Introduction

Ammunition is an essential war item that must be secured through a procurement plan for military logistics to maintain combat readiness for national defense1. Such ammunition that no longer serves its intended purpose must be appropriately disposed2. The use of simple disposal methods to manage the energetic materials and various compounds present in ammunition, which can weigh hundreds of kilograms, is impractical2. Moreover, handling idle ammunitions without a clear understanding of the weapon system, which may be due to differences in the manufacturing processes and raw materials used, poses additional economic viability and environmental protection challenges2,3,4. In such cases, information regarding the presence of energetic materials is often insufficient, hindering the determination of the nature of the weapon.

The pervasive threat of chemical weapons in such ammunition poses a significant challenge to international peace and security, emphasizing the need for effective disarmament strategies supported by the Chemical Weapons Convention, which is administered by the Organisation for the Prohibition of Chemical Weapons. This ongoing global concern amplifies the need for advanced technologies in chemical weapon disarmament, particularly in regions where such weapons are believed to be stockpiled. The chemical weapons presumably possessed by these suspected countries can be broadly divided into nerve and blister agents5. Nerve agents are highly toxic chemical compounds that disrupt the normal functioning of the nervous system, often leading to paralysis and potentially fatal outcomes. In contrast, blister agents are chemical substances that cause severe skin, eye, and respiratory tissue damage upon contact, resulting in painful blisters and other harmful effects6. Given the distinct application and processing methods for neutralizing each nerve and blister agent, a more comprehensive demilitarization approach is necessary7.

In scenarios where structural analyses of the ammunition cannot be performed or where internal analyses are difficult, analyzing the manufacturing state, technical characteristics, and preservation state of the ammunition becomes necessary. First, the selection of a disposal procedure requires the determination of whether the ammunition content is in the solid or liquid state. Understanding the state of the content is important for classifying the ammunition. For example, an ammunition content in the liquid state is a strong indication that the ammunition under observation is either a chemical or biological weapon. Chemical/biological weapons are characterized by leakage, spreading and concealment risks, rapid spreading, and ease of absorption; therefore, they must be cautiously handled during transportation and storage8. A liquid sensing technique can be applied to obtain relevant information for determining whether the ammunition content is liquid9. Liquid sensing is typically used to measure the liquid material volume in cans and other containers that store chemicals, fuels, or food in the liquid state in production processes to verify the material supply9,10,11. Several factors affect the measurement accuracy of liquid sensing, including the material and sealing of the container, viscosity and toxicity of the liquid substances, and explosive properties of the contained liquid, which should considered to ensure safety during inspection11,12. Radioactive13, electronic14,15, laser16, radar17, ultrasonic18, optical19, and hydraulic techiniques20 can be employed to measure the liquid properties such as substance identification. However, most of these techniques require direct contact with the liquid or the integration of the device into the container11.

Professional measurements and equipment are required for high-risk applications, such as the subject of this study21,22. The principle and method of level measurement in cases where the contents are flammable and explosive or have properties that are harmful to the human body have higher requirements because these materials must not interact with any other substance or be opened even for storage. Such substances must be stored away from other dangerous items such as chemicals, explosives, radioactive substances, and toxic substances that can cause chemical reactions, cross-contamination, or fires. Therefore, it is necessary to quantitatively identify internal information using a nondestructive method without physically damaging the ammunition. The structural analysis of unidentified ammunition is difficult because of the different manufacturing processes and raw materials used. The manufacturing status, technical characteristics, and preservation status of each ammunition type can be analyzed using radiographic or neutron techniques23,24. However, these techniques require appropriate protective gear and procedures, complex technology, considerable time, and large equipment. Numerous limitations hamper the portability of these methods for field application. Ultrasonic equipment, as compared with the abovementioned methods, offers the advantages of convenience and safety during usage, rapid installation and measurement, portability, and affordability9,25. Therefore, using ultrasonic equipment to determine the type and amount of contents by analyzing the signals received after transmitting ultrasonic waves through the ammunition has significant potential for easily acquiring internal information on ammunitions. Furthermore, the internal substances in ammunition can be used to identify the contents as a liquid or to classify it as a nerve/blister agent prior to the disposal process. The use of such systems can improve the efficiency and safety of the demilitarization process26,27.

