Metallized film capacitors (MFCs) are widely used in the power electronics industry due to their unique self-healing (SH) capability. SH performance is an essential assessment for MFC reliability verification in industrial production. The SH phenomenon of metallized films usually occurs rapidly in a very short period, and its real-time evolution details are often difficult to capture and analyze. In this paper, a test system for the SH performance of metallized films for capacitors was constructed. The system consists of three components: a voltage–current characteristic testing and current pulse capture device, a microscopic image real-time acquisition device, and an integrated analysis processing device. Through the voltage–current characteristic testing and current pulse capture device, the electrical parameters of the SH point, such as SH times, breakdown field strength, SH current, and SH energy, are obtained; through a microscopic image real-time acquisition device, the real-time spatial positioning of the SH point was obtained, and the interconnection between the morphology of the SH point and the electrical properties was established. The relationship between the SH point and the temperature distribution was further established using thermal imaging technology, which lays the foundation for a thorough and timely assessment and analysis of the failure mechanism and the real-time evolution of the metallized film SH process. This significantly improves the effectiveness of SH property research.

Metallized film capacitors (MFCs) play a crucial role in modern electronic devices for many applications, especially in pulsed power supplies,1–3 electric vehicles,4–6 and electromagnetic emission systems.7,8 These capacitors usually use polypropylene film as the base film. A metallic coating is formed on the surface by a vapor-phase deposition technique, generally in the order of tens of nanometers in thickness.9,10 This unique structure gives MFCs unique self-healing (SH) properties. SH is one of the main features that make it superior to other capacitors. It enables it to self-heal and clean up electrical weaknesses inherent in the base film or formed by the process, thus improving the operating field strength and prolonging the service life.11,12 Under the action of an applied electric field, the SH points on the surface of the metallized film form a short-circuit discharge channel, generating a high current. The joule heat generated by the high current causes the metal layer to evaporate rapidly, forming an SH point, which allows the film capacitor to résumé operation, a series of events known as the “SH process.”13–18 

The energy released during the SH process leads to physical and chemical changes in the metallized film, making SH extremely complex.19–23 Films with good SH properties return to their original state, but with a loss of capacity. In contrast, poorly SH films may fail to heal and increase the risk of explosion in MFCs. Therefore, reliability testing of metallized film SH performance becomes crucial.24,25 Previously adopted test methods, such as the voltage test method,26,27 the ultrasonic method,28,29 and the leakage current test method,30 can evaluate the SH performance of metallized films. Still, these methods tend to focus on the measurement of a specific performance parameter, which makes it difficult to reflect the SH performance of metallized films comprehensively. Due to the uncertainty of SH, a weak SH current signal may also be the result of SH in the film, which requires high sensitivity of the test equipment. High sensitivity means that a broader range of SH conditions can be detected, which is essential for different kinds of MFCs and applications under various operating conditions.

Therefore, this paper presents a test device and data analysis method for the self-healing performance of metallized film for capacitors, which can quickly assess the SH performance of metallized film for capacitors from different manufacturers or batches. The device integrates real-time video recording, thermal imaging technology, current pulse capture technology, and image recognition technology to provide multi-dimensional data for film breakdown time, breakdown field strength, breakdown current, and breakdown energy. These parameters will help to more comprehensively evaluate and analyze the real-time evolution of metallized films and failure mechanisms, providing a new approach to improving capacitor quality and production safety.

The working environment of this experiment is shown in Fig. 1(a). The equipment used for the SH test of metallized thin films mainly involves a real-time testing and analysis system for the electrical and voltammetric characteristics of thin films,31 which includes a voltage–current characteristic testing and a current pulse capture device, a microscopic image real-time acquisition device, and a comprehensive analysis processing device. The voltage–current characteristic testing includes a sample clamping base, a test probe, a digital source meter, and a test control module. The test probe is mounted on the sample clamping base, and the digital source meter can provide a stable voltage to the sample through the test probe. The test probe, the digital source meter, and the test control module are connected sequentially. The current pulse capture device includes a current waveform analyzer, a passive probe interface, and a current sensor. The current sensor is sequentially connected to the current waveform analyzer and digital source meter, which can accurately capture and display the dynamic current waveform. The real-time acquisition device for microscopic images includes a connected digital optical microscope module, a 3D infrared thermal imaging module, and an image acquisition module. The comprehensive analysis processing device connects the test control and image acquisition modules. It indirectly collects film data in the form of a user interface, analyzes and processes the raw data through the algorithms built into the system, and finally visually displays the SH information of the film in the form of a visualization.

