear elastic parameters during temperature change without (red) and with (black) deposited gallium. Temporal evolution of a) linear wave velocity, b) coefficient of nonlinear elasticity, c) coefficient of slow dynamics, d) linear damping, e) coefficient of nonlinear damping, f) distortion coefficient.