China piping solution supplier:

Detection technology of metal additive manufacturing defects

In the process of metal additive manufacturing, the material temperature changes violently and the behavior is complex. Due to the special process characteristics, all kinds of defects, including surface and internal defects, inevitably appear in the parts, which seriously endanger the performance of the parts, and become an important factor hindering the development of metal additive manufacturing technology, which limits the application of the technology in key fields. Timely acquisition of defect information helps to adjust the process parameters The defect information can also be used to guide the subsequent processing and processing of parts. Therefore, it is of great significance to study the forming mechanism of defects and the corresponding detection technology for improving the level of metal additive manufacturing technology. In view of this hot issue, this paper first briefly combs the common types of defects in metal additive manufacturing, and summarizes the main problems Then, the representative research work in the field of metal additive manufacturing defect detection in recent years is summarized, mainly focusing on process characteristic quantity monitoring and on-line non-destructive testing, and the application scope of different technologies is briefly described. Finally, the development trend of defect detection technology is prospected. It is pointed out that the quality of metal additive manufacturing parts can be improved through detection methods At present, many detection methods in terms of accuracy and real-time have not yet met the requirements of practical use, and the defect detection technology will develop towards integration and intelligence.

Additive manufacturing is the opposite of subtractive manufacturing. It is a process of manufacturing entities by adding materials based on three-dimensional model data, usually layered. According to different molding processes, metal additive manufacturing can be roughly divided into five categories, including laminate (SL), Material injection molding (MJ), binder injection molding (BJ), powder bed melting (PBF) and direct energy deposition (DED)[1]. Among them, PBF and DED are commonly used process types.
PBF manufactures parts by selectively melting the materials within a specific range of each layer of the powder bed; DED uses metal powder or wire to melt and solidify the surface of the workpiece. Both processes need to manufacture parts layer by layer, but different Formal heat sources such as laser, electron beam, etc. Compared with material injection molding, PBF and DED are less limited by the melting point of raw materials, and high melting point metals such as stainless steel and titanium alloy can be used as raw materials; compared with adhesive injection molding, PBF The post-processing process of the parts manufactured by the DED process is relatively simple and does not require complicated processes such as soaking [2]; because the lamination process is difficult to form a metallurgical bond between the layers, compared with PBF and DED, the parts are strong in the direction of material deposition Slightly insufficient. The content of this article mainly focuses on PBF and DED. In the following, metal additive manufacturing refers to these two processes.
The energy input of PBF and DED in the manufacturing process is concentrated, the material reaction process is violent and the molten pool behavior is complex, the temperature field of the workpiece is difficult to control, and the uncertainty is large. The combined effect of many factors makes the parts inevitably appear various types of defects, serious Endangering the performance of parts, increasing the difficulty of post-processing, increasing manufacturing time and cost. The quality and consistency of metal additive manufacturing parts are considered to be Achilles’Heel[3] of this technology. In response to this problem, this article briefly combs metal additive Common defects in manufacturing and their causes, and a review of recent representative research work in the field of defect detection in additive manufacturing, aiming to provide a useful reference for the development of more efficient detection technology, and hope to reveal the possible future development of this technology trend.

Common types and causes of defects

Surface defects

Surface defects include poor surface roughness of parts, surface oxidation, spheroidization, surface cracks, etc. The liquid metal in the metal additive manufacturing bath is affected by temperature gradients, surface tension and other factors. The movement is complicated, and the surface quality obtained after solidification is generally poor. .Poor surface roughness is also related to the process principle of layered accumulation. Generally, in order to obtain parts with better surface quality, a smaller layer thickness should be selected, and the angle between the side of the part and the deposition direction should be appropriately selected [4 ]. Appropriately reducing the layer thickness is also conducive to improving the accuracy of the parts. However, choosing too small a layer thickness will seriously reduce the material deposition efficiency and increase the manufacturing time. Therefore, this parameter is often the result of the balance between the quality of the part and the manufacturing time. In addition, during the manufacturing process Splashes of metal materials may occur, further worsening the surface roughness of the parts.
Oxidation is a defect formed by the chemical reaction between high-temperature metal materials and surrounding oxygen elements. In order to prevent oxidation, a protective gas is usually applied in the metal additive manufacturing process to isolate oxygen. Some special processes such as electron beam selective melting are required Carried out in a vacuum environment. Vacuum environment can reduce oxidation, but too low environmental pressure will cause serious splashing of materials [5]. In the metal additive manufacturing process, the liquid metal in the molten pool has a tendency to shrink under the action of surface tension, so Spheroidization may appear on the surface of the part [6]. Figure 1 shows the material deposition image of high temperature nickel-based alloy under PBF process at pulse laser energy 0.5J, frequency 40Hz, and scanning speed 100mm/min. It can be clearly seen from the figure. Spheroidization phenomenon. When the scanning speed is too high, the molten pool will be elongated, and unstable splitting will occur under certain conditions, which will aggravate the spheroidization defect [7-8]. The spheroidization defect will deteriorate the surface roughness of the part. While affecting the geometric accuracy of the part, it will also affect the deposition of the next layer of material, which can easily cause the internal material of the part to be discontinuous and endanger the mechanical properties of the part.
20210202105625 47453 - Detection technology of metal additive manufacturing defects
Figure 1 The micrograph of the spheroidization phenomenon on the surface of the part [9]

Internal defects

Typical internal defects of metal additive manufacturing parts can be divided into macro defects and micro defects according to different scales. This article regards the defects visible to the naked eye as macro defects, and vice versa. Common macro defects include cracks, non-fusion, inclusions and cold. Separation defects; other defects are mostly microscopic defects, including pore defects (usually referred to as pores) and directional growth associated with the material structure, solute segregation, and missing alloy elements [10-14]. It should be noted that cracks, unfused And inclusion defects have a large size distribution range, which may be invisible to the naked eye, and in some cases can also become microscopic defects.

