Cost effectiveness and feasibility considerations on the design of mini-UAVs for balloon takedown. Part 3: reliability and availability

Abstract

The reliability and availability of a Jet Unmanned Aerial Vehicle (JUAV) for balloon downing are evaluated in this third part of the paper. The paper delves into the challenges faced by traditional air defence systems in countering high-altitude balloons, leading to the proposal of substituting missiles with Unmanned Aerial Vehicles (UAVs), specifically the Jet Unmanned Aerial Vehicle (JUAV). The best solution found in the previous section is a derivation of the 1:6 scale RC model of the Lockheed F104C. This is designed for balloon downing, and the essay meticulously analyses its reliability, failure probabilities, and maintenance strategies for various subsystems. The engine is replaced with a more powerful micro-jet, and an 84 mm recoilless cannon is added at the base of the tail. The scaled-down model can use most data derived from the full-scale airplane flight manual, corrected for larger thrust-to-weight and lower wing-loading. Several adjustments and additions are made on sensors, autopilot, On-Line Diagnostic System, communication system, and firing system to increase the reliability of the system. Due to the relatively short design life of the RC model, the Time Between Overhaul (TBO) of the UAV is reduced to 25 h. Consequently, the study emphasizes the importance of simplicity, operational reliability, and adaptability in the development and assessment of modern missile defence systems. The reliability of the JUAVI is presented as 98.14% for 50 consecutive fully automatic missions, highlighting its potential significance in bolstering national security.

Keywords:

• Reliability
• availability
• UAV
• re-usable
• cost-effective
• balloon killer

Reviewing editor:

• D T Pham
• University of Birmingham
• UK

Subjects:

• Aerospace Engineering
• Mechanical Engineering
• Mechanical Engineering Design
• Mechanics
• Transport & Vehicle Engineering
• Design

1. Introduction

The world has been astonished by China’s recent audacious forays into U.S. airspace, employing outdated technology in the form of large-payload balloons for espionage purposes (Bera, 2023). In February 2023, the U.S. Air Force intercepted and destroyed a Chinese surveillance balloon that had traversed multiple states along the South Carolina coast. This event prompted inquiries into the extent of the U.S. government’s awareness regarding the entities operating within its expansive airspace (Posard et al., 2023). However, the fact that this technology is still in use and being updated is not a coincidence. During the First World War, observation balloons were employed on the front lines by all parties, additional balloons were utilized at sea, and balloon barrages along with balloon aprons were employed to defend high-value targets against air attacks (Bishop & Limpaecher, 2021). When we observe current technology, there are improvements in balloons technology. In August 2015, three aerostatic balloons, equipped with high-definition cameras utilizing military technology, were deployed for surveillance purposes in the municipalities (Tironi & Valderrama, 2021).

Sky balloons face challenges in their ability to evade traditional air defence systems due to their operational capacity at elevated altitudes (Piancastelli et al., 2023). Long-duration balloons can fly at heights exceeding 20 km (or 65,616.8 ft), enabling extensive surveillance over both land and water (D’Oliveira et al., 2016; Sóbester et al., 2014). Consequently, they are ideal for gathering intelligence and monitoring military operations, critical infrastructure, and other sensitive locations (Lloyd et al., 2016). Designed to remain airborne for extended periods, modern balloons can carry cameras, signals intelligence equipment, and other sensors (Darack, 2019), along with a payload capacity. In addition, balloons can elude many common air defences systems by flying at great altitudes, making them challenging to detect and engage. These capabilities pose a serious security concern. Attackers can use various weapon systems to maintain a state of conflict in the air littoral, which is located below 15,000 feet. These weapons include man-portable air defence systems, radar-guided anti-aircraft artillery, cruise missiles, dual-purpose drone technologies, and loitering munitions (Singh, 2022). Similar threats are also present at the highest altitudes of the air domain. Military and spy balloons have a long history of use at very high altitudes. The U.S. conducted several spy balloon missions over the USSR in the 1950s, and more recently, nationwide tests of mass surveillance balloons were conducted in the U.S. Thanks to a combination of technological advancements and business strategies, modern guided balloons can now enter and remain in the space littoral on a budget. Adversaries could use drone swarms or missiles fired from high-altitude balloons to attack known targets. In 2018, a test of a high-altitude balloon carrying hypersonic missiles was covered by Chinese state media (Col et al., 2023). The fact that balloons have a very small radar cross-section, and are therefore difficult to detect and destroy, could make them a persistent threat to airborne systems operating below them. Even if found, high-altitude balloons would still be challenging to deal with. In 2023, a surveillance balloon, most likely costing thousands of dollars, was shot down by a $250 million F-22 fighter armed with a $472,000 AIM-9 Sidewinder missile (Grieco & Bremer, 2023). If the adversary deployed hundreds of these balloons, traditional defence systems would easily be saturated.

There are many situations illustrating the use of missiles in destroying balloons in the military. For instance, Le Prieur missiles, which were unguided, were utilized to eliminate balloons within a distance of 100–150 m (Gibbons & Botha, 2015). As anti-missile technology evolves, there is a growing focus on cooperative multi-missile attack and defence. Researchers are now exploring the application of consensus theory to investigate the cooperative guidance and control of multiple interceptors, benefiting from advancements in multi-agent systems consensus (Kim, 2021).

