Airbus manufactures the A300, A310, A318, A319, A320, A321, A330, A340, A350, A380, and other aircraft. Among them, the A350 series consists of A350-800, A350-900, and A350-1000, which are small to large and carry 270, 312, and 350 passengers, respectively, as shown in Figure 1. In this paper, the Airbus A350-900, a large twin-engine ultra-long-range, two-aisle wide-body airliner, is selected as the object of study. The cabin layout of this type of airliner is shown in Figure 2.

After collecting and sorting out, the main basic parameters of Airbus A350-900, such as fuselage length, fuselage height, cabin length, and cabin width, are shown in Table 1.

In the passenger cabin of Airbus A350-900, the seating arrangement is a typical three-class cabin arrangement consisting of business class, super economy class, and economy class. Its seating arrangement is as follows: rows 11–18 are business class, with a total of 32 seats; rows 31–33 are super economy class, with a total of 24 seats; rows 34–63 are economy class, with a total of 256 seats. The basic parameters of specific cabin seats are shown in Table 2 below.

When building a fire scenario, the known conditions and potential risks should be taken into account, the parameters of the airliner should be understood, the safety influencing factors in the cabin should be identified, and the location of the fire source should be reasonably set around the purpose of the study (10). Civilian airliners generally have a central fuel tank or additional central fuel tank near the center of gravity of the fuselage. As early as 1989, the Civil Aviation Administration of the United States carried out a physical simulation experiment of cabin fire (1). The simulation scenario is: after the plane fell, the huge impact force destroyed the structure of the plane, the fuel tank on one side of the wing ruptured, and the fuel leaked, causing the external fuel to form a pool fire and then cause a cabin fire through thermal radiation.

This paper investigates the fire caused by the impact of a civil airliner on the ground causing damage to the front section structure of the main body of the airliner, resulting in the spillage of aviation fuel carried in the additional central tank of the airliner due to the cracking of the cabin. Further, it investigates the impact of the fire on the evacuation of passengers in the passenger cabin of the airliner. The fire scenario is designed around the issue of cabin passenger evacuation, and the fire scenario is designed by considering the maximum probability principle and the least favorable principle through literature research on airliner fires.

The air pressure in the cabin was set to 1 standard atmosphere, and the initial temperature was set to 20°C. The Airbus A350-900 wide-body airliner studied in this paper is a large airliner with a long fuselage and a large cabin space. The passenger cabin is a relatively closed, narrow, and long cavity. The external environment affects only a small area near the door and does not significantly affect the fire development in the entire cabin. In the standard case, the property parameters of aviation fuel are in Table 3.

The cabin bulkhead structure of civil airliner is often composed of aluminum alloy, which is flame retardant. To simplify this paper, the passenger cabin bulkhead is set to aluminum, and other materials such as seats are also made flame-retardant, and in the standard case, the property parameters of aluminum are shown in Table 4.

The size of the fire source is set to a rectangular oil pool of 2.0 × 2.0 m. The fire location is in the middle of the second pair of evacuation exits connected to the passenger cabin near the front side of the nose, with the red square representing the fire source. In contrast, the 4 pairs of doors of the airliner are numbered for the convenience of description, with the doors near Qmax=m″ΔHcχπD2/4 the nose side numbered Exit 1 and Exit 2, the doors in the middle section numbered Exit 3 to Exit 6, and the doors near the tail side numbered Exit 7 and Exit 8, as shown in Figure 3.

According to the plane layout of the Airbus A350-900 airliner, the three-dimensional model of the cabin can be established after reasonably simplifying the model, as shown in Figure 4.

In aviation accidents, once the oil tank is damaged and leaked, it is easy to cause oil injection or dripping, and an oil pool will be formed. The fire caused by aircraft fuel leakage can be divided into oil vapor fire, jet fire, and oil pool fire according to the fuel combustion mode. In practice, the combustion mode of fuel oil may be single or composite, which constitutes the complexity of cabin fire. In this paper, to simplify the fire calculation, only pool fire is considered for calculation. The calculation formula for the maximum heat release rate of oil pool fire is as follows (11):

Where, ΔH_c is the combustion heat, kJ·kg⁻¹; χ is the combustion efficiency; D is the diameter of the oil pool, m; m″is the mass combustion efficiency per unit surface area, kg·m⁻²·s⁻¹; The calculation formula is:

In the formula, m∞″ is the progressive mass burning rate of large-scale oil pool fire, kg·m⁻²·s⁻¹; k′ is the effective absorption coefficient.