Macro defects

The materials obtained by metal additive manufacturing technology usually do not have better plasticity and toughness. This feature makes the parts more prone to crack defects. Internal stress is the main reason for crack initiation during the manufacturing process. Material temperature in the metal additive manufacturing process Rapid changes, rapid melting and solidification process make the parts appear large temperature gradients, while the deformation is restricted by the surrounding materials, so large internal stresses are generated in the parts. When the internal stress reaches a certain value, it will cause cracks to initiate and make the parts Microcracks appear in the process. Some microcracks will further expand in the subsequent manufacturing process to form macroscopic cracks. The cracks will expand to a certain extent and even cause the parts to crack as a whole. [15]. Reasonably design the geometric shape of the part, select better process parameters and appropriate The preheat treatment can reduce the internal stress to a certain extent, which is of great significance for reducing the crack defects in the parts.
Unfused defect refers to the defect formed by the incomplete melting and bonding of the interlayer and interpass metal in the metal additive manufacturing process. The unfused defect is generally large in size and irregular in shape. If the raw material is metal powder, the defect may be accompanied by Unmelted powder particles. Figure 2 shows the non-fusion defects of low carbon steel in the arc deposition process. The non-fusion defects directly affect the interlayer bonding force of the parts and make the parts more prone to damage. At the same time, the edges of such defects are often It is relatively sharp and easily causes stress concentration. The tip is often affected by the thermal cycle in the subsequent manufacturing process to induce cracks [16], reducing the fatigue life of the parts. Unfused defects are difficult to remove by hot isostatic pressing technology, so the manufacturing should Try to avoid such defects. The main causes of non-fusion defects include too low heat source power, too fast scanning speed, too much material supply, abnormal defocusing, and too large scanning spacing.
20210202105752 77106 - Detection technology of metal additive manufacturing defects
Figure 2 Microscopic image of unfused defect [17]
Inclusion defects refer to the inclusion of other metals such as W, Mo, Nb, Ta, etc., which are different from the matrix material in the part; or the matrix contains carbides or oxides that are different from the surrounding structure. The former can be granular. It can be seen as a brighter tissue than the matrix material under the naked eye [14]; the latter is generally crescent-shaped, often consisting of unmelted powder particles and oxides in black and white intersecting tissue [18]. Examples of micrographs of inclusion defects are shown in Figure 3 (a) and (b) respectively. The inclusion defects will change the composition and structure of the matrix material distributed around them. Under the action of the applied stress field, micropores and cracks It is easier to nucleate at the junction of inclusions and matrix material, which will affect the fatigue life of parts. This type of defect generally occurs in contaminated metal raw materials. Unreasonable selection of manufacturing process parameters such as heat source energy density and shielding gas flow rate can also cause it to occur [18].
20210202111259 74024 - Detection technology of metal additive manufacturing defects
Figure 3 Microscopic image of inclusion defect
The cold barrier defect is the irregular shape of lack of meat in the metal additive manufacturing parts, which belongs to the type of body defect of the material. The size of the cold barrier defect is generally large, most areas have smooth surfaces, and some areas have uneven surfaces [13]. The root cause of the formation of barrier defects is that the feeding cannot be fully carried out during the molding process of the parts. If there are small-sized oxide particles in a local area, the flow and convergence of the molten pool liquid will be further hindered, increasing the possibility of cold barrier defects. Additive manufacturing The direct cause of the formation of cold barrier defects in the process may be insufficient heat source power or unreasonable printing parameter settings, resulting in insufficient liquid metal fluidity; or raw materials are contaminated and liquid metals are difficult to merge. Cold barrier defects can also cause stress concentration and promote cracks Initiation, which affects the fatigue performance of parts.

Micro defects

Hole defects are one of the most common microscopic defects. Their shape is mainly spherical or ellipsoidal, with a size generally <100 μm [11], and sometimes tiny holes with a size of <40 μm are distributed throughout the part [20]. Figure 4 shows the microscopic image of holes in Ti-6Al-4V parts obtained by the electron beam selective melting process. The main cause of hole defects is the large cooling rate of materials in the metal additive manufacturing process, and various types of liquid metals Gas cannot escape from the molten pool in time. These gases stay in the parts and are surrounded by solid metal to form holes. The main sources of gas include: 1) The raw material contains moisture. When the metal is melted, the moisture generates a lot of gas and enters the molten pool; 2) The raw material itself Hollow and wrapped with a small amount of gas, these gases are transported to the molten pool with the raw materials; or the protective gas enters the molten pool to form holes; 3) The power of the heat source is too high or the scanning speed is too slow to cause the low melting point alloy elements to evaporate and form pores. In addition, when When the heat source power is too high, the surface of the molten pool will form a deeper depression under the action of the recoil formed by the evaporation of the metal, which increases the possibility of holes [21-22]. Some hole defects in the parts can be treated by hot isostatic pressing Elimination, but some subsequent heat treatments will cause the holes to grow again [23]. The hole defects have a negative effect on the tensile properties of the material, but due to the small size and relatively regular shape, the fatigue strength of the parts is relatively affected. Smaller [11, 24].