In this study, the idea is to substitute UAV for missile in the military applications. The term UAV (unmanned aerial vehicle) is used to describe a reusable airborne vehicle that operates without a pilot on board. On the other hand, a ‘missile’ denotes a single-use vehicle without an on-board pilot. In the initial era of unmanned aircraft, there was no clear differentiation between UAVs and missiles. The advent of World War II spurred various innovations and trials in pilotless aircraft, emphasizing distinctions between missiles and UAVs. The pursuit of precise bombing has been a driving force behind missile advancements over the past six decades. Drawing upon enhancements in guidance and control derived from cruise missiles, as outlined in Newcome, UAVs evolved to become more reliable and autonomous. These technical enhancements were accompanied by shifts in nomenclature and conceptual understanding (Sullivan, 2006).

The developments in technology by improving the control systems of the UAV can increase the reliability for these kinds of solutions. Rapid advancements in wireless networking technologies have led to its widespread utilization in various applications, including disaster surveillance, military, civil, search and rescue (SAR) operations, disaster management, remote monitoring, and more (Khan et al., 2021, 2022a). There are current studies that prove UAVs collaborate with cloud technologies to bring about improvements.

A study provides us information about a pragmatic approach for the three-dimensional positioning of a cluster of HetNet (Heterogeneous Network) unmanned aerial vehicles (UAVs) equipped with mounted base stations (mBS). The study aimed to provide optimal wireless connectivity and coverage for terrestrial users in a specific area (Khan et al., 2022b). Another focused area is the collaboration between the UAV and Internet of Things (IoT). The Internet of Things (IoT) represents a swiftly evolving and advanced landscape where the central idea revolves around coordinating a diverse range of intelligent objects to enable their global utilization and operation. The simulation outcomes indicate that, in comparison to alternative trajectory planning algorithms, such as Dijkstra’s and HEA, the proposed technique achieves significant energy savings and time reduction (Banerjee et al., 2022). Another important technological improvement is the collaboration of UAV-Fog architecture within a four-tier network comprising smart devices, local UAVs, UAV-Fog, and a cloud server. It aims to manage UAV data and discusses security challenges associated with this cloud infrastructure (Gupta & Gupta, 2022). In an aerial network, limitations on swaps can restrict various unmanned flying devices, such as UAVs, spacecraft, or balloons, to specific operational altitudes. In general, a UAV is equipped with transceivers to establish a network with ground clients, providing flexible internet access, and the airspace around the unmanned aircraft serves as the corresponding broadcasting region (Wani et al., 2022). Overall, these advancements contribute to enhanced connectivity, energy efficiency, and effective management of UAV data, addressing various challenges and limitations in the field.

The aim of this study is to evaluate the reliability and availability of a Jet Unmanned Aerial Vehicle (JUAV) designed for balloon downing. The paper discusses the development of a scaled-down model based on the Lockheed F104C, emphasizing modifications, such as replacing the engine with a more powerful micro-jet and adding an 84 mm recoilless cannon. The essay provides a detailed analysis of reliability, failure probabilities, and maintenance strategies for various subsystems, including the airframe, jet engine, communication, autonomous control, manual control, fuel system, battery pack, and cannon system. The authors emphasize the importance of maintaining the JUAVI's simplicity, with a proposed maintenance strategy involving the replacement of critical components every 25 h. The reliability analysis considers factors, such as Time Between Overhaul (TBO), Mean-Time-Between-Failure (MTBF), Mean-Time-to-Repair (MTTR), and Scheduled Maintenance (SM). The authors also address the challenges associated with the mission, such as the need for real-time communication, authorization to shoot, and the impact of high-altitude conditions on equipment cooling.

2. Materials and methods

2.1. Summary of cost effectiveness and feasibility considerations on the design of mini-UAVs for balloon takedown. Part I: weapons and mission and part II: aircraft design approach selection