The fire source in the model is set in the middle of the connecting line between the second pair of hatch doors (Exit 3 and Exit 4) on the front side of the airliner. The rectangle with the size of 2.0 × 2.0 m is regarded as pool fire combustion. The fuel volume is assumed to be 0.1 m³, and the following formula can calculate its equivalent diameter:

Where S is the area of the liquid pool, according to the above, it can be calculated that the maximum heat release rate of the oil pool fire in this paper is 5.3011 MW.

The model in this paper sets the fire type as a very fast-growing fire, and the corresponding fire growth coefficient α = 0.1878 kW/s². According to the formula, it can be calculated that the time when the maximum heat release rate is reached is 168 s, and the combustion efficiency per unit surface area of the fuel is calculated to be 0.03424 kg·m⁻²·s⁻², from the total volume of fuel in the pool fire and the mass combustion efficiency of the fuel, it can be calculated that the burning time of the pool fire is 686 s. The decline of sump burning is a relatively complex process. The time when the fire source begins to decline is affected by the fire structure and other factors, and it is difficult to obtain an accurate solution. The decay rate of most fuels is 0.0001 m/s, and at some point the fuel is consumed to the point where the combustion cannot support the maximum heat release rate and begins to decay. At this time, the heat radiation is generally higher, and the fire decay process is faster. The schematic diagram of the fire scene's heat release rate development curve in this paper is shown in Figure 5.

FDS software simulates and calculates the CO concentration and flue gas diffusion according to the chemical equivalent combustion reaction of the specified fuel. Setting the co-production rate, flue gas production rate, and other parameters is necessary. Aviation fuel is a complex mixture, and aviation kerosene was chosen as a typical fuel for simplicity. Under standard conditions, its thermophysical parameters are as follows (14). The contents in brackets are the interpretation of the command line:

&REAC ID = ‘KEROSENE' (The fuel is called fuel oil)

FYI = ‘Kerosene, C_14 H_30, SFPE Handbook' (The molecular formula is C₁₄H₃₀)

MW_FUEL = 198.0 (The molar mass is 198)

NU_O2 = 21.5 (The coefficient of O₂ is 21.5)

NU_CO2 = 14.0 (The coefficient of CO₂ is 14)

NU_H2O = 15.0 (The coefficient of H₂O is 15)

EPUMO2 = 12700 (The energy released per unit mass of oxygen is 12700 kJ/kg)

CO_YIELD = 0.012 (The CO generation rate from combustion is 0.012)

SOOT_YIELD = 0.042/(The soot generation rate is 0.042)

The survey found that most passengers of medium and long-range wide-body airliners are adults. The average height of adults in China is about 1.70 m. It can be considered that the mouth and nose of passengers in the cabin during evacuation movement are about 1.65 m from the cabin ground. Therefore, this paper takes the cabin ground as the reference plane (H = 0 m), and the monitor is set at the height of H = 1.65 m, which is arranged concerning the layout of various types of seats in the cabin. A total of 312 observation points are set in the model. The positions of each observation point are shown in Figure 6. In the figure, the left side is the nose side of the aircraft, the observation points in business class are numbered from observation point 1 to observation point 32, and the observation points in super economy class are numbered from observation point 33 to observation point 56, and the observation points in economy class are numbered from observation point 57 to observation point 312. The numerical CO concentration, temperature, and visibility changes can be measured at each observation point. In the following chapters of this paper, these 312 observation points are selected to evaluate the safety indexes near all types of seats of the Airbus A350-900 wide-body airliner.