FIG. 1.

(a) Test setup for breakdown and SH characteristics of metallized thin films. (b) Micro-morphology real-time analysis software system and data acquisition.

FIG. 1.

(a) Test setup for breakdown and SH characteristics of metallized thin films. (b) Micro-morphology real-time analysis software system and data acquisition.

Close modal

The test user interface, as shown in Fig. 1(b), acquires video and image information from a real-time micro-image acquisition device about the surface topographical changes in a thin film sample before and after breakdown and intelligently labels and visualizes the micro-morphological changes in that sample. This phenomenon can also be traced back to a specific moment for in-depth study. The interface provides the lower and upper limit values of the voltage, and the user can set the voltage range by himself according to the needs of the test. When the test is finished, the file save path, video save path, and image save path can be set to facilitate subsequent data analysis and processing. In software design, the image recognition algorithm is used to capture and locate the SH phenomenon, which can determine the temporal and spatial location of the SH phenomenon, analyze the morphology of the SH points, and count and classify the number and area of the SH points. By recording the breakdown process through image processing and assisted analysis of the video, information can be obtained about the number, morphology, and composition of the SH points, as well as the energy relationships of specific SH phenomena. This information helps to study the particular morphological changes that occur at the moment of film breakdown, such as edge breakdown and the number of breakdown points.

The workflow diagram of the automatic real-time testing system is shown in Fig. 2. Starting the volt–ampere characteristic testing device system for the initialization of the breakdown test, the volt–ampere typical data of the film to be tested can be collected, and the I–V characteristic curve can be plotted. Turning on the current waveform analyzer can capture the details of the sudden change of the SH current to determine the release of energy when the film is SH. At the same time, open the microscopic image real-time acquisition device to observe the SH process of the film in real-time, capture the SH phenomenon of the film through the image recognition technology, and identify and locate the time and position where the SH occurs. Critical: The number of SH points and the area size of the SH points can be obtained by analyzing the image processing video key framescritical.

FIG. 2.

Flow chart of testing and data analysis.

FIG. 2.

Flow chart of testing and data analysis.

Close modal

The metalized film SH characteristics were realized on a DC voltage experimental platform. The film samples were selected as one-sided edge-retaining MFCs with a rated capacitance of 0.1 µF, a thickness of about 7 µm, and a film width of 75 mm, and the metal electrodes were selected with Zn as the metal coating. To investigate the whole process of SH instant on the surface of metallized thin films, we cut rectangular thin film samples with an area of 10 mm2 for testing, and the voltammetric characteristics under the microscopic images are reflected in Fig. 2(a). First, the film was placed on the ITO with the coated side facing down, and then a current loop was formed by contacting the film and the ITO surface with a probe, respectively. Thin film samples were observed under an optical microscope and recorded in real-time throughout the monitoring process. The physicochemical changes at each breakdown point of the sample are in real-time, and the optical microscope can capture the electric field. The electrical test setup provides a steady DC boost of 2 V/s, and a microscopic imaging unit was responsible for collecting video playback of the entire process.

The whole process is divided into three stages. The first stage is the edge breakdown. In Fig. 3(a), the voltage is ramped up from 0 to 859 V from 0 to 434 s, showing a nonlinear distribution. The film does not show any obvious SH phenomena and remains stable. At the same time, at 435 s, when the voltage reaches 860 V, the curve shows small spikes (region I) frequently. The real-time video recording of the microscopic image captures the film breakdown at the edge part, corresponding to region I in Fig. 3(e). This phenomenon is called the edge breakdown effect. The smaller radius of curvature of the edge part of the metallized film makes the electric field lines more concentrated in the region. The electric field intensity reaches a critical value near its surface, which causes the molecules on the surface of the film to be ionized. The formation of electrons and ions is continuously accelerated under the action of the electric field, which ultimately leads to the breakdown of the edge, whose macroscopic manifestation is that the electric arc is traced across the metal plating layer of the film. Then, the breakdown occurs within tens of nanoseconds, and the breakdown traces are left behind. The second stage is the SH corrosion breakdown, as shown in Fig. 3(a); when the voltage reaches 960 V, there is a current spike with a significant difference (region II), corresponding to region II in Fig. 3(f). At this moment, the breakdown occurs for the first time inside the film, but a small amount of metal plating remains on the surface. As the DC voltage increases, the SH corrosion continues until the voltage reaches 984 V, at which point the corrosion area no longer expands and remains relatively stable.