20210202111428 51748 - Detection technology of metal additive manufacturing defects
Figure 4 Microscopic image of hole defect [25]
The metal additive manufacturing process involves the rapid melting and solidification of the material. Therefore, generally, the higher cooling rate of the non-equilibrium structure can refine the grain to a certain extent and improve the yield strength, but the plasticity and toughness of the material are lower than that of the forging. The possibility of brittle fracture is greater [26]. Due to the different thermal processes of different parts of the part, the microstructure of the material is also different [27]. In addition, affected by the process characteristics, the material structure between layers and passes of additive manufacturing parts may appear The phenomenon of periodic changes [27], the mechanical properties may also show anisotropy [28].
In the additive manufacturing process of alloy material parts, due to the higher temperature of the molten pool, the metal will evaporate to a certain extent. However, the melting point of different element metals is different, resulting in the difference between the evaporated metal composition and the alloy composition, which leads to the alloy of the entire part. Deviations in composition, this type of defect is called the lack of alloying elements [29]. Changes in the content of alloying elements will affect the microstructure, mechanical properties and corrosion resistance of the material. In addition, directional growth and solute segregation appear in metal additive manufacturing parts. The possibility of such defects is also relatively high, and these can be regarded as defects in the structure of the material.

Defect detection technology

In order to ensure the quality of additively manufactured parts, it is necessary to detect defects and control them within the allowable range. Since surface defects can be removed by appropriate surface treatment and cutting processing, the detection and evaluation of internal defects is more important. The commonly used inspection methods in material manufacturing which can be divided into online inspection during manufacturing and non-destructive inspection after manufacturing. Timely detection of defects during the manufacturing process helps to adjust the manufacturing process and take certain measures (such as remelting) to remove Existing defects; or directly stop manufacturing to reduce material and time loss. Therefore, compared with post-inspection, efficient online inspection is of greater significance for improving the technical level of additive manufacturing. For online inspection, commonly used in published literature The methods can be divided into two categories, including: 1) Monitoring the additive manufacturing process, using process characteristics to monitor the possible conditions of defects, and then indirectly reflecting the stability of the defects and the manufacturing process; 2) Applying non-destructive testing technology to additive manufacturing In the process, existing defects are directly detected in the manufacturing process.

Afterwards non-destructive testing technology

The particularity of metal additive manufacturing technology puts forward the following requirements for non-destructive testing technology: low cost, rapid detection, adaptability to complex geometric structures and poor surface quality, and detection of multiple types of defects [30]. Currently, many applications in the industrial field Non-destructive testing methods include: vision, liquid penetration, ultrasound, eddy current, radioactive imaging, metal magnetic memory and permanent magnetic disturbance detection, etc. Archimedes drainage method is often used in research to measure the density of parts, and quantitatively through density Characterization of defects [31]. The advantage of post-testing is that the test results reflect the final quality of the part, which helps to ensure the reliability of the part in use. Post-testing can also be used to establish the relationship between process conditions and part quality. For example, Cunningham et al. [23] used synchrotron-based X-ray microtomography (μSXCT) to study the influence of powder and post-processing technology on hole defects. However, post-mortem non-destructive testing is not real-time, so testing results cannot guide the manufacturing process The real-time adjustment and the timely removal of defects will have a limited effect on the improvement of part quality and yield.

Online detection technology

Feature monitoring

Real-time monitoring of the additive manufacturing process is helpful to discover the instability in the process in time and adjust the corresponding parameters appropriately. The monitoring of the manufacturing process can be carried out by detecting process characteristic quantities [32]. These characteristic quantities can generally reflect the material behavior. And various instability phenomena. Instability is usually a prerequisite for the occurrence of defects, so it has a certain predictive effect on the occurrence of defects. Such methods generally have good real-time performance and are convenient to realize closed-loop control of the additive manufacturing process. At present, the most commonly used characteristic quantities are the parameters associated with the molten pool, such as the size of the molten pool, temperature, spectrum, etc. In addition, relevant information about the plasma near the molten pool is often involved. Process monitoring systems can generally be divided into two categories: coaxial Monitoring systems and off-axis monitoring systems usually require high sampling frequency for coaxial monitoring [33].
Berumen et al. [34] used a high-speed camera to detect the size of the molten pool and a photodiode to detect the average radiation of the molten pool, and established a coaxial monitoring system for the laser powder bed melting process. Research results It shows that the use of this system can effectively monitor the deviation information of the molten pool during the manufacturing process, and the measurement results can be used as feedback to control the manufacturing process. Clijsters et al. [35] used a similar system to detect the parameters of the molten pool. It was determined by experiments The 95% confidence interval of the weld pool parameters is established, and the area outside the confidence interval of the parameters is used as the area where the overheating phenomenon exists to judge the manufacturing quality. The research results prove that the hole defect has a higher frequency of occurrence at the edge of the part This is consistent with the research results of Choo et al. [36]. The possible reason for this phenomenon is that the laser at the edge has undergone a process of deceleration and then acceleration, which causes the material to be irradiated by the laser for too long, which leads to defects. Craeghs et al. [37] used this method to study the influence of the surrounding environment of the scanning path, the cantilever structure and the acute angle structure on the size of the molten pool and the manufacturing process. The results proved that the size of the molten pool increased at the edge, the cantilever structure and the acute angle and the parts Quality causes certain harm.
Cryptomeria fortunei [38] uses CCD sensors to establish a visual inspection system for monitoring the laser cladding molten pool. It extracts various parameters of the molten pool from the image, such as geometric shape, grayscale distribution, etc., and then performs manufacturing according to the state of the molten pool. State judgment. Chen Dianbing [39] built a vision-based molten pool detection system with the help of CMOS sensors. It mainly studied the relationship and law between the characteristics of the molten pool under different powder materials, the quality of parts and the changes of process parameters. Stockman et al. [ 40] The molten pool is monitored by a two-color pyrometer, and the characteristic information of the molten pool is extracted from the output signal through an algorithm, including the position offset of the molten pool, the external rectangular aspect ratio of the molten pool, the highest temperature, heating and cooling rates, etc. , Through abnormal data to identify possible defects. The xCT verification results show that this method can detect some holes with diameters exceeding 40 μm, but there are still missed inspections. Tan et al. [41] established a theoretical model of molten pool surface temperature and The actual molten pool temperature was measured with an infrared thermometer. The results show that the theoretical model and the actual measurement results are in good agreement, and the molten pool temperature can be used for real-time feedback control to improve the quality of parts.
During the metal additive manufacturing process, plasma will form above the molten pool, which will change the gas pressure above the molten pool, which will affect the nearby acoustic signals. At the same time, the plasma will affect the absorption of the laser by the material. Usually the plasma density fluctuates sharply and melts. The defects caused by the sputtering and overheating of the molten metal in the pool are closely related. The relevant information of the plasma near the molten pool can be used to predict the defects. Based on this principle, Ye et al. [42] used a microphone to collect the acoustic signals near the molten pool. And use the deep belief network to judge whether the defect appears. This method is only used to judge the defect related to spheroidization and overheating. Chen et al. [43] monitor the manufacturing quality by detecting the spectrum of the plasma. They study the intensity and intensity of the spectrum. The relationship between process parameters and the relative intensity of specific wavelengths of light to judge manufacturing quality.
Monitoring the process characteristics can predict part defects to a certain extent, and facilitate real-time feedback and closed-loop control to stabilize the manufacturing process. However, due to factors such as thermal stress, defects are most likely to occur during material cooling [44]. The basic process of metal additive manufacturing, especially the material structure evolution process and defect generation mechanism has not been fully revealed. There is no definite one-to-one mapping between process feature quantities and defects. Therefore, it is difficult to judge part defects only by process monitoring. It is completely convincing. It is necessary to use the detection technology that can directly reflect the defect information to detect the existing defects on the workpiece in the process, and complement the feature quantity monitoring to improve the online detection accuracy of the defect as much as possible.