A sensing model was developed in this study to determine the content of a sealed shell and to measure the liquid level by transmitting ultrasonic waves from outside of the container. The principle of the proposed technique is as follows: ultrasonic waves emitted from a transmitter attached to the outer wall of the shell penetrate the liquid chemical inside the shell, which is dependent on its propagation characteristics, and traverse the outer wall of the shell on the opposite side; the signal received by the receiver is then analyzed. Based on the differences in the ultrasonic sound velocity and acoustic impedance between the outer wall of the vessel and the liquid medium, the liquid level can be measured using the varying characteristics of the sound pressure considering the travel time calculated using ray tracing. This study proposes a method for designing a demilitarization process to safely neutralize chemical weapons by providing an effective identification process using an ultrasonic method to analyze possible chemical warfare agents (CWAs) sealed inside ammunition. Figure 1 provides a graphical illustration of the method proposed in this study.

Fig. 1
Fig. 1
Full size image

Graphical illustration of the method proposed in this study.

Results and discussion

Confirmation of the estimated time of arrival

Before comparing the different ultrasonic signal receiving times for each injected liquid in the absence of a burster, the estimated arrival times were calculated using Snell’s law. The estimated time values for water, chlorobenzene, and 1,2-dichloroethane were 93.9, 107.9, and 112.7 µs, respectively. The experimentally acquired ultrasonic signals were analyzed to verify the correspondence of the computed values to the actual measurements. Figure 2 shows the ultrasonic signals obtained when each simulant was injected into the mock ammunition under burster-free conditions.

Fig. 2
Fig. 2
Full size image

Ultrasonic wave signals for (a) water, (b) chlorobenzene, and (c) 1,2-dichloroethane in mock ammunition shells without bursters.

The waves transmitted through the wedge of the transmitter penetrated the ammunition shell, traversed the liquid, and re-entered the opposite shell, allowing the wedge of the receiver to analyze the ultrasonic signal. Because of the lack of interference from the burster, the amplitudes of the received ultrasonic signals following the desired path were clearly distinguishable, even to the naked eye. Moreover, the deviation between each acquisition time and the calculated expected arrival time was very small, and a clear time-of-flight difference was observed between the three types of liquids, facilitating a straightforward classification.

Quantitative and qualitative analyses of the internal contents of mock ammunition

Three different liquid samples were comprehensively analyzed in this study under three distinct burster conditions, resulting in nine datasets encompassing all possible combinations of liquids and burster conditions. First, the difference in the ultrasonic signals considering the amount of liquid was confirmed to verify whether the amount of liquid can be determined using the technique applied in this study, as shown in (Fig. 3). Because the highest amplitude values occurred at the designated angles of the two transducers corresponding to the expected paths, only the data for these angles are presented herein. The ultrasonic signal peaks received through the liquid are displayed in red, suggesting that the presence and level of the liquid can be estimated.

Fig. 3
Fig. 3
Full size image

Ultrasonic signals received at 2 mm intervals between the center of the transducers and the liquid in mock ammunitions containing: water with burster diameters of (a) 30, (b) 35, and (c) 40 mm; chlorobenzene with burster diameters of (d) 30, (e) 35, and (f) 40 mm; and 1,2-dichloroethane with burster diameters of (g) 30, (h) 35, and (i) 40 mm.

In most graphs in Fig. 3, when the liquid level aligned with the center position of the ultrasonic transducers (shown in the third signal in each set of individual signals), the ultrasonic signals traversing the liquid were immediately received. Although the peak amplitude values of all the signals received through the simulants were lower than those for water, the signals received through the liquids could be clearly visually distinguished. This is because ultrasonic signals received along a specific path through the projectile were consistently received and predictable regardless of the liquid level. Additionally, when the height of the transducers was sufficiently lower than the liquid level, higher amplitude values were observed. Signals received through the liquid often appeared weak, even when they were 2 mm above the water level; however, the peak amplitude values of these signals were very low. Therefore, applying an appropriate threshold for the axis of the peak amplitude can ensure an unambiguous measurement.

Figure 4 shows the data for the angles of the two transducers according to the liquid chemical being tested, indicating the time-of-flight when the signals were received and the positions of the two transducers at each specified angle. This result confirms that the burster condition and type of contained liquid can be predicted using the angle information between the two transducers and the time-of-flight.

Fig. 4
Fig. 4
Full size image

Time-of-flight of the ultrasonic signals as a function of the angle between the ultrasonic probes acquired using burster diameters of (a) 30, (b) 35, and (c) 40 mm in each liquid.