FIG. 3.

(a) Voltammetric characteristic curve and microscopic observation. (b)–(g) Video recording of the breakdown process of a film with an area of 10 mm2.

FIG. 3.

(a) Voltammetric characteristic curve and microscopic observation. (b)–(g) Video recording of the breakdown process of a film with an area of 10 mm2.

Close modal

As shown in Fig. 4(a), the video data captured in real-time by the microscopic image during the breakdown test can provide the approximate distribution of the SH points and display them after comprehensive analysis and processing. The built-in image recognition algorithm first intercepts the surface topography of the same batch of film samples before and after the breakdown. After gray scale processing, the two are subjected to image subtraction and binarization to obtain the critical location feature information. Subsequently, the location information is extracted, and the contour is drawn to show the user a non-binarized SH point distribution map. The number and area of the SH points are calculated. Figure 4(b) is the above microscopic image recognition and capture algorithm flow.

FIG. 4.

(a) Real situation of self-healing point microscopic image recognition. (b) Microscopic image recognition capturing algorithm flow.

FIG. 4.

(a) Real situation of self-healing point microscopic image recognition. (b) Microscopic image recognition capturing algorithm flow.

Close modal

Metallized films are subjected to electric field stress, and a breakdown point occurs when the electric field strength exceeds the dielectric strength. In the vicinity of the breakdown point, the electric field is strong enough to ionize the air molecules, generating positive and negative charges. Positive charges move toward the negative electrode, and when the collision energy is high enough, some electrons escape the metal surface, creating an SH current. The escaping electrons neutralize the positive charge, releasing ionization energy and causing the metal to evaporate. The positive charge continues to move toward the air gap region, colliding with air molecules and expanding the charge region. Eventually, the positive charge moves to the vicinity of the negative electrode, releasing free electrons and forming a self-healing current.32 The diffusion of heat often accompanies the process of SH, and when the surface of the film is punctured, the air molecules in the air layer are ionized with the SH current, and the charge travels through a larger area of diffusion, which leads to the evaporation of a larger area.

In order to research whether this phenomenon is widespread in every metallized film process, we used an infrared thermography analyzer to collect information on the process of film breakdown. Figures 5(a) and 5(b) show the microscopic images of the film at the moment of breakdown. Figure 5(c) shows the temperature distribution of the film before it is broken down, which will cause the phenomenon of non-uniformity of the temperature on the thermal imager due to the metal plating on its surface, which will reflect the temperature of the external environment. Figure 5(d) shows the temperature monitoring of the film at the moment of breakdown; when the thermal imager monitors a significant temperature change, it will show higher peaks on the 3D model constructed by the in-house program, which has been tested on several film samples.

FIG. 5.

(a) and (b) Surface temperature of the film before breakdown. (c) and (d) Surface temperatures of films at breakdown.

FIG. 5.

(a) and (b) Surface temperature of the film before breakdown. (c) and (d) Surface temperatures of films at breakdown.

Close modal
SH energy (Ws) is an important parameter for measuring the SH properties of metallized films. Typically, the SH energy is tens to hundreds of millijoules. The current pulse capture device can capture the SH current (is) of the film in real-time, and we can measure it with this device. At the same time, the voltage (Us) across the film is captured, so the self-healing process can be characterized by measuring the self-healing current (is) and voltage (Us) during the self-healing event. The self-healing energy can be expressed by the following equation:1,15,33–36
Ws=Usisdt.

To investigate the relationship between the SH point and the SH energy of the metallized films, 13 film samples with an area of 30 cm2 were cut for breakdown testing. Figure 6(a) depicts the surface texture after the breakdown of the film. Among numerous breakdown points, four primary locations were selected for detailed investigation, corresponding to the microscopic details in Fig. 6(b). At location I, there is a widespread and scattered distribution of breakdown points, which is a common SH phenomenon observed in metalized films. At location II, breakdown points with a diameter of around 10 mm emerge. The breakdown at III is more concentrated, with obvious burn traces; the outer layer has an obvious contour and complete boundary after breakdown; the inner layer is burned and deformed, and there are obvious dendritic discharge channels. IV is edge breakdown, mainly in the edge part of the film; and there is a regional type of corrosion phenomenon.