Online non-destructive testing

The layer-by-layer overlay of metal additive manufacturing materials makes it possible to implement online non-destructive testing. Appropriate testing methods are used to evaluate the quality of the workpiece in the process to help find defects as soon as possible and avoid unnecessary losses. At present, online non-destructive testing The main idea is to apply non-destructive testing to the printing process, to inspect the workpiece after each layer or several layers of material is deposited, and to ensure the quality of the parts through layer-by-layer printing and layer-by-layer inspection. And indirectly reflect defects through feature quantities Different, online non-destructive testing technology generally requires corresponding input, using the interaction between input and defect to directly reflect defect information. Common online non-destructive testing technology includes online ultrasonic testing, eddy current testing, temperature field-based testing and visual inspection Wait.
The propagation of ultrasonic in the tested part will be affected by defects, so it can reflect the defect information in the part. This method is mainly used to detect cracks, unfused and other physical defects. Traditional ultrasonic testing needs to input energy into the workpiece through couplant Restricted by factors such as part temperature, the application of this technology in real-time detection of defects in additive manufacturing is limited. Lopez et al. [17] used ultrasonic methods to detect defects in parts from the outside of the substrate, and the detection results were compared with X-ray and Liquid penetration detection can be a good agreement. Rieder et al. [45] used ultrasonic sensors to monitor the dynamic process of additive manufacturing from the outside of the substrate, but failed to infer the defect information in the part through the response signal.
In addition to traditional ultrasonic testing methods, many scholars have conducted research on laser ultrasonic testing methods. When the laser irradiates the surface of the part, part of the energy is transferred to the part in the form of heat. The thermal elastic effect or thermal erosion effect makes the part local area The stress field and strain field are generated, and then stress waves are generated on the surface and inside of the part [46]. Because the thermal corrosion effect will cause certain damage to the surface of the part, the detection is mostly achieved within the range of thermoelastic effect. Millon et al. [47] The laser ultrasonic detection method of crack defects is studied. It uses pulsed laser (pulse time 7ns) to generate ultrasonic waves in the part, uses the detection laser (interferometer) to detect the surface wave of the workpiece, and detects the part by B-scans. Surface defects. The laser ultrasonic inspection results of the 316L stainless steel strip defect specimens obtained by the DED process are shown in Figure 5. The defect location can be clearly seen from the figure. The research results prove that the depth of the detection by the laser ultrasonic inspection means is 0.5 mm and 0.1 mm with a width of 0.05 mm.
20210202112239 41913 - Detection technology of metal additive manufacturing defects
Figure 5 Laser ultrasound B scan results of different depth defects [47]
Everton et al. [48] tried to use laser ultrasound to detect irregular defects without processing the surface of the part. The experiment found that the interference noise was large and it was difficult to accurately determine the defect. Cerniglia et al. [49] used the laser ultrasonic method to detect the defect in the part. Sub-surface defects, and established a thermoelastic finite element model. The research shows that the measurement results are in good agreement with the finite element simulation. This method can detect small sub-surface defects. The laser ultrasonic method has great potential in real-time defect detection, but the above research All of them are still in the stage of principle verification, and online inspections have not been truly realized, and most of the workpiece surfaces need to be processed to eliminate the influence of surface roughness. Unlike the inspections carried out in the above-mentioned thermoelastic effect range, Levesque et al. [50] The laser ultrasonic inspection test is carried out from the outside of the substrate within the scope of the corrosion effect, the detection signal is processed by synthetic aperture focusing technology, and the X-ray inspection technology is used for verification. The results show that this method can detect some holes in the parts and defects such as unfused .
Eddy current testing is a testing method based on the principle of electromagnetic induction. Eddy current testing technology can be used in harsh environments and meets some of the requirements of additive manufacturing for online testing. It is suitable for detecting defects such as cracks and non-fusion. Du etc. [ 51] studied the eddy current detection technology for the defects of the composite machining parts of the increase and decrease of materials. The basic idea is to mill the plane after the deposition of each layer of material, remove the interference of the surface roughness, and use the eddy current method to carry out this layer. Detection. In order to verify this principle, it established a corresponding finite element model based on ANSYS, and studied the influence of factors such as excitation frequency, temperature, and lift-off height on the detection results. Defects were detected through experiments and X-ray detection technology was used. The measurement results are verified. The test results of the Ti-6Al-4V tape unfused defect samples obtained by the direct laser deposition process are shown in Figure 6. It can be seen from Figure 8 that the method can effectively detect the defects in the parts. On this basis, Wang et al. [52] studied the edge effect of eddy current testing. They drilled holes on the edge of the workpiece to simulate edge subsurface defects, and used eddy current sensors to scan the sample with different scanning paths. The analysis and testing results found that, Choosing a suitable scanning path can overcome the influence of edge effect to a certain extent. This research is helpful to expand the application of eddy current method in quality inspection of complex structure parts.
20210202112416 76582 - Detection technology of metal additive manufacturing defects
Figure 6 Test results of eddy current testing [51]
Todorov et al. [53] used the array eddy current sensor to detect the workpiece after the deposition of each layer of material, and realized the imaging function. The resolution of the detection image in the scanning direction of the probe is 0.1mm, and the resolution in the direction of the coil array Affected by the size and arrangement of the coil, it is 0.826 mm. It detects unfused defects in different directions, and the results are verified by X-ray inspection. Research shows that this technology can successfully detect unfused defects in different directions .
Although the eddy current detection technology can be used in harsh environments such as high temperature, because temperature has a great influence on the electromagnetic properties of materials, and additively manufactured parts often have complex temperature fields, it is still relatively difficult to accurately detect component defects through eddy current. Similar to laser ultrasonic testing, eddy current testing has certain requirements on the surface quality of parts. These characteristics bring challenges to the online use of this technology. Design a reasonable probe form and develop corresponding signal processing methods to overcome temperature and surface roughness for testing The resulting interference is urgently needed work.
Defects will affect the heat conduction in the part, and then affect the temperature field in the part. Use a certain method to input heat into the part and detect the temperature field of the part through various sensors to obtain defect information. Heat can directly come from the additive manufacturing process The heat source can also be input through laser and eddy current methods after the material is deposited. The types of defects that can be detected by this method are roughly the same as those of the eddy current method. Montinaro et al. [54] used a thermal imager to detect defects. Its Use a mobile laser as a heat source to heat the workpiece, use an infrared thermometer to measure the temperature, and use the average temperature and temperature standard deviation in the region of interest (ROI) as indicators to achieve defect detection. In order to verify the effectiveness of the method, it has established a limited Meta-model and experiments were carried out. The simulation and actual measurement results of the average temperature in the ROI of the Inconel600 single-pass cladding sample with defects according to the position of the laser thermal imaging method are shown in Figure 7. The results in the figure show that the method can be used to measure the defects Effective detection; the difference between simulation and test results can be explained by the different thermal initial conditions of the sample.
20210202112741 44001 - Detection technology of metal additive manufacturing defects
Figure 7 Simulation and test results of laser thermal imaging method [54]
Based on the abnormal increase in the surface temperature of the defect part, Xie et al. [55] used a two-color pyrometer to detect the temperature field of the part, and studied the quantitative relationship between the size of the defect and the temperature field. The test results show that the method can detect a width of 37.2 μm The defects of cracks and holes with a diameter of 88.6 μm. Schwerdtfeger et al. [56] performed infrared photography on each layer in the electron beam selective melting process, used radiation intensity to determine the location of the defect, and obtained the actual cross-section of the part to verify the inspection results Tests have proved that the infrared image has a good correspondence with the actual defects in the part, and the radiation intensity at the defect is higher. Bartlett et al. [16] used a long-wave infrared camera to photograph each layer in the manufacturing, and the temperature and the layer were averaged The area with a temperature difference of more than 1% is regarded as a defect area, and the defects are detected in real time during the manufacturing process. Experiments show that this method can detect 82% of the unfused defects in the part, and can detect all unfused defects with a size of more than 500 μm , But the detection rate of micro-hole defects is only 33%. For the DED process, Barua et al. [57] replaced the infrared camera with a SLR, characterized the surface temperature of the part by calibrating the RGB value in the photo, and used the vision system to detect the part defect.
In addition to the above studies, many scholars have explored other online non-destructive testing principles. Abdelrahman et al. [58] took pictures of each layer in the PBF process under 5 different lighting conditions and selected the ROI in the picture with the help of image processing technology .By calculating the relevant parameters and comparing the two adjacent layers of pictures to determine the possible defects and use CT to verify the detection results. Tests have shown that this method can determine the existence of defects to a certain extent, but the accuracy is poor and can only be judged Macroscopic body defects across multiple layers. Yao et al. [59] used multifractal theory to study defect images and explored the characteristics of different defect images. Grasso et al. [60] used CMOS sensors to capture the manufacturing process and used statistics The method processes the acquired images to detect defects caused by overheating.


