Radiographic testing of 3D-printed thermoplastics using Am-241 as a gamma-ray source

The industrial production of 3D printing is known as additive manufacturing (AM), in which a computer controls the process of producing 3D objects. Although x-ray computed radiography (XCT) is extensively used in the quality control and testing of additive manufacturing products, the gamma-ray radiography capabilities for these applications still need to be investigated. This study aimed to evaluate the performance of gamma-ray radiography using americium-241 (Am-241) as the gamma source. Here, we inspected fused deposition three-dimensional (3D) modeling products produced from polylactic acid (PLA) as thermoplastic samples. Radiographic testing of 3D-printed thermoplastic samples was performed using Monte Carlo simulations and validated by experimental studies. We used Am-241 (gamma-ray source) to conduct simulations and experiments investigations; two simulations were used: one by using 59.6 keV energy of gamma-ray and the other using all gamma-ray energies, including 16.96 keV, 26.3446 keV% 2.31 up to 662.40 keV. Also, we performed the x-ray radiography test to be used as a standard. The results showed that the defect detectability in the 3D-printed PLA samples using Am-241 as a gamma-ray source is comparable to that of x-ray results. This study concluded that the Am-241 could be used as the gamma-ray source to perform the radiography test for the products produced by 3D-printed thermoplastics.


Introduction
Additive manufacturing (AM) is a revolutionary and advanced manufacturing process. AM is classified into seven basic categories: binder jetting, directed energy deposition, material extrusion, material jetting, powder bed fusion, sheet lamination, and vat photopolymerization [1]. AM, which uses various materials [2][3][4][5][6], is predicted to significantly impact the industry by generating high-strength structural components. 3D printing techniques have become widely used in different important research topics in the last twenty years.
Fused deposition three-dimensional modeling (FDM) belongs to the material extrusion class, in which the material is drawn through a nozzle from a filament and deposited layer by layer to produce 3D parts. It is a clean process, sturdy manufacturing, and easy to operate with excellent repeatability and accuracy [7]. The system consists of an extrusion nozzle, a control system, and a build platform. Although various materials exceeding a hundred have been used in the FDM, processing some materials, such as thermoplastics, can induce a wide range of defects. These defects can be stress whitening, cavitation, or porosity, which can adversely affect the reliability Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. of the final product [8]. According to x-ray microtomography, FDM leads to a significant alteration in acrylonitrile butadiene styrene (ABS) filament geometry [9]. Therefore, the industry uses various nondestructive testing (NDT) methods to detect material defects. For example, although many methods exist to monitor pipelines, the NDT technique is deemed attractive for detecting corrosion inside pipelines [10]. The assurance of the NDT method's reliability in testing AM products has drawn the attention of scholars and industrial communities because these methods have been used for a different production approach, which is subtractive manufacturing [11]. NDT based on digital radiography performed after part production could aid in detecting non-acceptable flaws and minimize expenditures associated with the post-processing of nonfunctional components.
One of the widely used NDT methods for AM products is radiographic testing, in which x-ray computed tomography (XCT) has been used in quality control because it provides a layer-by-layer quantitative visualization. Its results are input to the raw material procedure, manufacturing parameter sets, inclusions analysis, and controlled porosity development [12]. The XCT technology is ideal for studying the outcome of AM [13]. In the last period, XCT has become the NDT method used in measuring dimensions for geometrical specifications and is commercially available [14]. The usage of XCT for AM is very important in advanced manufacturing. It was used to determine the orthotropic properties of thermoplastics reinforced by glass fiber [15].
There are two requirements to measure the AM porosity; increase the resolution in detecting small pores to get higher accuracy and pore distribution to more cost-effective measurement techniques [16]. XCT is timedemanding and only works with small amounts of feedstock powder, usually a few milligrams. For varied pore and particle dimensions, 3D digital radiographs were simulated to estimate the visibility of pores inside metallic particles; a validation experiment was carried out [17]; it was shown numerically and experimentally that typical gas pores in metallic microparticles larger than a particular size (here: 3 to 4.4 μm for the selected x-ray setup) could be successfully recognized by digital radiographs [17]. About twenty years ago, gamma radiography was developed [18] as an addition to x-ray radiography to offer a method of evaluating welded assemblies, castings, and other engineering structures in metal material to check the internal defects like cracks and blowholes by iridium-192; also, the experiences have shown that the radiation from radon or radium has high penetrating, making it particularly ideal for evaluating and examine thick casting of copper alloys or steel [19]. As well as, gamma-ray emitted from the Co-60 source was used to measure the pit depth on large-diameter pipes; the experiments showed that the radiography using the Co-60 could detect external and internal pits also; it has the ability to measure their depths on corroded large-diameter non-insulated carbon-steel and stainless steel pipes and insulated [20].
Although a recent study [21] has found that both x-ray and gamma radiation at the same dose possesses the same effects on the thermoplastics, the x-ray energies and equipment size limit its application to dense materials, thick products, and complex geometries, besides the needing for a power supply, which is also a shortcoming. Therefore, gamma-radiography of 3D-printed thermoplastics could be introduced as an alternative to XCT. Gamma-radiography can provide energy efficiency advantages since it does not require an electrical supply, flexibility to use with complex geometries, lower system cost, and higher photon energy that will be suitable for testing 3D-printed thermoplastics. Therefore, in this work, we used americium-241 (Am-241) for the first time as a gamma source for radiographic testing of 3D-printed thermoplastics. More details are included in the following sections.