The previous studies by the authors (Piancastelli et al., 2023), introduced small Jet Unmanned Aerial Vehicle Interceptors (JUAVI) equipped with an 84 mm recoilless cannon that can fly up to 28,956 m (95,000 ft) at 1.6 M. The use of the airburst shell of this cannon is more cost-effective than missiles (which are too expensive) and machine guns/automatic cannons (which are ineffective). Small off-the-shelf single-shaft jet engines can be improved to static thrusts up to 1200 N. This propulsion system, equipped with a variable geometry nozzle and a last-generation undirected air intake, maintains 10% of the static thrust at 28,956 m (95,000 ft). The weak point of this engine is the relatively low efficiency of 127 [gr N⁻¹ h⁻¹]. This best value is achieved at full throttle and decreases rapidly with partial loads. These jet engines are equipped with Full Authority Digital Electronic Control (FADEC) that controls the starting procedure and throttle. Altitude compensation can be programmed inside the tiny CPU of the FADEC. The control and systems are easily installed in the JUAVI. For real-time radio control, the maximum practical flight range is 320 km. Even if the JUAVI may be equipped with an autopilot, the eventuality to fly in Air Traffic Controlled (ATC) regions and the necessity of real-time shoot authorization impose a continuous connection with a ground station. The best JUAVI is a 1:6 F104C. The JUAVI will not have a landing gear and can be catapulted or can take off with a 4 m ski jump equipped with a jettisonable tricycle gear-style trolley. Alternatively, it can be launched from an airplane like the Lockheed C130. The typical mission would begin by turning on the electronic control and communication system of the airplane. This hardware has its own Electronic on Board Diagnosis System (EOBD) that will run continuously during the flight. The ground station will then input the necessary data into the Flight Control System (FCS). This operation can also be done after take-off. If properly authorized by the Air Traffic Controllers (ATC), the JUAVI will make a 0.9 M nearly vertical climb to approximately 12,192 m (40,000 ft). A dive down to 11,277.6 m (37,000 ft) will help the transition to supersonic speed and the acceleration up to 1.6 M. Then the climb will be resumed up to the required altitude to reach the target. The JUAVI will reach a position above the target and dive into the balloon. In fact, it is convenient to be fast to have a more stable shooting platform and to damage the top of the envelope. In this way, the helium, which is lighter than air, will naturally float to higher altitudes through the hole(s), deflating the balloon. The JUAVI would then dive toward the balloon and use the laser range finder to find the distance from the balloon. As the distance goes under 1000 m, it is possible to feed the time-to-burst into the shell and shoot. The rear blast will be avoided due to the faster-than-sound speed of the airplane. The nozzle/shell blast is a pressure cone that will be superimposed on the JUAVI one without adding much disturbance. As the cannon is fired, the JUAVI will slow down due to the increased drag caused by the loss of the aerodynamic plug installed on the nozzle. At this point, the JUAVI will make a steep turn upwards to avoid the High Explosive (HE) round blast. Then it will fly to the arresting gear of a rescue station (trap) and will remain suspended on the wire. The JUAVI will then be refuelled and reloaded for a new mission. For reliability, availability, and cost-effectiveness, it is convenient to keep the design of the JUAVI as simple as possible.

2.2. 1:6 F104C cannon

The authors found that the most promising JUAVI is the downsized F104C. The ideal scale factor is 6. This means the JUAVI has a size that is 1:6th the original airplane. Table 1 summarizes the main data of this JUAVI.

Table 1. UAVI basic data with F104C (*) as a full-scale reference airplane (Figure 1).

Parameter	F104C (F-104 A, B, C & D Flight Manual, 1968)	F104 1:6 single cannon
Length (m)	16.66	3.3
Wingspan (m)	6.63	1.1
Wing area WA (m^2)	18.22	0.5
Empty weight (kg)	5788	24
MTOW (kg)	12634	52
Half Internal Fuel Mass (kg)	1138	8
Reference mass (kg)	6926	32
Wing loading at reference mass (kg m^-2)	380.1	64
Maximum static thrust dry (N)	44,500	1251
Maximum static thrust with afterburner (N)	70,300	–
Trust loading-reference mass dry (N kg^−1)	6.42	39
Trust loading-reference mass with afterburner (N kg^−1)	10.1	–
Maximum speed (M)	2.2 M at 15,240 m (2.2 M at 50,000 ft)	1.6 M at 28,956 m (1.6 M at 95,000 ft)(***)
Ceiling	17,678.4 m (58,000 ft)	28,956 m (95,000 ft)

^(***) Limited by the high-altitude fixed-geometry air intake.

2.3. Availability, reliability, MTBF, MTTR, and TBO

It is necessary to clarify a few definitions since different parameters are referred to by the same name in the literature. In this paper, availability is a characteristic of items or systems that can be kept operational through maintenance or upgrade. It depends on parameters, such as Time Between Overhaul (TBO), Mean-Time-Between-Failure (MTBF), Mean-Time-to-Repair (MTTR), and Scheduled Maintenance (SM). Reliability is the probability of fulfilling the mission. The JUAVI considered is the 1:6 F104C equipped with the Carl Gustav 84 mm recoilless cannon. It can be subdivided into nine main sub-assemblies (Table 2): the airframe with the main servos (Yaw, Pitch, Roll) and the fixed air intake (1); the jet engine section equipped with FADEC (Full Authority Digital Control) and nozzle (2); the communication package (3); the autonomous control package with the basic sensors (4); the manual control system with additional backup sensors (5); the rescue package with the hook and servo (6). The emergency parachute and its ejection system (7); the fuel system with tanks (8); the battery (9), and the cannon pack (10). (1) $$R_i=e^{-\frac{\mathit{\text{Life}}}{\mathit{\text{MTBF}}}}\mathrm{ }$$(1) (2) $$F_i=1-R_i\mathrm{ }$$(2) (3) $$R_{\mathit{\text{Total}}}=\prod _{i=1}^n{F}_i\mathrm{ }$$(3) (4) $$F_{\mathit{\text{Total}}}=\prod _{i=1}^n{R}_i$$(4)

Table 2. UAVI sub-assemblies list.

Sub-assembly	Position	Reference number (#)
Airframe	Entire airplane	1
Jet engine	Rear	2
Communication	Centre-nose	3
Autonomous control	Centre	4
Manual control	Centre	5
Rescue system (trap)	Centre	6
Emergency parachute	Rear	7
Fuel system	Centre	8
Battery pack	Nose	9
Cannon pack	Rear	10

Where R_i is the reliability of the i component. F_i is the failure probability of the i component. R_total is the total reliability of n components in a serial arrangement (Lusser’s Law). F_total is the failure probability of n components in a parallel arrangement when one single component will perform the required function. Life represents the life in hours of the assembly.

