In this section you will find information on specific applications of Dragonfly software, using examples that may be relevant to your application of interest.

If you do not find your application here, please get in touch so that we can showcase what is possible in your area of interest. Each application note includes a short “how to” video on YouTube to assist in using the software more effectively.

Deep learning

Everything you always wanted to know about Dragonfly Deep Learning but were too afraid to ask

This application note dives deeper into Dragonfly’s deep learning capabilities, exploring three popular misconceptions about the technology, and covering the most common applications for Computed Tomography (CT) image data: image segmentation, image enhancement and super-resolution. The application note also covers some frequently asked questions about the technology, and contains a useful glossary of terms for future reference.


Inspection of printed circuit boards (PCBs)

X-ray inspection by computed tomography (CT) allows for the quality evaluation of printed circuit boards (PCBs). This is a routine method to inspect and quantify the presence of voids, cracks, broken connections, bridges, and typical PCB flaws. In this application note, we showcase the electronics inspection capabilities of CT using Dragonfly software.

Industrial NDT

Industrial inspection

This application note illustrates the most popular applications for the industrial inspection of metal castings, injection-molded parts, concrete, or any manufactured part requiring quality control.

Battery Inspection for Overhang Distances

X-ray inspection by computed tomography (CT) allows for the quality evaluation of batteries and is especially useful for lithium-ion batteries. CT provides easy checking of voids, cracks, delaminations, and overhang distances between anode and cathode components. In this application note, we showcase the battery inspection capabilities of CT using Dragonfly software. In particular, a deep learning model is trained to segment the overhangs automatically.

Turbine blade segmentation using deep learning

The industrial inspection of turbine blades plays an important role in power generation and aerospace industries, either for manufacturing quality control or for evaluating wear over time. Failure of a turbine blade during operation can lead to extensive damage and potential loss of life, making their inspection of vital importance. However, turbine blades are challenging for CT-based inspection due to the density and geometry involved, which can result in image artifacts and make quantitative evaluation difficult using traditional image analysis tools.

Crack detection and evaluation

Crack detection using non-destructive X-ray computed tomography (CT) is one of the key capabilities that make the technique useful for industrial applications. The presence of cracks or planar flaws, such as delaminations in engineering components, are a major problem for the continued performance of critical systems in the energy, aerospace, and other industries. Cracks in structural components can lead to system failures, downtime, high repair costs, and even loss of life. Many NDT tools are available for crack detection, and CT is increasingly used for this purpose due to its high sensitivity. It works on all material types and allows full characterization of the crack location and extent in the sample, providing spatial context and quantitative information.

Additive manufacturing

Quality control of an additive manufacturing artifact

Additive manufacturing (AM) is transforming the way supply components for various industries, including aerospace and medical, are manufactured. With new technologies come new inspection challenges, and CT is often used for this application due to its high sensitivity to detect typical AM defects and flaws. In this application note we demonstrate a metal AM test part containing some of the typical defects of interest for such inspections: cracks, trapped powders, and porosities.

Additive manufactoring powder characterization

Additive manufacturing using metal powder feedstock is the most widely used AM method in industry, and it is now widely recognized that the powder quality is key to high quality manufacturing. There are many ways to test and characterize powders – all with some limits and assumptions. X-ray computed tomography is increasingly used due to its ability to provide full 3D morphological information in addition to simple identification of contamination or excessive irregularities that may be missed in other methods. This application note shows some of the capabilities of Dragonfly for this sample type.

CAD deviation of AM bracket

Additive manufacturing (AM) allows unique complex geometries to be produced, but sometimes these do not perfectly follow the design intent. These deviations could be due to warping because of the high temperatures and residual stresses induced in the material, or they could be due to dimensional errors in the AM toolpath/scanning hardware, or due to slicing errors of the CAD model. Other unexpected errors may also occur which necessitate geometry evaluation compared to the design intent – commonly referred to as CAD deviation, nominal-actual comparison, or 3D deviation maps.

Porosity quantification for Additive Manufacturing

Porosity in additive manufacturing is well known to occur widely in the form of very small pores, usually evenly spaced across the part. Local concentrations or clusters of pores can occur, and larger pores can occur due to a variety of reasons. Despite the huge potential of X-ray tomography for this application, porosity quantification is not often performed correctly or reliably. This application note demonstrates simple and reliable segmentation and quantification of porosity in a 10 mm cube of AlSi10Mg manufactured by laser powder bed fusion (L-PBF).