Methodology details
2.1. Experiment study 2.1.1. Materials This work used a polylactic acid (PLA) filament, SainSmart PRO-3 (SainSmart, USA), as a thermoplastic polymer. Compared to the other thermoplastics, PLA was used in this study because it has specific features, such as being environmentally friendly and used for various purposes, from food packaging to medical implants [22,23]. We used PLA due to its attractive properties, such as physical, mechanical, biocompatibility, and biodegradability features [24]. Besides, it is appropriate for constructing phantoms, which can be used for radiation dosimetry [25]. Because of PLA's suitability as a substitute for soft human tissue, PLA was used in radiation applications and therapy [23,26].
The PLAs' filament was printed extremely smoothly and is supported by technology layer by layer PRO-3 without 100% tangle. Each inch of filament is precisely manufactured to an error rate of ±0.02 mm. The filament was biodegradable and non-toxic, as it was made from a natural renewable resource.

Design and manufacturing of specimens
An FDM 3D printer Ender-3 (Creality 3D company, China) with a nozzle diameter standard 0.4 mm printing size of 220 × 220 × 250 mm was used to produce the two PLA samples described in the experimental study.

Sampling and irradiation
We prepared two samples, 5 and 15 mm thick, and each one had a length as well as a width of 100 mm. Each sample contained eight artificially induced wire-like defects having diameters of 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm. Figure 1 shows the front view of the samples with 5 and 15 mm thicknesses, respectively, designed by SOLIDWORKS version 22 (Dassault Systèmes, France). Figure 2 represents the top view of the 15 mm thickness sample with the same length and width of 100 mm; both samples with thicknesses 5 and 15 mm have the same top view. The artificially induced wire-like defects were used to evaluate the detectability of flaws in the additively manufactured thermoplastics.
The tested samples were radiographed with an Am-241 gamma-ray emitter with an energy of 59 keV (J L Shepherd and Associates, USA) and an x-ray machine. The imaging receptor was an imaging plate used in medical applications. The software used for processing and analyzing scientific images obtained by radiography was the ImageJ software.

Simulation study
The parameters were defined for the simulation study: two 3D FDM product samples, a phosphor imaging plate (35 cm × 35 cm), a directional x-ray source with an energy of 40 kVp, and a gamma-ray source with an energy of 59 keV. The material was polylactic acid (PLA) with the composition of (C 3 H 4 O 2 ). Figure 3 shows the specimens and simulation setup.