When a fire occurs, there will be many adverse factors that will cause harm to the human body. In this paper, the commonly used indicators in the fire are selected to analyze the available safe evacuation time of cabin fire. To ensure that passengers in the cabin can complete safe evacuation in the case of a cabin fire, the following specific quantitative standards are set as the critical value of danger for further analysis:

After investigation, it is found that the forced landing of passenger aircraft is often accompanied by incidental fire accidents, and the passenger aircraft is often in a horizontal and inclined state during the forced landing. Among them, the front landing gear of the passenger aircraft bears a huge impact force when landing, and the phenomenon of failure is not uncommon. When this happens, the passenger plane is often prone to lean forward, and the dynamic development of the fire at this time is also quite different from that when the passenger plane is in a horizontal and normal state. Therefore, it is necessary to carry out a comparative study of the two states.

When the airframe of the airliner is functioning normally, the front and rear landing gears are not damaged and can work normally, and the pilot drives the airliner properly so that the airliner is forced to land normally without tilting. As shown in Figure 7, the airliner remains horizontal after the forced landing. The length of the simulation time considers the specific simulation objects and scenarios. According to the relevant airworthiness regulations and practical experience of civil aviation aircraft, and this paper studies the passenger evacuation problem in the cabin fire scenario, the fire simulation time can be set as 200 s.

To obtain the ASET at each seat in the cabin, it is also necessary to calculate and analyze the time for the three indicators of CO concentration, temperature, and visibility to reach the dangerous critical value at the height of 1.65 m at the 312 observation points above the 312 seats under the two states of the airliner (15, 16). To select the time when any of the three safety evaluation indicators at each observation point first reaches the critical value of danger as the available safe evacuation time for the observation point.

When the airliner is in a horizontal state, the minimum time when any of the three safety indicators at each seat of the airliner reaches the critical value is calculated as the available safe evacuation time there, and the spatial distribution map of the airliner in the horizontal state is drawn, as shown in Figure 8.

From Figure 8, the distribution of ASET in the horizontal state of the passenger aircraft is analyzed as follows:

When the airliner is forced to land, if the pilot does not operate properly and the uncertainty of the forced landing position is added, the probability of collision damage and failure of the nose landing gear is very high. The Airbus A350-900 aircraft studied in this paper adopts the front three-point landing gear design. The disadvantage of this landing gear design is that the front landing gear bears a large load. If the aircraft adopts an abnormal landing attitude, it can easily cause damage and failure. The passenger will lean forward, as shown in Figure 9.

Similarly, the distribution of ASETs around the cabin when the aircraft is tilted forward can be calculated, as shown in Figure 10.

From Figure 10, the analysis of the ASET distribution in the forward tilt state of the passenger aircraft is as follows:

In the emergency evacuation scenario of a real aircraft forced landing, passengers need to be evacuated from the Exit of the cabin door through the escape slide. Therefore, the evacuation efficiency of passengers is closely related to the capacity of the escape slide. The evacuation flow of the escape slide is generally smaller than that of the cabin door, which will cause passengers to wait at the cabin door and reduce the evacuation efficiency of the people in the cabin. Therefore, it is necessary to analyze the evacuation capacity of the escape slide, but the Pathfinder software does not support pedestrians escaping by The slide carries out the motion simulation of evacuation behavior by sliding down, and there are few evacuation experiments on aircraft escape slides, so the relevant slide flow experimental data has not been obtained yet. In the actual situation, there is a large section in the process of people sliding down the slide in the suspended state and doing parabolic motion, so the actual friction force is ~0. Based on this, when the passenger cabin is in a horizontal state, this paper takes the aircraft escape slide equivalent to a smooth slope with a depression angle of 40° to carry out the ideal calculation, and its equivalent schematic diagram is shown in Figure 11.

In this paper, it is assumed that the length of the slide is fixed at 7 m, and the pedestrian is accelerated by gravity from the top of the slide with an initial velocity of 0 at the exit of the passenger cabin at a height of 4.5 m from the ground, and its sliding acceleration a = g × sin[(40 × π)/180], from this, it is calculated that the time for the passenger to fall from the top of the slide to the bottom is about 1.49 s. Assuming that the next passenger immediately falls from the top after the previous passenger leaves the slide, ignoring the influence of other factors, it is calculated that The corresponding flow is 1 divided by 1.49 equals 0.68 ped/s. When the cabin is tilted forward, the calculation method of the evacuation flow of slides with different angles from the ground is the same.