Various defects produced in the metal additive manufacturing process seriously affect the performance and reliability of parts, and hinder the application of additive manufacturing in key areas. Research on defects and their detection technology is of great significance for improving the level of metal additive manufacturing technology. The specific conclusions are as follows :

  • 1) The metal additive manufacturing process is special, the defects are difficult to avoid, and the variety, size and distribution are complex, which brings challenges to the inspection. In order to accurately detect the defects, it is necessary to further study the material structure evolution process and the defect generation mechanism to deepen the defect detection. Recognition.
  • 2) The monitoring process feature quantity has a certain predictive effect on the occurrence of defects, and the real-time performance is good, which is convenient to realize closed-loop control. However, because there is no definite one-to-one mapping between the defect and the process feature quantity, the part cannot be guaranteed by this method alone. The final quality needs to be complementary to online non-destructive testing and post-mortem non-destructive testing techniques.
  • 3) The existing online non-destructive testing technology does not meet the actual use requirements in terms of detectable defect types, detection accuracy and efficiency, real-time performance and robustness. It is expected that the technology will change from offline principle verification to actual online application. The transition from a single detection principle to multi-principle multi-sensor integration. The development of artificial intelligence also provides an opportunity for the intelligence of detection technology.
  • 4) In future research, defect detection, online defect removal, and real-time adjustment of manufacturing parameters can be combined to improve the quality of metal additive manufacturing parts.

Author: GUO Zhengya, XIONG Zhenhua

Source: China Metal Flanges Manufacturer – Yaang Pipe Industry (

(Yaang Pipe Industry is a leading manufacturer and supplier of nickel alloy and stainless steel products, including Super Duplex Stainless Steel Flanges, Stainless Steel Flanges, Stainless Steel Pipe Fittings, Stainless Steel Pipe. Yaang products are widely used in Shipbuilding, Nuclear power, Marine engineering, Petroleum, Chemical, Mining, Sewage treatment, Natural gas and Pressure vessels and other industries.)

If you want to have more information about the article or you want to share your opinion with us, contact us at [email protected]