Comparison with existing methods

A comparison of the proposed method with existing techniques for managing unknown weapons is necessary to contextualize the findings of this study and highlight the contributions of the proposed method. As mentioned in the Sect. 1, conventional methods, such as X-ray and portable isotopic neutron spectroscopy, rely on radiation-based analyses, which can be challenging to implement in the field owing to the need for radiation-handling experts and specific protective facilities. In contrast, the proposed nondestructive ultrasonic testing approach offers several advantages. First, the proposed method is user-friendly and portable, making it suitable for field applications where the rapid assessment of unknown weapons is crucial. The ease of installation and measurement and the affordability of ultrasonic equipment make it a practical choice for on-site inspections. Second, the proposed method avoids the use of radiation, eliminating the associated safety risks and concerns. The proposed method provides a safer alternative for managing unknown weapons, particularly when chemical or biological agents that pose health and environmental risks are present. Finally, the proposed method can not only identify the presence of liquids; it can also potentially classify the internal substances, such as nerve or blister agents, providing valuable information for demilitarization processes. This capability enhances the efficiency and safety of managing unknown weapons.

Accordingly, the comparison of the proposed ultrasonic testing method with existing techniques emphasizes its relevance and contribution to the field of weapon identification and security. In summary, this study demonstrates that the proposed method can serve as a valuable tool in various scenarios, including warfare, military operations, and counterterrorism, where the rapid and accurate assessment of unknown weapons is essential for informed decision-making and safety.

Conclusions

An ultrasonic testing approach for the management of unknown weapons obtained from warfare, military operations, and terrorist situations was developed and investigated in this study. Conventional nondestructive evaluation methods, such as X-rays or portable isotopic neutron spectroscopy, are effective for identifying unknown weapons. However, their reliance on neutron- and radiation-based analyses hampers their field application owing to the need for radiation-handling experts and specific protective facilities. Therefore, the proposed ultrasonic testing method, which can identify the presence of substances contained within unknown acquired weapons and optimize the verifiability of their states, was developed and demonstrated to overcome these limitations. The experimental system enabled quantitative and qualitative analyses of the internal chemicals by analyzing the time difference between the acquired signals, particularly when the results identified the internal substances as liquids. The proposed approach is useful for both determining the state of the internal substance within a sealed body and for unconventionally identifying whether the chemical is a nerve or blister agent.

Extensive experiments were conducted using mock ammunitions containing different liquids with varying burster conditions, leading to the collection of datasets representing all possible combinations of liquid and burster conditions. Controlling the incident angle of the ultrasonic waves allowed the interference from bursters to be feasibly avoided and the expected penetration path to be predicted, which enabled the quantification and classification of internal liquid substances and provided valuable insights for the identification and management of unknown weapons.

The proposed ultrasonic testing method offers a promising solution for analyzing and classifying unknown weapons without the need for radiation-based analyses. The ease of use, portability, and affordability of this approach render it a useful tool for field applications, thereby reducing the potential risks associated with handling radioactive materials. Additionally, the ability of this method to identify substances, such as liquids, indicates new possibilities for a more accurate and comprehensive assessment of unknown weapons. This study contributes to the advancement of nondestructive testing methods in the field of weapon identification, with significant implications for enhancing security and safety measures in warfare, military operations, and counterterrorism. The results of this study support the adoption of ultrasonic testing as a viable and efficient approach for managing unknown weapons and facilitating timely and effective decision-making processes.

Materials and methods

Mock ammunition and ultrasonic transducers

The initial ammunition assessment and categorization phase for unidentified weapons necessitates determining the composition of its contents and distinguishing between the solid and liquid states. Therefore, understanding the structural makeup of ammunition is critical for determining the presence of solid or liquid materials. A solid content within weaponry is often associated with highly explosive bombs, whereas a liquid content is predominantly linked to chemical weapons. Notably, chemical weapons exhibit a distinctive feature wherein bursters responsible for generating explosions that are capable of rupturing the bomb casing and dispersing their contents over a broad area are arranged vertically from the bottom to the top at the center of the interior of the ammunition. Although exceptions may exist, the force generated by the primary explosive material in explosive bombs ruptures the entire casing, resulting in the dispersal of the bomb contents6,28.