3. Failure probability

The reliability analysis starts from the final condition of the mission: the take-down of the balloon (Black, 2017). For this purpose, it is necessary that the HE shell explodes within a few meters’ distance from the balloon to pierce and tear the balloon envelope. The airburst will produce a huge pressure wave and a pattern of fragments that will dig a hole in the envelope. The damage should take place on the top of the balloon to be effective. In fact, helium is lighter than air and will naturally escape from the holes in the top of the envelope. Side holes are acceptable since helium will find its way outside the damage envelope, even if the deflation speed is reduced. Bottom holes are ineffective since the helium will remain inside the balloon for a long time. The life of the Carl Gustav cannon is 500 rounds. The carbon fibre-reinforced cannon cannot be replaced in the JUAVI airframe, and it limits the JAVI life to 500 missions. Each mission lasts a maximum of 30 min; therefore, the JUAVI life is limited by the cannon to 250 h. The firing procedure is as follows: the rangefinders measure the distance of the JUAVI from the balloon. The system acquires consecutive sequential distances. If the number of acquisitions below 1000 m and closing is sufficient, they are fed into the ballistic software. In this way, the time to burst is calculated and fed to the shell, which is fired immediately. The reliability of the process depends on the rangefinder, the ballistic computer/software, and the shell. Another condition for a successful mission is that the JUAVI at the firing position receives the ‘permission to shoot’. This fact depends on the communication system, which is an airplane-radio-communication unit. The airplane radio reliability at 250 h is very low: 0.73 (Pettit & Turnbull, 2001). This is a problem to be addressed. The airplane should arrive at the target. For this purpose, it will need to use all the subsystems of Table 1 apart from sub-assemblies reference numbers (6) and (7). Starting from the rear of the JUAVI, we have the jet engine (2) with the fixed or variable geometry nozzle. The variable geometry nozzle is a 3D print made by laser sintering from metallic powder. The unit may be quite expensive since hot hyping is needed for quality control. However, the weak point for reliability is the single servo system that controls the nozzle opening. The servo reliability will be discussed later. The jet engine has a TBO of 25 h. After this time, the ball bearings should be replaced. The replacement of the turbine is a depot maintenance operation that requires approximately 6 h, all included. Therefore, 25 h is the actual flight limit of the JUAVI without maintenance. The jet FADEC requires a triple redundancy (Kumar & Kumar, 2023). It should be noted that it is possible to control the failure figure by evaluating engine degradation before failure. This can be done by calculating the thrust from speed and air data and by positioning an accelerometer on the bearing stations. Theoretically, it is also possible to evaluate fuel consumption, but this will require the installation of a fuel meter. In addition, the FADEC comes with sensors and monitoring software that will emulate failed sensors and provide a history log of the flight. This will enhance engine reliability by controlling anomalies after the flight. While thrust calculation is inexpensive, monitoring bearing and fuel consumption would require time and skilled operators. The fuel system is composed of a main tank with a secondary ‘aerobatic’ tank with an air trap. The fuel pump will work only during engine start. In fact, the main fuel tank will be pressurized with bleed air taken from the compressor of the jet engine upstream of its fuel-burning section. Since the engine will run most of the time at full throttle, the pressure regulation system is simple. A fuel-to-air heat exchanger can be used to cool down the bleed air. In this way, the fuel fed into the engine will be preheated with beneficial effects on the overall efficiency of the jet propulsion system. When the throttle is ‘pulled back’, the pressure valve on the fuel line after the filter will automatically restart the electric pump. This fact is an anomaly in the climb phase and will probably signal a main problem in the fuel system. In fact, the engine should always work at full throttle to keep thrust at altitude. Speed can be reduced only with the airbrake. Therefore, the throttle is pulled back only in the descent phase of the mission, after the single shell has already been fired. For this reason, a fuel pump failure would affect the mission only at engine start. With a 10-min work time per mission, the fuel pump will have an MTBF of 80,000 h (SAFRAN, 2020). A heat sink in the fuel tank should be used for the equipment that operates continuously at high altitudes. The roll, pitch, yaw, radio, controller, etc., would require cooling to function at such altitudes. The reliability of these cooling systems is included in the reliability of the other subsystems. The airframe reliability depends on the roll, pitch, and yaw servos and control surfaces. As discussed in paragraph 3.1, servo reliability is 0.99999967478861961906 for each servo (Life = 25 h). This implies a reliability of 0.99999999349577139979 for each mission (0.5 h). The servo failure probability is assumed to be 0.0013 in 100,000 h (Fortna Murtha, 2009). The control surface system reliability is 0.0002 (Fortna Murtha, 2009). For the nozzle servo and the nozzle assembly, the same reliability figures were adopted. Table 3 summarizes the airframe (1) reliability. A structural failure of the JUAVI is not considered, as it is a remote possibility statistically not relevant for a life of 250 h with Non-Destructive Controls (NDC) conducted every 25 h. The overall reliability figure is calculated using Lusser’s law. Table 4 summarizes the jet engine reliability. The jet engine with the alternator has an MTBF of 50,000 h. This is because every 25 h, the engine undergoes a full overhaul where it is disassembled, the bearings are replaced, and all parts are thoroughly inspected with non-destructive (ND) methods, after which the engine is reassembled. The starter/generator is also replaced.