FIG. 6.

(a) Surface texture of a thin film sample after breakdown. (b) Thin film texture and magnified details before and after the breakdown at the same position. (c) SH area and the SH energy of the same thin film sample. (d) Statistics of SH point and SH energy of the thin film samples.

FIG. 6.

(a) Surface texture of a thin film sample after breakdown. (b) Thin film texture and magnified details before and after the breakdown at the same position. (c) SH area and the SH energy of the same thin film sample. (d) Statistics of SH point and SH energy of the thin film samples.

Close modal

Figure 6(c) shows the relationship between the area of the SH point and the SH energy on this film, and it can be seen that the SH energy is proportional to the SH area. Due to the existence of several defects or impurities on the surface of the film, which is very much related to the process of industrial vacuum vapor deposition of metal particles, there are some differences in the results brought about by the breakdown test of each batch of films. According to the results counted by the comprehensive processing and analysis device, as shown in Fig. 6(d), it can be seen that all film samples basically have an SH energy of not less than 30 mJ when SH occurs. The number of SH points is not the same.

During real-time testing, the current in the metallized film changes from moment to moment as the DC voltage continues to rise, and when a breakdown event occurs, the current suddenly increases and forms a peak. At this point, it is known as the SH current. The peak magnitude of the SH current usually reflects the SH speed and efficiency of the metallized film; the higher the peak value the faster the SH. The duration of the SH current can assess the speed and stability of the SH process. In order to study the current variations at the moment of the breakdown of the metallized film, the SH currents were analyzed and processed using a current waveform analyzer. As shown in Fig. 7, the interface of the device contains the scanning and capturing of the current over a period of time, the analysis and processing, the statistical methods, etc. Due to its high sensitivity and accuracy up to the nano-ampere level, the small changes in current occurring during the breakdown process can be quickly acquired and saved in the form of a file.

FIG. 7.

Current waveform analyzer interface.

FIG. 7.

Current waveform analyzer interface.

Close modal

Combining the voltage–current characteristic testing with the microscopic image setup allows further observation of the details of the occurrence of the SH event in the film. As shown in Figs. 8(a)8(d), which show the pulse capture of the breakdown current of the metallized thin film, it can be seen that the peak value of the current acquired in Fig. 8(a) is 6.091 µA, which is the lowest of the three. The duration of the current is about 16 µs, which is the highest of the three. The current applied to the thin film induces lower energy, and the SH duration is longer so that the breakdown region corresponding to the thin film exhibits a dotted-type structure. The peak current in Fig. 8(b) is the highest of the three and is about 8 µs. The SH current in Fig. 8(c) has the lowest duration, which implies that the short-time SH is very fast and can quickly restore the state before SH.

FIG. 8.

(a)–(d) Capture of breakdown current pulse and micro-morphology of self-healing point.

FIG. 8.

(a)–(d) Capture of breakdown current pulse and micro-morphology of self-healing point.

Close modal

Table I shows the film statistics of the samples from three different manufacturers; 15 films were taken from each, and the film samples were analyzed in terms of electrical parameters such as SH time, SH energy, SH current, and dielectric strength. According to the SH time, film C is slightly better than film B, while film A has a relatively longer SH time. Based on the SH energy, film A performs slightly better than film B and film C. Based on the peak breakdown voltage, film A has the best performance, followed by film B and film C, which have the worst performance. Based on the average DC dielectric strength, film A performs slightly better than films B and C. From the point of view of breakdown strength and electrical insulation properties, it can be judged that film A performs better.

TABLE I.

SH breakdown statistics values and breakdown distributions, including a comparison of statistical analyses of samples from three different manufacturers (A, B, and C).

ParametersFilm-type
ABC
Thickness (μm) 7.59 7.14 7.20 
SH points (dots/mm24.24 5.31 6.60 
SH time (μs) 13.43 11.67 10.92 
SH energy (mJ) 43.3 41.8 39.2 
Peak SH current (μA) 5.49 5.47 5.47 
Peak breakdown voltage (VDC) 930.3 914.8 885.8 
DC dielectric strength (VDC/μm) 132.6 130.1 128.1 
ParametersFilm-type
ABC
Thickness (μm) 7.59 7.14 7.20 
SH points (dots/mm24.24 5.31 6.60 
SH time (μs) 13.43 11.67 10.92 
SH energy (mJ) 43.3 41.8 39.2 
Peak SH current (μA) 5.49 5.47 5.47 
Peak breakdown voltage (VDC) 930.3 914.8 885.8 
DC dielectric strength (VDC/μm) 132.6 130.1 128.1 