  • [1] CHUA Z Y, AHN I H, MOON S K. Process monitoring and inspection systems in metal additive manufacturing:Status and applications[J]. International Journal of Precision Engineering and Manufacturing-Green Technology, 2017, 4(2): 236. DOI:10.1007/s40684-017-0029-7
  • [2] GIBSON I, ROSEN D, STUCKER B. Additive manufacturing technologies[M]. New York: Springer, 2015: 206. DOI:10.1007/978-1-4939-2113-3
  • [3] TAPIA G, ELWANY A. A review on process monitoring and control in metal-based additive manufacturing[J]. Journal of Manufacturing Science and Engineering, 2014, 136(6): 060801. DOI:10.1115/1.4028540
  • [4] RAHMATI S, VAHABLI E. Evaluation of analytical modeling for improvement of surface roughness of FDM test part using measurement results[J]. The International Journal of Advanced Manufacturing Technology, 2015, 79(5-8): 823. DOI:10.1007/s00170-015-6879-7
  • [5] GUO Qilin, ZHAO Cang, ESCANO, et al. Transient dynamics of powder spattering in laser powder bed fusion additive manufacturing process revealed by in-situ high-speed high-energy x-ray imaging[J]. Acta Materialia, 2018, 151: 171. DOI:10.1016/j.actamat.2018.03.036
  • [6] LI Ruidi, LIU Jinhui, SHI Yusheng, et al. Balling behavior of stainless steel and nickel powder during selective laser melting process[J]. The International Journal of Advanced Manufacturing Technology, 2012, 59(9-12): 1025. DOI:10.1007/s00170-011-3566-1
  • [7] KRUTH J P, LEVY G, KLOCKE F, et al. Consolidation phenomena in laser and powder-bed based layered manufacturing[J]. CIRP Annals, 2007, 56(2): 733. DOI:10.1016/j.cirp.2007.10.004
  • [8] GUSAROV A V, SMUROV I. Modeling the interaction of laser radiation with powder bed at selective laser melting[J]. Physics Procedia, 2010(5): 393.
  • [9] MUMTAZ K, HOPKINSON N. Top surface and side roughness of Inconel 625 parts processed using selective laser melting[J]. Rapid Prototyping Journal, 2009, 15(2): 101. DOI:10.1108/13552540910943397
  • [10] EVERTON S K, HIRSCH M, STRAVROULAKIS P, et al. Review of in-situ process monitoring and in-situ metrology for metal additive manufacturing[J]. Materials&Design, 2016, 95: 431. DOI:10.1016/j.matdes.2016.01.099
  • [11] ZHANG Bi, LI Yongtao, BAI Qian. Defect formation mechanisms in selective laser melting:A review[J]. Chinese Journal of Mechanical Engineering, 2017, 30(3): 515. DOI:10.1007/s10033-017-0121-5
  • [12] MALEKIPOUR E, EI-MOUNAYRI H. Defects, process parameters and signatures for online monitoring and control in powder-based additive manufacturing[C]//Conference Proceedings of the Society for Experimental Mechanics Series. Cham: Springer, 2018: 83
  • [13] WEN Yi. Research on microstructure and defects of two-phase titanium alloy with 3D printing[D]. Nanchang: Nanchang Hangkong University, 2016
  • [14] LI Yongtao. The study on defect formation in laser additive manufacturing titanium alloy[D]. Dalian: Dalian University of Technology, 2017
  • [15] GRASSO M, COLOSIMO B M. Process defects and in situ monitoring methods in metal powder bed fusion:A review[J]. Measurement Science and Technology, 2017, 28(4): 044005. DOI:10.1088/1361-6501/aa5c4f
  • [16] BARTLETT J L, HEIM F M, MURTY Y V, et al. In situ defect detection in selective laser melting via full-field infrared thermography[J]. Additive Manufacturing, 2018, 24: 599. DOI:10.1016/j.addma.2018.10.045
  • [17] LOPEZ A, BACELAR R, PIRES I, et al. Non-destructive testing application of radiography and ultrasound for wire and arc additive manufacturing[J]. Additive Manufacturing, 2018, 21: 304. DOI:10.1016/j.addma.2018.03.020
  • [18] CAO Lin, CHEN Suiyuan, WEI Mingwei, et al. Effect of laser energy density on defects behavior of direct laser depositing 24CrNiMo alloy steel[J]. Optics&Laser Technology, 2019, 111: 541. DOI:10.1016/j.optlastec.2018.10.025
  • [19] MONTAZERI M, YAVARI R, RAO P, et al. In-process monitoring of material cross-contamination defects in laser powder bed fusion[J]. Journal of Manufacturing Science and Engineering, 2018, 140(11): 111001. DOI:10.1115/1.4040543
  • [20] ALESHIN N P, GRIGOR E M V, SHCHIPAKOV N A, et al. Using nondestructive testing methods for in-production quality control of additive manufactured parts[J]. Russian Journal of Nondestructive Testing, 2016, 52(9): 533. DOI:10.1134/S1061830916090023