The attachment of ultrasonic transducers to warheads is difficult owing to their predominantly conical shape. The bottom part is typically flat, but owing to its elongated length, a greater the amount of content increases the attenuation of the ultrasonic waves traveling back and forth to the surface. Additionally, if the munitions are upright, sensor attachment to the bottom requires an initial step of heavy lifting, which not only diminishes the advantages of the easily deployable ultrasonic technique in the field but also carries the risk of weapon activation. In addition to these constraints, side-mounted sensors are more conducive to assessing the fill level as they enable both quantitative measurements and qualitative content classification, even in situations where the contents are unknown. However, the presence of bursters significantly complicates ultrasonic fill-level measurements and content estimation. The burster obstructs the ultrasonic transmission path, causing reception interference. The cylindrical shape of the munitions obstructs the ultrasound beam, causing it to spread while preventing effective signal reflection and rendering the acquired ultrasound signal unfavorable for analysis. Considering these constraints and limitations, performing precise fill-level measurements and content classification in the presence of bursters using ultrasonic techniques are considerably challenging.

Nevertheless, ultrasound technology exhibits significant advantages, and research efforts to overcome these limitations are worthwhile. Controlling the incident angle of ultrasonic waves allows for the path of the transmitted waves to be directed and their expected path of penetration to be predicted. Consequently, the estimated time of arrival of the ultrasonic signals at the receiver can be determined. That is, sufficient information to predict the interior contents can be collected by adjusting the incident angle such that the propagation path of the ultrasonic waves is as far from the burster as possible.

The measurement procedure proposed to identify specific chemical weapons based on these conditions is as follows. First, a customizable jig capable of attaching transducers matching the curvature of the ammunition shell is prepared. The jig is then installed on the shell surface to adjust the position and angle between the transmitter and receiver. The jig is then slowly raised vertically while observing changes in the signal. Owing to the presence or absence of a burster and other internal conditions, the signal may not be received. Therefore, it is advantageous to adjust the transmitter-receiver angle and jig height to confirm the presence or absence of a signal. The height at which the signal first traverses the internal material and is received corresponds to the liquid level position, allowing for quantitative assessment of the liquid volume. Additionally, the angle between the two transducers and the time-of-flight information at the signal reception point can be used to infer the type of internal material. This measurement procedure is expected to enable precise and effective identification of the liquid quantity and type within the ammunition. The proposed method can be used as a core technology for the analysis of chemical weapons.

Rolled steel (SM-45 C) mock ammunition with an outer diameter of 155 mm and thickness of 20 mm was used in this study. Experiments were conducted using 30, 35, and 40 mm-diameter bursters to assess the performance and feasibility of the ultrasonic technique in the presence of differently sized bursters. Two 5 MHz ultrasonic transducers utilizing a 14 mm (0.55 in)-diameter Pb(ZrxTi1−x)O3 piezoelectric element were employed. A polymethyl methacrylate (PMMA) wedge was meticulously fabricated and attached to the transducer to conform to the curvature of the mock ammunition. Figure 5 illustrates the arrangement of the ultrasonic transducers for analysis of the internal liquid and classification of the mock ammunition. The red line in this figure represents the angle between the transmitting and receiving transducers.

Fig. 5
Fig. 5
Full size image

Position of the two ultrasonic transducers for the analysis of the internal liquid and classification of the mock ammunition.

The speed of sound in the ammunition material was approximately 5,740 m/s, whereas that in PMMA was 2,757 m/s. The incident angle of the ultrasonic waves on the wedge was determined using Snell’s law to be approximately 20.7°, enabling the transmitted wave to enter the ammunition at an angle of approximately 47.39°. An incident angle of approximately 47.39° was selected based on considerations such as the specific angle formed by the transmitting and receiving ultrasonic transducers. Incident angles greater than this value hindered the ultrasonic waves from propagating within the ammunition, transmitting them into the liquid medium. In contrast, incidence angles smaller than this value resulted in the transmitted wave exiting the ammunition and propagating through the liquid medium owing to refraction, leading to obstruction by the burster. The determined incidence angle of approximately 47.39° effectively balanced the requirements, thereby avoiding the burster while simultaneously allowing the ultrasonic waves to penetrate the ammunition and for relevant information to be acquired. Figure 6 illustrates the process of incident-angle selection to avoid the burster and provides a comprehensive explanation of the methodology employed in the experiment. The red lines in the figure represent the path traversed by the ultrasonic wave. The green dotted lines in Fig. 6a,c visually indicate the difference between the suitable angle (indicated by the black dotted line) and each unsuitable angle (smaller or larger than the suitable angle).