Figure 9(a) shows the energy dispersive spectra (EDS) of the corresponding SH point after film breakdown; the darker color in the figure represents the higher elemental content in the region. Figure 9(b) shows the result of the surface scan of the corresponding SH point, and it can be seen that a high content of zinc element and a trace amount of aluminum element are detected at the SH point after a breakdown. The metal layer of the experimental metallized film capacitor is made of zinc–aluminum particles by vapor deposition. Still, the content of aluminum is low, so it is difficult to distinguish the change in the content of aluminum in the scanning diagram of the aluminum element. The scans of zinc and oxygen elements show that the zinc levels within the self-healing site are lower than in the surrounding area, while the oxygen levels are relatively higher. When a metallized film undergoes breakdown, an arc discharge occurs, during which the energy is highly concentrated, resulting in a sharp increase in temperature in the local area. The metal coating evaporates, and the metal elements spread around, reducing the zinc content in the breakdown zone. At high temperatures, the zinc particles are subjected to evaporation and oxidation with oxygen in the air, resulting in zinc oxide. It can also be seen from the scan of the oxygen elements that the high oxygen content is evenly distributed inside the SH point, in contrast to the surrounding area of the SH point.

FIG. 9.

(a) EDS energy spectrum of the SH point of the metallized film. (b) Surface scanning results of the corresponding SH point.

FIG. 9.

(a) EDS energy spectrum of the SH point of the metallized film. (b) Surface scanning results of the corresponding SH point.

Close modal

Technical considerations are as follows:

  1. Real-time localization of film breakdown phenomena.

Metallized films undergo significant SH during testing, but it is difficult to predict the timing and spatial positioning of the SH. Our microscopic real-time video recorder allows us to observe the SH of metallized films in real-time and to determine the time and location of the SH points. We can analyze the SH points with relevant statistics and process them with the integrated analysis and processing device.

  1. Image recognition of film SH events.

In order to intuitively count the data information of the SH points of the film, we introduced the image recognition technique, which can quickly and accurately identify the critical information in the SH event, such as the number of SH points and the area statistics. This technique improves the testing efficiency of metallized thin films and reduces the cost of manual counting, while it is expected to be able to be applied in large-scale testing.

  1. Real-time monitoring and analysis of local temperatures.

In order to explore the temperature distribution of the film at the moment of breakdown, we use 3D infrared monitoring technology to observe the energy distribution released at the moment of breakdown in real-time and generate planar and 3D imaging effects in real-time to understand the process of energy release more deeply.

  1. Real-time, high-precision breakdown current pulse capture.

In our study, we introduced a current waveform analyzer to provide new insights and analytical tools for studying the SH properties of metallized films. This allowed us to gain a more comprehensive understanding of the temporal evolution and waveform characteristics of the current during the SH process and to observe small variations in the current waveform. Real-time current data are also provided, allowing us to accurately measure the peak and duration of the current at the point of SH.

This paper considers the error analysis of real-time test systems, which is attributed to system, test environment, operational, and algorithmic errors. The following section is a discussion of its error analysis:

  1. Instrument Errors.

The various instruments used in this experiment, such as the microscope camera, 3D infrared thermal imager, digital source meter, etc., may have inevitable errors. The measurement accuracy, response time, and other properties of these devices will impact the experimental results.

  1. Test Environment Errors.

There are some temperature or humidity changes in the test environment, and fluctuations in these factors may lead to instability in the results. Electromagnetic interference in the laboratory may affect the performance of equipment testing, such as the lack of means of shielding electromagnetic interference at the terminals, which may affect the distortion of the current waveform signal.

  1. Sample differences.

There may be some sample differences in the metallized films supplied in different batches, including variations in cut size, surface topography, degree of edge bending, etc. All these differences may have an impact on the test results.

  1. Algorithmic Errors.

The algorithm provided in the integrated image processing device is capable of performing an essential image recognition task, but since the breakdown points do not have a fixed shape and the video information returned from the digital optical microscope is limited by its picture quality as well as its pixels, it is necessary to adjust the threshold size of the algorithm a number of times to calculate the number and the area of the breakdown points, which will deviate from the actual number and area.