Simulation results
The simulation study used two different sources, Am-241 and a 40 kVp x-ray, to radiograph two different samples with different thicknesses. For Am-241, we carried out the simulation twice, one using 59.6 keV energy of gamma-ray and the other using all gamma-ray energies, including 16.96 keV, 26.3446 keV% 2.31 up to 662.40 keV. Figure 4(a) shows the simulation results of gamma-ray radiography from Am-241 for the sample of 5 mm thickness with a length and width of 100 mm of each using 59.6 keV energy gamma-ray from Am-241 source; all the eight defects with the diameters of 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm were clearly detected in the simulation of gamma-ray radiography. At the same time, figure 4(b) represents the simulation results of gammaray radiography from Am-241 for the sample of 15 mm thickness. This figure shows seven defects, which were detected clearly by gamma-ray radiography, whereas the smallest defect with a diameter of 0.2 mm was not detected. By comparing the results from figures 4(a) and (b), it can be observed that the detection ability decreases when thickness increases. Also, it can be found from both figures that the defect with the smallest diameter, i.e., 0.2 mm, was not detected at higher thickness samples, suggesting the increase in the sample thickness could deteriorate the testing sensitivity.    4(b)). These seven peaks represent seven artificial defects where the peaks were prominent. These peaks correspond to the seven defects with diameters of 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm, whereas the defect with the lowest diameter, i.e., 0.2 mm, was not seen, suggesting the smallest defects can not be detected at higher thicknesses. Figures 6 and 7 below show, respectively, the simulation and the recorded line profile of the image results of gamma-ray radiography from Am-241 by using all energies for the sample of 5 and 15 thicknesses. As shown in these figures, the defects with diameters of 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm were clearly detected. These results mean that all defects were detected even at the smallest artificial flaw (diameters 0.2) when applying the simulation using all energies gamma rays emitted from Am-241. Figure 8 shows the simulation image results obtained using x-ray radiography for samples of 5 and 15 thickness with a length and width of 100 mm each; all the eight defects with diameters of 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm, where figures 8(a) and (b) display the simulation image results of the samples of 5 and 15 thicknesses, respectively. As can be seen in this figure, similar results to the obtained results by gamma-ray radiography, where all defects were clearly detected in the sample with 5 mm in thickness ( figure 8(a)). At the   figure 9(a), we can see eight peaks that represent eight defects in the sample with 5 mm thickness. However, figure 9(b) shows only seven peaks representing seven defects in the sample with 15 mm thickness. Although the results using 40 kVp x-ray (figures 8 and 9) seem to be the same as those in figures 4 and 5, resulting from the gamma-energy of 59.6 keV, suggesting that the gamma-ray radiography from Am-24 could bring the same results as x-ray radiography; however, the results of using all gamma energies emitted from Am-24 (figures 6 and 7) seem to be better than those obtained using the gamma-ray of 59.6 keV energy ( figures 4 and 5) and using x-ray radiography of 40 keV (figures 8 and 9). Therefore, the detectability of the smallest artificial defects could be suitable and clear when all energies were incorporated into the simulation.

