This work describes the use of nondestructive evaluation (NDE) techniques to characterize defects such as rust, voids, etc. in materials, and to analyze and predict strain and stress-induced breakdown. The use of terahertz (THz) imaging as an evaluation tool for novel materials and systems has found much success in recent years. While suggested as a potential NDE tool for use in the field of ceramic and ceramic matrix composite materials, the use of THz spectroscopy and imaging in the examination of the effects of mechanical and thermally induced strain on ceramic composite materials is not well established. In order to assess the effectiveness of THz imaging for use in the analysis of ceramic composite material health, it is necessary to determine if THz spectroscopic imaging can clearly highlight areas of the samples that have been affected by mechanical and thermal stress.

Spectroscopic Image of SiNC Samples based on the maximum time-domain amplitude. Samples include, from left to right: aluminum reference, baseline, heat treated 100 hours, fatigue treated, and heat treated 10 hours. Note that the textured weave on the surface of the sample is clearly visible.
Ceramic matrix composite materials are the ideal choice for the next generation of thermal protection systems due to their high strength and thermal resistance. The first round of data acquisition consisted of spectroscopic imaging of the samples to receive any heat or mechanical fatigue treatment. Subsequent rounds consisted of sample imaging following treatments of either thermally or mechanically induced stress of varying magnitudes.

Combinations of the two differing types of stress have also been studied. Comparison of the data both in the time-domain and frequency- domain acquired from imaging of the untreated and treated samples assesses whether or not the magnitude and extent of stress-induced changes can be monitored using terahertz imaging and spectroscopy.

THz time-domain images were acquired using a commercial system. Ultrafast laser pulses with an 800-nm center wavelength and 100-fs pulse width triggered a fibercoupled GaAs photoconductive antenna (PCA). Collimated THz light from the PC antenna transmitter was focused via a 50-mm focal length lens (f#=2) onto the samples at a near-normal incident angle. The reflected radiation was detected by a PCA receiver module, based on LT-GaAs, with an identical lens configuration. When the system is optimized and calibrated using a metal reference reflection target, the typical bandwidth of the detected THz pulse exceeds 3 THz.

Tests were conducted on both oxidebased and silicon nitride carbon-based CMC samples. A 5.9 × 17.4 cm area was scanned, containing both an aluminum reference and the CMC sample. A full time-domain waveform, 250 ps long, was acquired for each 0.5 × 0.5 mm pixel. The samples of dimension 15.7 × 1.3 × 0.2 cm are segmented into 11 blocks of equal area. This allows for comparative measurements to be isolated to the same area of the sample. Certain treatments are only applied to a specific area of the sample, and analysis is focused on those sections rather than on the entire sample.

The heat-treated samples appeared to have a different surface roughness than the undamaged baseline or fatigue-treated samples. In the oxide samples, however, no change in surface roughness is visibly apparent following any type of treatment.

Changes in the reflectivity of these samples pre- and post-stress treatment over smaller spatial sub-domains were noted. A grid system has been developed that allows for point-by-point analysis of the imaging data, which makes it possible to compare specific spatial points, with a spatial resolution of less than 1 mm, on samples prior to and following stress treatment. This analysis approach assesses the spatial variance of any stress-induced changes in the THz reflectivity. In addition, in-depth statistical analysis on the reflectivity data is currently underway that will be used to aid in determining if the data overlapped in error bars is distinguishable.

Following heat treatment of 1200 deg. C for 100 hours, the magnitude of the reflectivity maintained the same general trend of the baseline with a decrease in reflectivity by approximately 0.1 for the entire frequency span was shown. In the lower frequencies, out to 600GHz, there is no apparent change in the magnitude of reflectivity. However, at higher frequencies, the reflectivity diverges, with the fatigue treated sample exhibiting a lower reflectivity.

Dwell fatigue is a combination of fatigue and heat treatment induced simultaneously, and is concentrated only on the middle of the sample. Both fatigue treatments and dwell fatigue treatments have no effect on reflectivity, as no measureable change was measured.

This work was done by Adam Cooney of the Air Force Research Laboratory, and Lindsay Owens and Jason A. Deibel of Wright State University. AFRL-0217

This Brief includes a Technical Support Package (TSP).
Characterization of Ceramic Composite Materials Using Terahertz Nondestructive Evaluation Techniques

(reference AFRL-0217) is currently available for download from the TSP library.

Don't have an account? Sign up here.