In this study, we developed a testing apparatus for the SH properties of metallized films for capacitors, along with complementary data analysis methods. Integrated with an automated real-time testing system, the setup dynamically records the SH process of metallized films under a strong electric field. The study explores the relationship between the morphology of SH points and electrical characteristics. The system enables real-time observation of breakdown field strength, SH energy, SH current, and SH time. These methods contribute to the analysis of the real-time evolution process and failure mechanisms of metallized film breakdown, offering novel approaches for the rapid assessment of film performance.

This work was supported by the Guangdong Basic and Applied Basic Research Foundation (Grant No. 2022A1515140004), the Guangdong Science and Technology Plan (Grant No. 2022A0505020022), and the key laboratory in the higher education institutions of the Educational Commission of Guangdong Province (Grant No. 2021KSYS008).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Shaopeng Zhou: Writing – original draft (equal). Deping Chen: Writing – review & editing (equal). Baoyu Du: Resources (equal). Pan Wang: Resources (equal). Xiucai Wang: Funding acquisition (equal). Wenbo Zhu: Funding acquisition (equal). Si Liu: Funding acquisition (equal). Peng Xiao: Funding acquisition (equal). Jianwen Chen: Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
V.
Belko
et al, “
Current pulse polarity effect on metallized film capacitors failure
,”
IEEE Trans. Plasma Sci.
45
(
6
),
1020
1025
(
2017
).
2.
C.
Joubert
et al, “
Electric field and equivalent circuit in all-film capacitors
,”
J. Appl. Phys.
81
(
10
),
6579
6584
(
1997
).
3.
Z.
Li
et al, “
Temperature rise of metallized film capacitors in repetitive pulse applications
,”
IEEE Trans. Plasma Sci.
43
(
6
),
2038
2045
(
2015
).
4.
N.
Kim
et al, “
Experimental comparisons and evaluations of different types of DC-link capacitors for VSI-based electric compressors in battery electric vehicle systems
,”
Electronics
9
(
8
),
1276
(
2020
).
5.
K.
Voloshin
et al, “
Applicability of energy storage units to electric transport
,” in
E3S Web of Conferences
(
EDP Sciences
,
2019
).
6.
D.
Montanari
et al, “
Film capacitors for automotive and industrial applications
,” in
Proceedings of the CARTS USA
,
2009
.
7.
T.
Scholz
et al, “
Development of a high energy density storage capacitor for NIF
,” in
Digest of Technical Papers. 12th IEEE International Pulsed Power Conference. (Cat. No. 99CH36358)
(
IEEE
,
1999
).
8.
L.
Hua
et al, “
Study on metallized film capacitor and its voltage maintaining performance
,”
IEEE Trans. Magn.
45
(
1
),
327
330
(
2009
).
9.
R. W.
Brown
, “
Linking corrosion and catastrophic failure in low-power metallized polypropylene capacitors
,”
IEEE Trans. Device Mater. Reliab.
6
(
2
),
326
333
(
2006
).
10.
D. V.
Lackey
, “
A versatile capacitor web coater with multiple source capability
,”
Anti-Corros. Methods Mater.
39
(
2
),
4
8
(
1992
).
11.
W.
Sarjeant
, “
Capacitors
,”
IEEE Trans. Electr. Insul.
25
(
5
),
861
922
(
1990
).
12.
C.
Reed
and
S.
Cichanowskil
, “
The fundamentals of aging in HV polymer-film capacitors
,”
IEEE Trans. Dielectr. Electr. Insul.
1
(
5
),
904
922
(
1994
).
13.
J.
Bond
, “
A new mitigation strategy for failures in metallized polypropylene capacitors
,”
Electronic Concepts Inc.
,
2012
.
14.
N.
Valentine
et al, “
Metallized film capacitors used for EMI filtering: A reliability review
,”
Microelectron. Reliab.
92
,
123
135
(
2019
).
15.
V.
Belko
et al, “
Energy characteristics of self-healing process in metallized film capacitors
,” in
E3S Web of Conferences
(
EDP Sciences
,
2019
).
16.
Y.
Chen
et al, “
Study on self-healing and lifetime characteristics of metallized-film capacitor under high electric field
,”
IEEE Trans. Plasma Sci.