  • [21] CUNNINGHAM R, ZHAO Cang, PARAB N, et al. Keyhole threshold and morphology in laser melting revealed by ultrahigh-speed x-ray imaging[J]. Science, 2019, 363(6429): 849. DOI:10.1126/science.aav4687
  • [22] PARAB N D, ZHAO C, CUNNINGHAM R, et al. Ultrafast X-ray imaging of laser-metal additive manufacturing processes[J]. Journal of Synchrotron Radiation, 2018, 25(5): 1467. DOI:10.1107/S1600577518009554
  • [23] CUNINAHAM R, NICOLAS A, MADSEN J, et al. Analyzing the effects of powder and post-processing on porosity and properties of electron beam melted Ti-6Al-4V[J]. Materials Research Letters, 2017, 5(7): 518. DOI:10.1080/21663831.2017.1340911
  • [24] STUGELMAYER E J. Characterization of process induced defects in laser powder bed fusion processed ALSI10MG alloy[D]. Montana: Montana Technological University, 2018
  • [25] CHERN A H, NANDWANA P, YUAN Tao, et al. A review on the fatigue behavior of Ti-6Al-4V fabricated by electron beam melting additive manufacturing[J]. International Journal of Fatigue, 2019, 119: 173. DOI:10.1016/j.ijfatigue.2018.09.022
  • [26] GUO Peng. Study on mechanical properties and milling performance of stainless steel manufactured by laser additive manufacturing[D]. Ji’nan: Shandong University, 2017
  • [27] BAI Jiuyang. Microstructure evolution of 2219-Al during GTA based additive manufacturing and heat treatment[D]. Harbin: Harbin Institute of Technology, 2017
  • [28] YANG Kaike, ZHU Jihong, WANG Chuang, et al. Experimental validation of 3D printed material behaviors and their influence on the structural topology design[J]. Computational Mechanics, 2018, 61(5): 596. DOI:10.1007/s00466-018-1537-1
  • [29] DEBROY T, WEI H L, ZUBACK J S, et al. Additive manufacturing of metallic components-Process, structure and properties[J]. Progress in Materials Science, 2018, 92: 133. DOI:10.1016/j.pmatsci.2017.10.001
  • [30] ALBAKRI M I, STURM L D, WILLIAMS C B, et al. Impedance-based non-destructive evaluation of additively manufactured parts[J]. Rapid Prototyping Journal, 2017, 23(3): 589. DOI:10.1108/RPJ-03-2016-0046
  • [31] OBATON A, LE M, PREZZA V, et al. Investigation of new volumetric non-destructive techniques to characterise additive manufacturing parts[J]. Welding in the World, 2018, 62(5): 1050. DOI:10.1007/s40194-018-0593-7
  • [32] REPOSSINI G, LAGUZA V, GRASSO M, et al. On the use of spatter signature for in-situ monitoring of laser powder bed fusion[J]. Additive Manufacturing, 2017, 16: 35. DOI:10.1016/j.addma.2017.05.004
  • [33] KRAUSS H, ZEUGNER T, ZAEH M F. Layerwise monitoring of the selective laser melting process by thermography[J]. Physics Procedia, 2014, 56: 65. DOI:10.1016/j.phpro.2014.08.097
  • [34] BERUMEN S, BECHMANN F, LINDNER S, et al. Quality control of laser and powder bed-based Additive Manufacturing (AM) technologies[J]. Physics Procedia, 2010, 5: 617. DOI:10.1016/j.phpro.2010.08.089
  • [35] CLIJSTERS S, CRAEGHS T, BULS S, et al. In situ quality control of the selective laser melting process using a high-speed, real-time melt pool monitoring system[J]. The International Journal of Advanced Manufacturing Technology, 2014, 75(5-8): 1089. DOI:10.1007/s00170-014-6214-8
  • [36] CHOO H, SHAM K L, BOHLING J, et al. Effect of laser power on defect, texture, and microstructure of a laser powder bed fusion processed 316L stainless steel[J]. Materials and Design, 2019, 164: 107534. DOI:10.1016/j.matdes.2018.12.006
  • [37] CRAEGHS T, CLIJSTERS S, YASA E, et al. Determination of geometrical factors in Layerwise Laser Melting using optical process monitoring[J]. Optics and Lasers in Engineering, 2011, 49(12): 1440. DOI:10.1016/j.optlaseng.2011.06.016
  • [38] YANG Liushan. Study on CCD-based detection systemfor laser cladding[D]. Changsha: Hunan University, 2011
  • [39] CHEN Dianbing. Experimental research on the molten pool image detection during laser cladding process[D]. Shanghai: Shanghai Jiao Tong University, 2015
  • [40] STOCKMAN T, KNAPP C, HENDERSON K, et al. Stainless steel 304L LENS AM process monitoring using in-situ pyrometer data[J]. JOM, 2018, 70(9): 1835. DOI:10.1007/s11837-018-3033-7
  • [41] TAN Hua, CHEN Jing, ZHANG Fengying, et al. Estimation of laser solid forming process based on temperature measurement[J]. Optics&Laser Technology, 2010, 42(1): 47. DOI:10.1016/j.optlastec.2009.04.016
  • [42] YE Dongsen, HONG G S, ZHANG Yingjie, et al. Defect detection in selective laser melting technology by acoustic signals with deep belief networks[J]. The International Journal of Advanced Manufacturing Technology, 2018, 96(5-8): 2791. DOI:10.1007/s00170-018-1728-0