Table 3. Airframe (1) reliability.

Sub-assembly	Life reliability (25 h)	Mission reliability (0.5 h)
Structure	1	1
Rudder servo	0.99999967478861961906	0.99999999349577139979
Rudder	0.99999994999500063647	0.99999999899989999713
Elevator servo	0.99999967478861961906	0.99999999349577139979
Elevator	0.99999994999500063647	0.99999999899989999713
Left aileron servo	0.99999967478861961906	0.99999999349577139979
Left aileron	0.99999994999500063647	0.99999999899989999713
Right aileron servo	0.99999967478861961906	0.99999999349577139979
Right aileron	0.99999994999500063647	0.99999999899989999713
Variable nozzle servo	0.99999967478861961906	0.99999999349577139979
Variable nozzle	0.99999994999500063647	0.99999999899989999713
Airframe (1) total	0.9999981239195904613	0.9999999624783575803

Table 4. Jet engine (2) reliability.

Sub-assembly	Life reliability (25 h)	Mission reliability (0.5 h)
FADEC (3-time redundancy)	0.99999999655165490476	0.99999999999997235187
Air to fuel heat exchanger	0.99999999655165490476	0.99999999655165490476
Fuel pump	0.99989583875849218034	0.99999791666883680404
Pressure regulator for air bleed	0.99989583875849218034	0.99999791666883680404
Jet assembly with alternator	0.99975003124739599608	0.99999500001249996822
Jet engine (2) total	0.99954176479181383610	0.99999082992700594945

In Table 4, the fuel pump and the air pressure regulator have the same reliability. The FADEC has the average reliability of a car injection system with an Electronic Control Unit (ECU) and sensors (Knight et al., 2001). It is assumed that a billion kilometres are equivalent to 18 million hours; this means that an average car takes 4500 h to cover 250,000 km. The MTBF of a car injection system is 16,535 h. The radio communication equipment (3) needs a three-time redundancy to achieve acceptable mission reliability (see Table 5).

Table 5. Communication package reliability.

Sub-assembly	Life reliability (25 h)	Mission reliability (0.5 h)
Comm. package (3) total	0.9999794392580720352	0.99999999982867504572

The backup system is a CAT II autopilot that can execute the mission if the shoot is authorized. If communication problems arise, the autopilot will guide the airplane to the landing site, but the mission fails. The autopilot system fulfills other tasks that are crucial to improving the reliability and maintainability of the JUAVI. In fact, it will run pre-flight checks and command the FADEC to start the engine. In addition, it will run a diagnostic system and send warning or caution information to the ground station. It will also produce a log file of the flight with all the anomalies detected during the flight. The reason we are considering manual control of the JUAVI for the reliability assessment is that the reliability of the JUAVI on manual control is higher than on autopilot. The autopilot MTBF is assumed to be 31,250 h. The reliability of this data, which comes from a blog, is dubious, but in our case, the autopilot is used only as a backup and monitoring system. Table 6 summarizes the Autonomous control (4) subsystem reliability data.

Table 6. Autonomous control (4) reliability.

Sub-assembly	Life reliability (25 h)	Mission reliability (0.5 h)
Autonomous control (4) total	0.99920031991468373060303	0.99998400012799926179

Table 7 summarizes the reliability data from the other subsystems. The data are derived from reference (Fortna Murtha, 2009). The fuel system reliability is that of the whole propulsion system of reference (Fortna Murtha, 2009).

Table 7. Sub-assemblies reliability total.

Sub-assembly (#)	Life reliability (25 h)	Mission reliability (0.5 h)
Manual control (5)	0.99999999749998750304	0.99999999749998750304
Fuel system (8)	0.99998717675862158755	0.99999974353356091411
2-Times redundant battery pack (9)	0.99999999998999999980	0.99999999999980004878

Table 8 summarizes the reliability data for the cannon system, which is composed of the cannon, the shell, the shell retaining system, the fire system, the time-to-target feed system, the ballistic computer (software), and the range finder(s). The laser-range finder will have an MTBF of 11,000 h at 40 °C (system powered and firing 25% of the time at 1 Hz Firing Rate). Therefore, it is not necessary to install a redundant system. The ballistic computer is software installed on the autopilot and in the ground station (for redundancy). This software will acquire a defined number of airplane-to-balloon distances below 1000 m to define the airplane trajectory and will control the JUAVI to compensate for ballistic parameters (wind, shell trajectory, etc.). If authorized, this software will fire the cannon at the right instant and will feed the time-to-target into the shell. The cannon has an electric priming system. The reliability of the software is considered unitary. The shell, the cannon assembly, and the firing system have a 500 h and 500 rounds MTBF (US Department of Defence data). In this case, it is not possible to improve the cannon reliability. In fact, the cannon is a commercial item.

Table 8. Cannon system (10) reliability.

