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.
Electronics
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.
Fast Porosity Analysis of Castings
X-ray computed tomography (CT) allows the detailed quality evaluation of castings. Of particular interest is porosity and its distribution and extent within the castings in three dimensions. Since this is often done as a spot check for foundries, high throughput is needed, and hence a simple and fast 3D image processing method is important here. In this application note, we showcase three simple (and fast!) methods in Dragonfly 3D World, all of which can be automated by macro’s. They are demonstrated on three different castings representative of typical castings, with varying porosity amounts. The time for the analysis by macro is shown in each case as well. The method can be applied identically to any other sample types with closed (internal) porosity
Additive manufacturing
New! |
Dragonfly Basic Automated Porosity Analysis For Additively Manufactured Parts
This application note shows some simple cases where porosity analysis is possible using automated segmentation tools in Dragonfly 3D World. In particular, these three examples all use the automated porosity – advanced Otsu method. This method is applicable in cases of simple single-material CT scans – the threshold is then applied using an automated Otsu method, based on all grey values of the part and the internal pores, excluding external air space. This often makes a simple, accurate segmentation of most porosity types, including that found in additive manufacturing. The segmentation is then used to calculate pore sizes and much more.
Dragonfly's AI-powered Solution for Microporosity Analysis in Additive Manufacturing
In this application note, the focus is on small additive manufacturing porosity and challenges in detecting them in CT images. Often CT scans are not perfect, resulting in varying bright and dark areas in the material. This makes it hard to segment small porosities correctly or even at all with traditional image segmentation methods. Here we show with two different examples how AI segmentation can solve this problem.
Deviation Analysis using Dragonfly 3D World
The focus of this application note is deviation analysis – also known as nominal-actual comparison, part to part deviation analysis or 3D comparison. It allows to visualize deviations between two parts that are well aligned in 3D space, using color coding to highlight the size of the deviation. In Dragonfly 3D World this is achieved using a function called “signed deviation map” and is applied to mesh representations of the “before” and “after” datasets. Two examples are used to demonstrate this: a lattice structure sample that was hit multiple times with a hammer to induce damage (part to part comparison), and a bracket that warping during manufacturing compared to its design intent (part to design comparison).
Additively Manufactured Lattice Structure Characterization using Dragonfly 3D World
In this application note, a full detailed analysis of an additively manufactured lattice structure is demonstrated. From basic measurements of material/air volume fractions to advanced measurements such as strut thickness, microporosity in the material, maximal spheres of macroporosity, statistical analysis and more.
Local Wall Thickness Analysis on a Single Additively Manufactured Lattice Strut
This application note uses the example of a high-resolution X-ray CT scan of a single lattice strut, to demonstrate local wall thickness analysis using two different methods in Dragonfly 3D World. Wall thickness analysis is useful to understand the local thickness of manufactured parts in general, but additively manufactured parts often include surface-attached particles that can hide the underlying thickness by measuring small local thickness values of the particles. This application note shows both methods and how they can be applied to gain different types of information.
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.