Radiography (x-ray) and ultrasound technology have been two nondestructive evaluation (NDE) techniques that have the ability to provide users with useful volumetric data. The NDE data obtained from these processes can be quite diverse in nature from quantifying defect sizes to measuring microstructural texturing in specific alloys. In addition, radiographic testing (RT) and ultrasonic testing (UT) have been rivals of each other and there is a debate on which one reigns supreme in the NDE community. As with any particular process in any application, there are advantages and disadvantages of the implementation of both processes. But there have been studies conducted to combine both x-ray and ultrasound for in-process (or in situ) monitoring.

I recently read a journal article titled, “In situ characterization of laser-generated melt pools using synchronized ultrasound and high-speed X-ray imaging” where the authors studied if X-ray and ultrasound can be combined to effectively monitor additive manufacturing (AM) builds. The authors provided detailed analysis on 2D and 3D finite-element (FE) simulations, which are used to set up/optimize their synchronous X-ray and ultrasound experiments on laser powder bed fusion (L-PBF) AM aluminum substrate. As a side note, FE simulation or modeling is a numerical method to solve differential equations that mimic engineering applications [1]. L-PBF is a AM (or 3D printing) process that uses lasers to melt a powder to create a build layer during manufacturing. The laser beam dynamics can either create a melt pool in conduction (ideal) or keyhole mode during the AM process [2]. The paper focused on measuring depth of the keyhole because the keyhole can sometimes lead to complications during the AM process builds.

The experimental configuration of the X-ray and ultrasound sensor are positioned in a way that both the X-ray beam and the ultrasonic wave interact with the melt pool made by the laser beam [2]. The main points of this paper are, first, X-ray and ultrasound can be synchronized via their frequencies to measure the laser beam dynamics. Secondly, X-ray can be used as a visual measurement tools for laser beam dynamics while ultrasound can provide depth measurements from the amplitude scattered data. Third, FE simulations have great accuracy and agreement with physics of ultrasonic wave propagation. Additionally, FE can be used to optimize ultrasonic wave propagation experiments to ensure that valuable data is captured to advance NDE technology forward. FE simulations are very powerful but they do not always capture every variable that is tested in actual experiments. For instance in the featured paper of this post, some variables had to be left out because of computational time or assumptions had to be made based on the overall complexity of the actual system.

One point that the authors highlight about this paper is that there is a linear relationship to the ultrasonic amplitude and the depth of the melt pool [2]. The authors of the paper also prove that this relationship held true in their early FE simulations. Linear relationships are valuable because they offer simplicity to complex system if they are proven. The experimental set up only allowed for a depth of 400 μm and the authors plan to capture larger depths in future experiments. There may be a case where the linear relationship breaks down at greater depths due to higher temperatures or longer processing time. But the aforementioned comment is just speculation on my part.

So what do these experiments of coupled X-ray and ultrasound mean for the future of NDE? For starters, this study along with countless other studies on hybrid NDE methods have proven that combining two NDE methods together is possible and can be tailored to provide the necessary resolution for various engineering application. The emergence of AM has brought in more innovation to NDE in-process monitor to close the gap between manufacturing advancements. Lastly, computation methods and computer technology has come a long way from the 1960s and is still making exponential growth in hardware and software. NDE experiments and products will also advance in terms of optimization and capabilities.

I hope you found this helpful and as always thank you for your time!

–DB PhD

References:

  1. Nikishkov, G. P. (2004). Introduction to the finite element method. University of Aizu, 1-70.
  2. Gillespie, J., Yeoh, W. Y., Zhao, C., Parab, N. D., Sun, T., Rollett, A. D., … & Kube, C. M. (2021). In situ characterization of laser-generated melt pools using synchronized ultrasound and high-speed X-ray imaging. The Journal of the Acoustical Society of America, 150(4), 2409-2420.