Sub-assembly	Life reliability (25 h)	Mission reliability (0.5 h)
Cannon, shell, firing system	0.98347145382161748947	0.9996667222160499077
Laser range finder	0.99924271113048742165	0.9999848485996321034
Total	0.98272668183615502487	0.9996515758653071495

It is assumed that the cannon system will work for only 1/3 of the mission time. Multiple ‘heat sinks’ are frequently used by modern airplanes. Cooling is a significant problem at high altitude, where the air is so thin that the mass flow becomes minimal. Ram Air (RA) can be used for the equipment racks or separate cooling at altitudes lower than 12,192 m (40,000 ft) via protruding ram scoops or flush inlets (Piancastelli, 2023). Due to the thinness of the air, this solution cannot be used in high-altitude aircraft. Heat sinks are often used for high-altitude flying. Liquid gases that can absorb heat at normal/low temperatures (below the water freezing point) and endothermic chemical processes are examples of expendable heat sinks. When other options are impractical during flight phases, expendable heat sinks are used. For example, should-fired missiles have a supply of cryogenically cooling using high-pressure gaseous argon injection that unfortunately lasts less than a minute. Therefore, it is impractical for the JUAVI. Positioning an external duct at the main fan’s (Engine Fan Air-EFA) exit is another typical fix; however, this solution is also impractical due to the ‘pure’ jet propulsion system of the JUAVI. The Skin Heat Exchanger (SHX) rejects thermal load heat to the environment directly through the aircraft fuselage. Again, this solution is impractical due to the high temperature linked to the high speed of the JUAVI. High speed for the JUAVI is convenient at low altitudes to avoid problems with winds and turbulence linked to the very low wing loading of the downsized F104C (see Table 1). At high altitudes, the air intake works as a compressor to increase the air density at the intake of the jet engine. As a result, SHX is not useful for JUAVIs. For a very long time, fuel was used by aircraft as a heat sink. A fuel-oil heat exchanger is frequently used to cool aircraft. Recirculation systems are installed if the fuel flow required for cooling is greater than the fuel flow demand of the engine. This is especially true in low-thrust flight conditions like descent and taxi. In addition, considerable effort has been put into improving the fluid properties of jet fuel to increase its heat capacity and Thermal Management Systems (TMS). In the case of the JUAVI, most of the flight is done at full throttle, due to the very low efficiency of the jet at partial loads. In the case of our JUAVI, Peltier-effect cells can be used to transfer the heat accumulated in servos, sensors, electronic cards to the fuel tank. A Peltier effect cell has a MTBF of 100,000 h. Typically, one cell is used in the cooled equipment while another is sunk in the fuel tank. A thermostatic controlled electric energy supply will transfer the heat when necessary. The thermostatic control will also have an MTBF of 100,000 h. The single heat sink will have a total ‘life’ reliability of 0.99999999956570510832 and of 0.9999999999999965218 for the single mission. The heat sink can also be used to heat the equipment. It is probable that the firing system will need to be heated during the mission.

3.1. Servo reliability, TBO, and maintenance strategy

Servos are an extremely critical component in airplanes. RC servos are known to be not very reliable and are usually a weak point in the system’s reliability. To improve servo reliability, two different strategies are possible. The first one is to purchase ‘industrial’ servos provided by most RC servo manufacturers. A second strategy is to purchase high-quality RC servos with steel gearings. It is then possible to ‘exercise’ the servo by cycling it at limited load (1/3 maximum load). This will improve the fatigue life of the gears. Another strategy is to make the servos work at the maximum allowed temperature for 24 h, as was done for the Apple Macintosh and most Formula 1 Electronic Control Units. These strategies will eliminate infancy mortality. The design life of an RC servo is about 25 h. It is then possible to replace the servos on the JUAVI every 25 h, which is also the life of the microjet bearings. The JUAVI will, therefore, fly 25 h (50 missions) without maintenance, relying on the Electronic On-Board Diagnostic (EOBD) of the Autopilot and the FADEC. It will then necessitate Depot level Maintenance for a full overhaul that will keep only JUAVI structure and cannon, which will be controlled with Non-Destructive (ND) methods. Every other component will be replaced. As we will see in the next paragraphs, this strategy will ensure the necessary reliability and availability for a successful design. The operator must perform a complete ‘walkaround’ before each flight, and the Autopilot and the FADEC will perform the basic checks in connection with the ground station. After every flight, an automatic check of the FADEC and Autopilot log history will also be performed. Before every take-off, the batteries will also be checked. Battery replacement is an on-site maintenance.

4. Manual mission reliability

For a manually controlled ‘Balloon-killer’ mission, the JUAVI will need the aileron-rudder-elevator servos, the laser range finder, the firing system, the three communication packs, the engine FADEC, the manual receiver, and 11 Peltier’s heat rejection units. The reliability for the JUAVI is 0.982241719906063273 for life (25 h) and 0.99964211243382391 for a single mission (0.5 h). This means an MTBF of 1396 h. In this way, the JUAVI will be able to kill the balloon, but the return is not included. It is the mission reliability of a missile or a Kamikaze airplane.

4.1. Complete manual mission reliability

To complete a manually controlled ‘Balloon-killer’ mission, the JUAVI will also need the rescue system (6). The rescue system reliability is shown in Table 9. It is assumed that the JUAVI will collimate the camera to the landing aim. As the ‘target-wire’ is sufficiently close, it will deploy the airbrake, the hook, and the parachute. In this way, the approach can be done at a sufficient speed, and the ‘final approach’ will be with the largest reduction of the speed possible.

Table 9. Rescue system (6) reliability.