  • [43] CHEN Bo, YAO Yongzhen, TAN Caiwang, et al. A study on spectral characterization and quality detection of direct metal deposition process based on spectral diagnosis[J]. The International Journal of Advanced Manufacturing Technology, 2018, 96(9-12): 4231. DOI:10.1007/s00170-018-1889-x
  • [44] GARCIA D L Y A, PFLEGER M, ARAMENDI B, et al. Online cracking detection by means of optical techniques in laser-cladding process[J]. Structural Control and Health Monitoring, 2019, 26(3): 2291. DOI:10.1002/stc.2291
  • [45] RIEDER H, DILLHOFER A, SPIES M. Online monitoring of additive manufacturing processes using ultrasound[C]//Proceedings 11th European Conference on Non-Destructive Testing. Prague: ECNDT, 2014: 16533
  • [46] CHAUVEAU D. Review of NDT and process monitoring techniques usable to produce high-quality parts by welding or additive manufacturing[J]. Welding in the World, 2018, 62(5): 1105. DOI:10.1007/s40194-018-0609-3
  • [47] MILLION C, VANHOYE A, OBATON A, et al. Development of laser ultrasonics inspection for online monitoring of additive manufacturing[J]. Welding in the World, 2018, 62(3): 653. DOI:10.1007/s40194-018-0567-9
  • [48] EVERTON S, DICKENS P, TUCK C, et al. Using laser ultrasound to detect subsurface defects in metal laser powder bed fusion components[J]. JOM, 2018, 70(3): 378. DOI:10.1007/s11837-017-2661-7
  • [49] CERNIGLIA D, SCAFIDI M, PANTANO A, et al. Inspection of additive-manufactured layered components[J]. Ultrasonics, 2015, 62: 292. DOI:10.1016/j.ultras.2015.06.001
  • [50] LEVESQUE D, BESCOND C, LORD M, et al. Inspection of additive manufactured parts using laser ultrasonics[C]//Proceedings of 42nd Annual Review of Progress in Quantitative Nondestructive Evaluation. Minneapolis: AIP, 2016: 130003. DOI: 10.1063/1.4940606
  • [51] DU Wei, BAI Qian, WANG Yibo, et al. Eddy current detection of subsurface defects for additive/subtractive hybrid manufacturing[J]. The International Journal of Advanced Manufacturing Technology, 2018, 95(9-12): 3185. DOI:10.1007/s00170-017-1354-2
  • [52] WANG Yibo, BAI Qiai, DU Wei, et al. Edge effect on eddy current detection for subsurface defects in Titanium Alloys[C]//Proceedings of the 8th International Conference on Computational Methods, Guilin: [s.n.], 2017: 1445
  • [53] TODOROV E, BOULWARE P, GAAH K. Demonstration of array eddy current technology for real-time monitoring of laser powder bed fusion additive manufacturing process[C]//Proceedings of Nondestructive Characterization and Monitoring of Advanced Materials, Aerospace, Civil Infrastructure, and Transportation XII. Denver: SPIE, 2018: 1059913. DOI: 10.1117/12.2297511
  • [54] MONTINARO N, CERNIGLIA D, PITARRESI G. A numerical and experimental study through laser thermography for defect detection on metal additive manufactured parts[J]. Frattura ed Integrita Strutturale, 2018, 12(43): 231. DOI:10.3221/IGF-ESIS.43.18
  • [55] XIE Ruidong, LI Dichen, CUI Bin, et al. A defects detection method based on infrared scanning in laser metal deposition process[J]. Rapid Prototyping Journal, 2018, 24(6): 945. DOI:10.1108/RPJ-04-2017-0053
  • [56] SCHWERDTFEGER J, SINGER R F, KORNER C. In situ flaw detection by IR-imaging during electron beam melting[J]. Rapid Prototyping Journal, 2012, 18(4): 259. DOI:10.1108/13552541211231572
  • [57] BARUA S, LIOU F, NEWKIRK J, et al. Vision-based defect detection in laser metal deposition process[J]. Rapid Prototyping Journal, 2014, 20(1): 77. DOI:10.1108/RPJ-04-2012-0036
  • [58] ABDELRAHMAN M, REUTZEL E W, NASSAR A R, et al. Flaw detection in powder bed fusion using optical imaging[J]. Additive Manufacturing, 2017, 15: 1. DOI:10.1016/j.addma.2017.02.001
  • [59] YAO Bing, IMANI F, SAKPAL A S, et al. Multifractal analysis of image profiles for the characterization and detection of defects in additive manufacturing[J]. Journal of Manufacturing Science and Engineering, 2018, 140(3): 031014. DOI:10.1115/1.4037891
  • [60] GRASSO M, LAGUZZA V, SEMERARO Q, et al. In-process monitoring of selective laser melting:Spatial detection of defects via image data analysis[J]. Journal of Manufacturing Science and Engineering, 2017, 139(5): 051001. DOI:10.1115/1.4034715


Leave a Reply



Inquery now



  • Email me
    Mail to us