In this paper, it is assumed that the length of the slide is fixed at 7 m, and the pedestrian is accelerated by gravity at the exit of the passenger cabin at the height of 4.5 m from the ground at the top of the slide with an initial velocity of 0. From this, it is calculated that the time for a passenger to fall from the top to the bottom of the slide is about 1.49 s. Assuming that the next passenger immediately falls from the top after the previous passenger leaves the slide, ignoring the influence of other factors, the corresponding flow rate is calculated to be 0.68 ped/s. Therefore, to consider the influence of the evacuation capacity of the slide on the evacuation efficiency of the aircraft, it is necessary to perform an equivalent calculation on the width of the hatch according to the calculated flow of the slide. In this section, a single-exit room scenario is established to analyze the evacuation flow of pedestrians under different Exit width conditions. The room size is 10 × 10 m, the personnel density is uniformly set to 2 ped/m², and other personnel parameter settings are consistent with the above.

By setting different outlet widths (the width is shortened from 0.9 to 0.4 m in turn), the changes in the point section flow under the condition of different outlet widths are obtained, and the average evacuation section flow of different outlet widths is obtained according to the statistics of the outlet flow distribution, as shown in Figure 12.

It can be seen from the figure that when the width of the exit is 0.7 m, the average flow rate of the evacuation is close to 0.68 ped/s. Therefore, in this paper, considering the evacuation flow of passengers via the evacuation slide, the width of the exit of the passenger door is revised to 0.7. The equivalent width of m is used to carry out the evacuation simulation of the passenger aircraft when it is horizontal.

In the same way, after the airliner tilts forward, the distance between Exit 1 and Exit 2 from the ground is reduced to about 3 m. After calculation, the angle between the slide and the ground is 26°, and the time it takes for passengers to slide from the top to the bottom of the slide is About 1.80 s, the corresponding flow is calculated to be about 0.56 ped/s. The distance from Exit 7 and Exit 8 to the ground increases, about 5.8 m. After calculation, the angle between the slide and the ground is 56°, and the time for passengers to slide from the top of the slide to the bottom is about 1.31 s. Calculate the corresponding flow rate of About 0.80 ped/s. Therefore, this paper also considers the evacuation flow of passengers through the evacuation slide when the passenger aircraft is tilted forward, and the width of the passenger door Exit 1 and Exit 2 is corrected to an equivalent width of 0.5 m, and the width of the passenger door Exit 7 and Exit 8 is corrected to an equivalent width of 0.8 m, and the height of Exit 5 and Exit 6 does not change significantly when the passenger aircraft tilts forward, so the equivalent width of 0.7 m is still selected, and the following is the case after the passenger aircraft tilts forward? Personnel evacuation simulation.

By studying the internal structure of Airbus A350-900 wide-body passenger aircraft, evacuation dual aisles, door position and passenger seat distribution information, and related parameters, combined with other parameter settings above, the construction of the cabin personnel evacuation model is completed. The schematic diagram of the model is shown in Figure 13.

When the passenger cabin is in a horizontal state, the evacuation speed of personnel decreases to a certain extent in the middle of the evacuation, and the evacuation efficiency of passengers in the cabin is not high during some periods. The specific evacuation situation is shown in Figure 14.

As seen from Figure 14, in the middle of the evacuation, all passengers moved to the evacuation channel or near the evacuation exit, which caused a certain congestion phenomenon, resulting in a decrease in the evacuation rate. At about 25 s, the number of people evacuated at Exit 1 and Exit 2 no longer increased, and at about 55 s, The number of evacuees at Exit 7 and Exit 8 no longer increased, indicating that the above four exits were idle at the corresponding time, and the average utilization rate of the exits was not high. In the later evacuation period, the remaining passengers were evacuated only through Exits 5 and 6. The flow at the exit is shown in Figure 15. After 119.8 s, all 312 passengers were evacuated.