Sub-assembly	Life reliability (25 h)	Mission reliability (0.5 h)
Airbrake servo	0.99999967478861961906	0.99999999349577139979
Airbrake	0.99999994999500063647	0.99999999899989999713
Parachute deployment system	0.99999994999500063647	0.99999999899989999713
Hook servo	0.99999967478861961906	0.99999999349577139979
Hook	0.99999994999500063647	0.99999999899989999713
‘Pilot error’	0.99999238522712130534	0.99999984770397420947
Rescue system (6) total	0.9999915847956684401	0.9999998316952195228

The ‘pilot error’ takes into account the uncertainties in the ‘trap’ process, due to the fact that the last part is performed at very low speed with the very low wing loading of the JUVI (see Table 1). This pilot error is the highest reliability data for pilot error of reference (Fortna Murtha, 2009). Airbrake, parachute, and hook have the same reliability as control surface mechanisms of reference (Fortna Murtha, 2009). The full-mission reliability for the JUAVI is 0.982233454141287280 for life (25 h) and 0.99964194418927762 for a single mission (0.5 h). This means an MTBF = 1390 h. The availability of the JUAVI for 50 consecutive manual missions is 98.22%, or the probability that a single JUAVI will not perform 1 mission every 25.

4.2. Manual and automatic complete mission reliability

The reliability of the ‘Balloon-killer’ with the autopilot system fully functional [Autonomous control system (4)] is 0.981447981608879082 (Life = 25 h) and 0.99962595004612402 (Mission = 0.5 h). The availability of the JUAVI for 50 consecutive fully automatic missions is 98.14%. The MTBF is 1336 h. This data is based on an MTBF of the autopilot that comes from an uncertain source.

5. Results

The study delves into the failure probability analysis, focusing on the key requirement of taking down a balloon. The mission’s success hinges on factors, such as the reliability of the rangefinder, ballistic computer/software, and the shell. The limitation imposed by the cannon’s 500 rounds life expectancy is discussed, impacting the overall JUAVI lifespan, capped at 250 h. In-depth analysis of each subsystem’s reliability is presented, addressing issues related to the jet engine, communication equipment, autonomous control, and the cannon system.

The reliability analysis extends to subsystems, including airframe, jet engine, communication, autonomous control, and the cannon system. Tables provide a comprehensive summary of the reliability data for each subsystem, aiding in understanding the intricate network of dependencies. The discussion on high-altitude challenges introduces potential solutions, such as heat sinks and the use of Peltier-effect cells for thermal management. Servo reliability is addressed as a critical element, with strategies proposed to enhance their performance and longevity. It must be outlined that a maintenance strategy, emphasizing the importance of regular checks, diagnostics, and replacement intervals to ensure sustained reliability. The manual mission reliability section assesses the JUAVI's capability to accomplish a ‘Balloon-killer’ mission, considering various components and their reliability. The inclusion of a rescue system further enhances the mission’s success, demonstrating a holistic approach to reliability.

The concept of reliability brings us to the point of evaluating the success of the system produced based on the proportion dependent on its usage. Some missiles that are currently in use and have been tested, along with their capacity details, are provided in Table 10.

Table 10. Comparison between the current missiles and the system offered in this study.

	Success rate (%)	Capabilities
Russian air-launched cruise missile	40–80	The missile can cover a distance of 4500 km. The Russian Air Force employs the Kh-102 missile in conjunction with the Kh-101. The Kh-BD is a forthcoming cruise missile anticipated to have a range exceeding 3000 km.
North Corean ballistic missile	85	This missile has a range of 4500 km, making the US island of Guam accessible. Additionally, North Korea has tested the Hwasong-14 ballistic missile with a reported range of 8000 km, but studies propose it might reach up to 10,000 km, potentially reaching New York.
US missiles in ‘endgame’ mission	43
Iron dome	>90	Over 5000 interceptions, effective against rockets, mortars, artillery shells, aircraft, helicopters, and UAVs.
AIM-9X Block II Sidewinder	99.7	Significant advancements from the older AIM-9B version, including helmet-mounted display and improved manoeuvrability.
Stinger missile	>90	Supersonic speed, agility, accurate guidance and control system, used on Apache helicopters for air-to-air engagements.
Mini-UAV interceptor	98	Each mission lasts 30 min max, therefore the JUAVI life is limited by the cannon to 250 h.

The global landscape of missile capabilities reveals a spectrum of successes and challenges, reflecting a diverse landscape of strengths and challenges. Russian missiles have faced substantial setbacks, with failure rates ranging from 20 to 60%, notably affecting cruise missiles, raising concerns about reliability, quality control, and a fusing problem (Schneider, 2022). North Korea, in contrast, has conducted over 214 missile tests since 2012, boasting an 85% success rate and expanding its missile testing infrastructure, posing a formidable threat with missiles capable of reaching distances up to 10,000 km (Woo, 2023). It is initially asserted an 88% endgame success rate for missile defence programs based on 25 tests, but a more detailed analysis reveals a 71% success rate, with a midcourse-specific success rate of 61% and an overall midcourse system success rate of 41%, increasing marginally to 43% after the successful October 14 test of the ground-based midcourse system (Lewis & Li̇, 2020). On the defensive side, the Iron Dome, historically successful with over 5000 interceptions, has seen a recent decline in its success rate, prompting an investigation into its performance by the Israeli Defence Forces. The AIM-9X Block II program, a fifth-generation missile, achieves an impressive 99.7% success rate, showcasing advancements in target acquisition and manoeuvrability. Conversely, the Israeli news reports a decline in the Iron Dome’s success rate from 96 to 67%, prompting an inquiry into the system’s recent performance lapses (AIM-9X Blk II, 2019; Fitzpatrick, 1985; Israel Defence Forces Probe High Rate of Iron Dome Missile Interception Failures, 2023). The Stinger missile, a lightweight air defence system, maintains a reliability rate exceeding 90%, providing a significant operational advantage against various threats, including cruise missiles and aircraft (Stinger Missile, 2023). The reliability of the JUAVI, a jet unmanned aerial vehicle interceptor, is 98.14% for 50 consecutive fully automatic missions, with an MTBF of 1336 h. These systems underline the importance of simplicity, operational reliability, and adaptability, as exemplified by historical systems like the Sidewinder, in the development and assessment of modern missile systems. The diverse array of missile capabilities underscores the complex dynamics of global security and the ongoing pursuit of technological advancements in missile defence and offense.