40
(
8
),
2014
2019
(
2012
).
17.
N.
Klein
, “
The mechanism of self-healing electrical breakdown in MOS structures
,”
IEEE Trans. Electron Devices
ED-13
(
11
),
788
805
(
1966
).
18.
D.
Shaw
et al, “
A changing capacitor technology—Failure mechanisms and design innovations
,”
IEEE Trans. Electr. Insul.
EI-16
(
5
),
399
413
(
1981
).
19.
M.
Yao
et al, “
Anodic oxidation in aluminum electrode by using hydrated amorphous aluminum oxide film as solid electrolyte under high electric field
,”
ACS Appl. Mater. Interfaces
8
(
17
),
11100
11107
(
2016
).
20.
X.
Chen
et al, “
TiO2 nanoparticle/polyimide nanocomposite for ultrahigh-temperature energy storage
,”
Nanomaterials
12
(
24
),
4458
(
2022
).
21.
M.
Yao
et al, “
Ionic transport and barrier effect of anodic oxide layer in a solid-state Al2O3 capacitor under high electric field
,”
Electrochim. Acta
224
,
235
242
(
2017
).
22.
D.
Hu
et al, “
Critical electric field stabilizing structure of Al2O3/TiO2/Al2O3 thin film for achieving high energy density
,”
Ceram. Int.
49
(
11
),
17296
17304
(
2023
).
23.
Q.
Cao
et al, “
Nonsolid TiOx nanoparticles/PVDF nanocomposite for improved energy storage performance
,”
ACS Appl. Mater. Interfaces
14
(
6
),
8226
8234
(
2022
).
24.
J.
Tortai
et al, “
Self-healing of capacitors with metallized film technology: Experimental observations and theoretical model
,”
J. Electrost.
53
(
2
),
159
169
(
2001
).
25.
M.
Min
, “
Study on charging characteristics and voltage maintaining characteristics of metallized film capacitors
,”
J. Fusion Energy
34
(
5
),
1016
1019
(
2015
).
26.
S.
Ul-Haq
and
G. G.
Raju
, “
Weibull statistical analysis of area effect on the breakdown strength in polymer films
,” in
Annual Report Conference on Electrical Insulation and Dielectric Phenomena
(
IEEE
,
2002
).
27.
M.
Zhu
et al, “
Experimental investigation of self-healing properties of metalized film capacitors
,” in
2020 4th International Conference on HVDC (HVDC)
(
IEEE
,
2020
).
28.
Y.
Ma
et al, “
Detection of self-healing discharge in metallized film capacitors using an ultrasonic method
,”
Electronics
9
(
11
),
1893
(
2020
).
29.
L.
Zheng
et al, “
Research on self-healing discharge detection method of metallized film capacitor based on ultrasonic and ultra high frequency signals
,” in
2022 IEEE International Power Modulator and High Voltage Conference (IPMHVC)
(
IEEE
,
2022
).
30.
M.
Carlen
et al, “
Electrical endurance characterization of polypropylene winding capacitors for traction applications: A new experimental method
,” in
7th International Conference on Dielectric Materials, Measurements and Applications (Conf. Publ. No. 430)
(
IET
,
1996
).
31.
D.
Hu
et al, “
Automated real-time study of the defect-induced breakdown occurring on a film–electrode system under a high electric field
,”
Rev. Sci. Instrum.
92
(
12
),
123906
(
2021
).
32.
A.
Schneuwly
et al, “
Uncoupling behaviour of current gates in self healing capacitors
,”
Mater. Sci. Eng.: B
55
(
3
),
210
220
(
1998
).
33.
I. O.
Ivanov
et al, “
Characteristics of self-healing processes in metallized film capacitors with all-over and segmented electrodes
,” in
2022 Conference of Russian Young Researchers in Electrical and Electronic Engineering (ElConRus)
(
IEEE
,
2022
).
34.
Z.
Chen
et al, “
Prediction of breakdown field strength for large-area and multilayer film dielectrics
,”
IEEE Trans. Dielectr. Electr. Insul.
29
(
2
),
470
477
(
2022
).
35.
D. Y.
Glivenko
et al, “
Metallized film capacitors operation under the ramp voltage test
,” in
2023 5th International Youth Conference on Radio Electronics, Electrical and Power Engineering (REEPE)
(
IEEE
,
2023
).
36.
I.
Rytöluoto
and
K.
Lahti
, “
Effect of inter-layer pressure on dielectric breakdown characteristics of metallized polymer films for capacitor applications
,” in
2013 IEEE International Conference on Solid Dielectrics (ICSD)
(
IEEE
,
2013
).