Microwave nondestructive testing (NDT) has emerged as a promising technique for evaluating the quality of additive manufacturing (AM) metal and polymer powder feedstock, surface inspection of conducting parts, and volumetric inspection of polymer prints. Ex-situ or in-situ microwave measurements link voltage or S-parameters to inherent electromagnetic properties of the material-under-test, such as complex permittivity. Varied measurement setups can help adapt the inspection to shape and space requirements. Subsequently, electromagnetic mixing models are used to relate these properties to corrosion, density, infill percentage, packing fraction, etc. These monitoring techniques can be used to improve in-line process control and the print’s overall quality, as well as support the development of advanced materials tailored for specific applications. This presentation will cover the use of electromagnetic mixing models and measurement setups in several example case studies, including packing fraction of a ceramic powder and the compression of a woven polymer print.

Air-coupled ultrasonic testing (ACUT) has become an attractive non-contact technique for inspecting composite structures where conventional contact or immersion methods are not feasible. Typical ACUT configurations include: (a) through-transmission compressional and shear waves, (b) pseudo pulse-echo, and (c) through-transmission or pitch-catch Lamb wave modes. However, all of these approaches suffer from inherently low signal-to-noise ratio (SNR) due to the strong acoustic impedance mismatch between the transducers, the air gap, and the specimen, combined with high internal attenuation in materials such as honeycomb, CFRP, and GFRP.
To address these challenges, advanced hardware and signal processing strategies are required—particularly those that enhance sensitivity and SNR. Matched filtering is one such approach: it correlates the transmitted excitation signal with the received waveform, maximizing SNR and improving time resolution. While traditionally difficult to implement due to the need for precise waveform generation and synchronization, recent ultrasonic systems equipped with arbitrary waveform generators (AWG) now make matched filtering practical for industrial NDT applications.
This study presents a comparative analysis of matched filtering applied to the main ACUT techniques for the inspection of honeycomb and CFRP structures. Experimental results demonstrate the SNR improvement achieved through matched filtering relative to conventional signal averaging, and the implications for inspection speed and reliability in industrial environments are discussed.

Automated Fiber Placement (AFP) has become a leading technique for manufacturing primary composite structures due to its high speed, repeatability, and process efficiency. However, it introduces unique defects—such as gaps, overlaps, and wrinkles—that are uncommon in traditional fabrication. A persistent challenge is the in-situ evaluation of tape quality and compaction during layup, both of which are critical to the final part performance yet remain difficult to assess in real time. This study proposes an advanced non-contact ultrasonic sensing approach for simultaneous detection of AFP defects, compaction levels, and prepreg tape quality. Ultrasonic guided waves are generated by a pulsed laser and recorded using dual-output air-coupled transducers. The Green’s function of the composite part is extracted efficiently and robustly via Normalized Cross Power Spectrum (NCPS) analysis, from which phase velocity and attenuation dispersion curves are derived. These measurements are utilized to identify the composite viscoelastic constants (storage and loss moduli) that are related to the state of cure of the part, as well as to detect and quantify common AFP defects such as gaps and overlaps.

Recent research from our group has shown that metal additive manufacturing (AM) can be used to create rollers for railroad bearings with excellent performance when subjected to rolling contact fatigue (RCF). In this presentation, we discuss the reasons that the AM rollers performed well in comparison with rollers that were conventionally manufactured. The AM rollers were created using laser powder bed fusion from 8620HC steel powder after the print parameters were optimized in terms of the laser power, laser spot size, laser scan speed, and layer thickness. Then, while still on the build plate, the rollers were stress-relieved. Subsequently, they were cut from the build plate using wire electrical discharge machining, machined to the appropriate geometry, carburized, case hardened, and finish ground to achieve the required surface finish. The rollers underwent simulated service life tests with RCF at levels typical of in-service bearings and reached a test equivalent of 250,000 miles with no visible damage. RCF of metals, in general, is inherently dependent on the sample microstructure, an aspect of metal AM for which many challenges remain. Ultrasonic inspection and X-ray computed tomography (XCT) were the primary NDE methods used to understand their performance. XCT provides information about the sample porosity and ultrasound, using both longitudinal and shear backscatter, was used to assess information about the microstructure. These results allow possible acceptance criteria to be proposed for AM parts subjected to RCF. Metal AM has made great inroads in many fields, but qualification and certification of parts is still a challenge. The results of this study provide a clear foundation for inspection strategies for AM parts.