The necessary evacuation time consists of three parts: early warning time, passenger response time, and movement time (24–26). Figure 16 shows the distribution of the necessary evacuation time for passengers at different seats in the passenger cabin when the passenger cabin is in a horizontal state under the condition of full load. The higher the evacuation efficiency of the individual, the shorter the time required for evacuation and the smaller the safety risk borne by the movement of the safety exit.

Combined with the analysis of the evacuation situation in the horizontal state of the aircraft, it can be seen from Figure 16 that:

Corresponding to the PyroSim fire model of the passenger aircraft cabin in the forward tilt state, the cabin is tilted forward 5.6°, the nose side touches the ground, and the schematic diagram of the cabin model in the forward tilt state is shown in Figure 17.

Figure 18 shows the distribution of necessary evacuation time for passengers at different seats in the cabin under full load when the passenger cabin is tilted forward.

Combined with the analysis of the evacuation situation when the passenger aircraft is tilted forward, it can be seen from Figure 18 that:

When the airliner is in a horizontal state, there are 86 seats in the three classes of cabins, and the risk of safe evacuation is high. Figure 19 shows the distribution of the remaining evacuation time of passengers at different positions in the fully-loaded cabin.

Combined with the previous analysis of the cabin in a horizontal state, it can be seen from Figure 19 that:

When the airliner is tilted forward, there are 97 seats in the three classes of cabins, with a high risk of safe evacuation. Figure 20 shows the distribution of the remaining evacuation time for passengers at different positions in the fully-loaded cabin.

To sum up, when the cabin is tilted forward, the safe evacuation risk in the cabin is higher than the risk when the cabin is in a horizontal state, and the evacuation situation when the cabin is tilted forward, and a fire occurs is a more dangerous situation than the conventional horizontal state. Therefore, putting forward a targeted cabin evacuation optimization scheme has important reference significance in the engineering practice in the field of aircraft safety.

The evacuation pressure of Exit 5 and Exit 6 in the middle is high, and the utilization rate of the four cabin doors available for economy class passengers is uneven, resulting in different degrees of congestion, resulting in low overall evacuation efficiency. Given the above situation, it is proposed to add zoning guidance to rationally allocate the use of available evacuation exits instead of using the overall passenger to select exits based on the nearest principle. In two states, horizontal and forward, this section analyzes the optimization method in the horizontal state of the cabin.

The specific method is to divide the business class into area A, and guide all passengers in this area to evacuate to Exit 1 and Exit 2 during evacuation; divide the super economy class and the economy class set in the middle of the cabin into area B, and guide all passengers in this area during evacuation. Evacuate to Exit 5 and Exit 6; divide the remaining economy class into area C, and guide all passengers in this area to evacuate to Exit 7 and Exit 8 during evacuation. The specific guidance optimization scheme is shown in Figure 21.

The distribution of the remaining evacuation time after optimization is as follows:

It can be seen from Figure 22 that this scheme improves the evacuation efficiency of this part of the passengers by alleviating the congestion in the middle section of the cabin where the risk of cabin evacuation is the highest, effectively reducing the evacuation risk of passengers in area B, and optimizing the remaining cabin. Evacuation time distribution. The optimized cabin evacuation situation is shown in Figures 23, 24:

After adding evacuation guidance, the overall evacuation time of the cabin is reduced to 111.3 s compared with 119.8 s before optimization, and the reduction is 8.5 s. As seen from the above figure, the evacuation flow of Exit 5, Exit 6, Exit 7, and Exit 8 is more balanced, and the utilization rate of available exports improved. Passengers in area C no longer use Exit 5 and Exit 6 for evacuation, which reduces the evacuation pressure on the pair of exits and reduces the congestion of people near the pair of exits. After optimization, the evacuation situation in area A has not changed from the previous one. There are 6 passengers with a high risk of safe evacuation; 64 passengers in area B have a high risk of safe evacuation; 8 passengers in area C have a high risk of safe evacuation, with a total of 63 passengers. The risk of safe evacuation is high, a total of 8 people have been reduced compared with before optimization, and the reduction ratio is 9.3%.