Experimental results
The experiments were performed using a 40 kVp x-ray and gamma-ray from Am-241 with radioactivity of 1 Ci. First, we found a gap between the results of the real experiments and simulations. The main reason for this gap could be the bottom noise. Therefore, we reduced noise by processing the images using ImageJ software to avoid the bottom noise that comes from the low resolution and accuracy that is inevitable when 3D printing PLA, which is prone to inter-layer gap or surface roughness. Figure 10 shows the recorded radiographic images using x-ray and gamma-ray from Am-241 for the sample with 5 mm thickness. Where figure 10(a) shows the recorded radiographic image obtained using the x-ray; this image clearly shows verticle lines representing the artificial  defects with diameters of 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm; the widest diameter is on the right, while the narrowest diameter is on the left. At the same time, figure 10(b) displays the recorded radiographic images of the same sample obtained using the gamma-ray from Am-241 source. From both images, it was found that the line representing the defect became more contrasted as the diameter of the defect increased. Besides, a darker image was found in the recorded radiographic image obtained using gamma-ray radiography from Am-24 source; this result could relate to the higher energy from Am-241 that enables more photons to reach the detector. From comparing the images in figures 10(a) and (b), it can be found that the gamma-ray radiography using the Am-241 source showed a radiographic image with acceptable quality comparable to that obtained with x-ray radiography. Figure 11 exhibits the recorded line profiles of the images of the sample with 5 mm thickness shown in figure 10, in which the images were recorded using both x-ray and gamma-ray radiography. The x-axis represents the distance in the pixel units, and the y-axis represents the grayscale values. Where figures 11(a) and (b) show the line profile of the recorded radiographic image using x-ray and gamma-ray, respectively; it can be seen in both figures 11(a) and (b) eight peaks, which represent the eight defects with diameters of 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, and 2.5 mm in the sample. By comparing figures 11(a) and (b), it can be observed that the gray values were higher for the recorded line profile of the image obtained using gamma-ray radiography from Am-24 source, reflecting more darkness in this image. The defects detection ability is evident in both line profiles of the recorded radiographic images obtained using x-ray and gamma-ray radiography. The peak with the highest intensity on the right represents the defect with the widest diameter; in contrast, the peak with the lowest intensity on the left belongs to the defect with the narrowest diameter. This result could indicate that the intensities of the peaks increase as the defect diameters increase. Also, in the recorded radiographic image gamma-ray radiography, the line profiles of the three defects with the narrowest diameters, i.e., 0.6, 0.4, and  Similarly, figure 12 exhibits the recorded radiographic images using x-ray (figure 12(a)) and gamma-ray from Am-241 ( figure 12(b)) for the sample with higher thickness, i.e., 15 mm. In both images, the eight verticle lines can be noticed; these lines represent the eight artificial defects. However, the lines in the obtained radiographic image using the x-ray seemed to be more contrasted ( figure 12(a)). The low contrast in the obtained radiographic image using gamma-ray from Am-241 could result from the darkness that comes from more gamma-ray photons with higher energy reaching the detector. Again in this thicker sample, it is evident that as the diameter of the defect increase, the line representing the defect becomes more contrasted.
Similarly, figure 13 displays the recorded line profiles of the 15 mm-thick sample shown in figure 13, where figures 13(a) and (b) show the line profiles of the radiographic images x-ray and gamma-ray from Am-241source, respectively. Both line profiles exhibit many peaks, representing the samples' defects; the higher peaks represent the defects with wider diameters, meaning the first peak on the right belongs to the defect with a diameter of 2.5 mm. In the recorded radiographic image using x-ray, it can be seen that the peaks' intensities for the thicker sample ( figure 10(a)) decreased compared with those of the thinner sample ( figure 12(a)). Line profiles showed some noise ( figure 13(a)), meaning good contrast. However, in the recorded radiographic image using gamma-ray from Am-241source, it was found that the detection of smaller defects was better than that of using the x-ray. This result could be due to the higher thickness enabling better interaction between the gammaray, sample, and detector. It was also observed that gamma-ray gave higher grayscale values (darker image) ( figure 13(b)), which also relates to the higher energy from Am-241 that enables more photons to reach the detector. However, the disadvantages of higher gamma-ray energy come with higher noise and lower contrast expenses that need more investigation to optimize the testing method. In the summary of figure 13, for thicker  samples, x-ray provides better contrast; on the other hand, gamma-ray radiography enables better detection sensitivity.

Conclusion
The current study's main goal was to use Am-241 as the gamma-ray source in the radiographic testing to evaluate the defects in thermoplastic products produced by FDM printers. The obtained radiographic testing using Am-241 as the gamma-ray source was compared to x-ray radiography as the standard. The environmentally friendly polylactic acid was used as a thermoplastic polymer for performing this work. The results have shown that gamma-ray and x-ray detection abilities were comparable. However, x-ray provides better contrast; on the other hand, as the sample thickness increases, gamma-ray radiography enables better detection sensitivity. This research highlights the usefulness of using Am-241 as the gamma-ray source in the radiographic testing of 3D printed products. However, there are some restricted points where the greater penetration power of the higher gamma-ray energies is not suited for the radiography of small-thickness and low-density materials. Moreover, the higher energies produce more scattered radiation that deteriorates the radiographic image definition and contrast; these two factors could govern the radiographic sensitivity of using Am-241 gamma-ray source.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).