Fig. 6
Fig. 6
Full size image

Effect of various incident angles on the ultrasonic wave propagation paths: (a) small, (b) suitable, and (c) large.

Simulants (mock agents)

Conducting ammunition experiments using actual CWAs is not feasible for ethical, legal, and safety reasons. Substitute materials or chemical agent simulants (also known as mock agents) are commonly used for such experiments. This approach prevents the risks associated with CWAs and is internationally regulated to discourage illegal use. Accordingly, chemical agent simulants were used in this study. Ideal chemical agent simulants are typically chosen to mimic the chemical structures and physical properties of actual CWAs without their lethal and hazardous characteristics. However, the simulants employed in this study, representing several significant properties of CWAs, were chosen to appropriately conduct ultrasonic testing29,30.

Compressibility is the dominant property determining the speed of sound in liquid media. Low-compressibility liquids typically exhibit high sound velocities. That is, easily compressible liquids tend to have lower sound velocities, whereas those with limited compressibility tend to have higher sound velocities. Compressibility is related to the density of a liquid and is influenced by the molecular structure and intermolecular interactions of the medium. These properties play crucial roles in governing the speed of sound in a liquid.

In addition to compressibility, the physicochemical properties of actual CWAs were also considered. Because CWAs can be broadly divided into two major categories, namely, nerve and blister agents, the indirect identification of these two CWAs is necessary to confirm the feasibility of the proposed ultrasonic method. The effectiveness of the method proposed in this study was demonstrated by filling mock ammunition with two types of simulated agents representing nerve and blister chemical agents and by performing qualitative and quantitative ultrasonic analyses. The simulants used in this study were selected to simulate the physical properties of these two types of CWAs. Chlorobenzene (99.8%, anhydrous, Sigma-Aldrich, St. Louis, MO, USA) and 1,2-dichloroethane (99%, Samchun Chemical, Seoul, South Korea) were selected to mimic the physicochemical properties of the nerve agent GB (Sarin, O-isopropyl methylphosphonofluoridate) and the blister agent HD (Sulfur mustard, 1-chloro-2-(2-chloroethylsulfanyl)ethane), respectively. Table 1 lists the characteristics of the simulant substances and corresponding CWAs at a temperature of 25 °C.

Table 1 Properties of the simulants and their original CWAs at 25 °C31,32,33.
Full size table

Experimental setup

An ultrasonic pulser and receiver (HIS2, Krautkramer, GmbH, Cologne, Germany) were employed to generate ultrasounds and acquire ultrasonic signals. An oscilloscope (WaveRunner 640Zi, Teledyne LeCroy, NY, USA) was used for signal acquisition. A ring-shaped jig with a protractor was used to vertically scan the mock ammunition while maintaining a consistent angle. Custom-made covers with holes of the same diameter as the bursters were fabricated to accurately position the mock bursters at the center of the mock ammunition. Figure 7 shows a photographic image of the experimental setup.

Fig. 7
Fig. 7
Full size image

Photographic image of the experimental setup to analyze the internal liquid and classify the mock ammunition.

Initial experiments were conducted without bursters to compare the differences in the ultrasonic signal reception times for the three different liquids. Snell’s law was applied to estimate the ultrasonic paths and calculate the estimated time of arrival for each liquid. Subsequently, the actual reception times were validated to confirm their correspondence with the estimated values. This analysis served as the basis for predicting the expected arrival times of the ultrasonic signals in the presence of bursters. The presence of bursters disrupts the ultrasonic paths, causing discrepancies in the estimated time of arrival. However, quantitatively determining the exact impact of round bursters on wave propagation is challenging. However, despite the disruptive influence of bursters and considering that each liquid exhibited distinct and sufficient time differences, determining the type of liquid present inside the ammunition is still feasible. Furthermore, to assess the exact amount of liquid present inside the ammunition with unknown contents, vertical scanning was performed in 2 mm increments while monitoring the ultrasonic signals. Anticipating that the difference in the angles between the transmitter and receiver would lead to amplitude variations in the received signals, a thorough analysis of the signal differences was conducted while systematically changing the angle in 5° increments.