6. Discussion

The JUAVI, with a very short TBO of 25 h, provides acceptable reliability and availability. It should also be considered as a possibility to dispose of the JUAVI after 25 h or 50 missions, given the low cost of the vehicle (below $100,000). These missions will be performed without maintenance, with 98 JUAVIs available in a fleet of 100. The solution to use small JUAVIs as ‘balloon killers’ seems to be very promising and cost-effective. However, the manual approach to such a high-speed mission requires very good skills from the pilot and has a relatively high probability of human errors (human factor). In the USAF Predator, 67% of the accidents were due to human factors compared to 47% due to airplane problems (Stinger Missile, 2023). Therefore, the manual approach is also subject to these unknown factors. General aviation airplanes stay below 18,288 m (60,000 ft). Therefore, any airplane that would intercept and destroy the enemy balloon has an extremely low probability of causing collateral damage or hitting the wrong target. It may be convenient to define a temporary no-fly area for the climb and descent of the JUAVI in the ATC airspace. Then, the balloon may autonomously find and destroy the balloon. The position of the slow-moving balloon is known at take-off; therefore, the interception would be relatively easy and safe.

There may be some challenges for this application. The study focuses on the use of a specific technology (JUAV) for countering surveillance balloons, but the effectiveness of this technology may be subject to technological advancements or changes in threat strategies. The operational conditions and capabilities of the JUAV may be limited to specific scenarios, and its effectiveness may vary under different environmental conditions or against different types of threats. The reliability analysis focuses on specific subsystems of the JUAV, but it may not cover all potential failure scenarios or provide a holistic view of the system’s reliability under various operational conditions. When all the other possible scenarios are taken into consideration, these can guide future studies that can be conducted in this field.

7. Conclusions

The third section of this paper introduces the reliability and availability of a modified 1:6 F104 RC jet model armed with an 84 mm airburst cannon. This very small UAV, with a length of <3 m and a weight of 45 kg, can fly up to 28,956 m (95,000 ft) and destroy a balloon with an up-to-down attack mission. The airplane will outperform the original F104C and be able to reach the target in <20 min. The choice of a life of ‘only’ 25 h (50 missions) without maintenance is dictated by the fact that RC models and parts are not designed for long life. An accurate pretesting of most parts will avoid the infant mortality part of the failure curve, giving a high system reliability. The availability of the JUAVI is estimated at 98 out of every 100 airplanes.

Comparisons with existing missile systems highlighted the importance of simplicity, operational reliability, and adaptability in the development and assessment of modern missile systems. The JUAVI demonstrated a reliability of 98.14% for 50 consecutive fully automatic missions, with an MTBF of 1336 h. Ultimately, this paper contributes valuable insights into the reliability assessment of a Jet Unmanned Aerial Vehicle designed for specific mission requirements, providing a foundation for further research and development in the field of missile systems.

Author contributions

Conceptualization and methodology, L.P., C.L., and M.S.; software, L.P.; validation, C.L. and M.S.; formal analysis and investigation, L.P., C.L., and M.S.; resources, L.P.; data curation, writing—original draft preparation, and writing—review and editing, M.S. and C.L.; visualization, C.L.; supervision, project administration, and funding acquisition, L.P.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Not applicable.

Figure 1. F104 1:6 single cannon (schematic). The cannon is the red cylinder at the root of the tail.

Additional information

Notes on contributors

Luca Piancastelli

Luca Piancastelli Full professor since 2000, currently his research is focused on both land and air vehicles, energy generation systems from renewable sources, advanced vehicle interfaces, autonomous driving system, restoration of monuments using additive technologies, innovative aids for the disabled and vehicle emergency systems.

Christian Leon-Cardenas

Christian Leon-Cardenas is a Ph.D. Student of the Department of Industrial Engineering, at Alma Mater Studiorum University of Bologna. Christian is involved in Methods for product design based on overall Optimization, Composites, 3D Printing applications and Augmented Reality studies.

Eugenio Pezzuti

Eugenio Pezzuti is an Associate Professor at University of Rome Tor Vergata | UNIROMA2 · Dipartimento di Ingegneria dell'Impresa

Merve Sali

Merve Sali is a Ph.D. Student of the Department of Industrial Engineering, at Alma Mater Studiorum University of Bologna. Merve is involved in Stylistic Design Engineering and Generative Design related studies.